Permalink
Cannot retrieve contributors at this time
Fetching contributors…
Cannot retrieve contributors at this time
| <?xml version="1.0" encoding="UTF-8"?><!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.1d1 20130915//EN" "JATS-archivearticle1.dtd"><article article-type="research-article" dtd-version="1.1d1" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><front><journal-meta><journal-id journal-id-type="nlm-ta">elife</journal-id><journal-id journal-id-type="hwp">eLife</journal-id><journal-id journal-id-type="publisher-id">eLife</journal-id><journal-title-group><journal-title>eLife</journal-title></journal-title-group><issn publication-format="electronic">2050-084X</issn><publisher><publisher-name>eLife Sciences Publications, Ltd</publisher-name></publisher></journal-meta><article-meta><article-id pub-id-type="publisher-id">03952</article-id><article-id pub-id-type="doi">10.7554/eLife.03952</article-id><article-categories><subj-group subj-group-type="display-channel"><subject>Research article</subject></subj-group><subj-group subj-group-type="heading"><subject>Neuroscience</subject></subj-group></article-categories><title-group><article-title>Widespread correlation patterns of fMRI signal across visual cortex reflect eccentricity organization</article-title></title-group><contrib-group><contrib contrib-type="author" corresp="yes" id="author-16098"><name><surname>Arcaro</surname><given-names>Michael J</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff2">2</xref><xref ref-type="corresp" rid="cor1">*</xref><xref ref-type="other" rid="par-1"/><xref ref-type="other" rid="par-2"/><xref ref-type="other" rid="par-3"/><xref ref-type="other" rid="par-4"/><xref ref-type="fn" rid="con1"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-16324"><name><surname>Honey</surname><given-names>Christopher J</given-names></name><xref ref-type="aff" rid="aff3">3</xref><xref ref-type="other" rid="par-2"/><xref ref-type="fn" rid="con2"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-16325"><name><surname>Mruczek</surname><given-names>Ryan EB</given-names></name><xref ref-type="aff" rid="aff4">4</xref><xref ref-type="other" rid="par-1"/><xref ref-type="other" rid="par-2"/><xref ref-type="other" rid="par-3"/><xref ref-type="other" rid="par-4"/><xref ref-type="fn" rid="con3"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-2436"><name><surname>Kastner</surname><given-names>Sabine</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff2">2</xref><xref ref-type="other" rid="par-1"/><xref ref-type="other" rid="par-2"/><xref ref-type="other" rid="par-3"/><xref ref-type="other" rid="par-4"/><xref ref-type="fn" rid="con4"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-16326"><name><surname>Hasson</surname><given-names>Uri</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff2">2</xref><xref ref-type="other" rid="par-2"/><xref ref-type="fn" rid="con5"/><xref ref-type="fn" rid="conf1"/></contrib><aff id="aff1"><label>1</label><institution content-type="dept">Princeton Neuroscience Institute</institution>, <institution>Princeton University</institution>, <addr-line><named-content content-type="city">Princeton</named-content></addr-line>, <country>United States</country></aff><aff id="aff2"><label>2</label><institution content-type="dept">Department of Psychology</institution>, <institution>Princeton University</institution>, <addr-line><named-content content-type="city">Princeton</named-content></addr-line>, <country>United States</country></aff><aff id="aff3"><label>3</label><institution content-type="dept">Department of Psychology</institution>, <institution>University of Toronto</institution>, <addr-line><named-content content-type="city">Toronto</named-content></addr-line>, <country>Canada</country></aff><aff id="aff4"><label>4</label><institution content-type="dept">Department of Psychology</institution>, <institution>Worcester State University</institution>, <addr-line><named-content content-type="city">Worcester</named-content></addr-line>, <country>United States</country></aff></contrib-group><contrib-group content-type="section"><contrib contrib-type="editor"><name><surname>Behrens</surname><given-names>Timothy</given-names></name><role>Reviewing editor</role><aff><institution>Oxford University</institution>, <country>United Kingdom</country></aff></contrib></contrib-group><author-notes><corresp id="cor1"><label>*</label>For correspondence: <email>marcaro@princeton.edu</email></corresp></author-notes><pub-date publication-format="electronic" date-type="pub"><day>19</day><month>02</month><year>2015</year></pub-date><pub-date pub-type="collection"><year>2015</year></pub-date><volume>4</volume><elocation-id>e03952</elocation-id><history><date date-type="received"><day>10</day><month>07</month><year>2014</year></date><date date-type="accepted"><day>21</day><month>01</month><year>2015</year></date></history><permissions><copyright-statement>© 2015, Arcaro et al</copyright-statement><copyright-year>2015</copyright-year><copyright-holder>Arcaro et al</copyright-holder><license xlink:href="http://creativecommons.org/licenses/by/4.0/"><license-p>This article is distributed under the terms of the <ext-link ext-link-type="uri" xlink:href="http://creativecommons.org/licenses/by/4.0/">Creative Commons Attribution License</ext-link>, which permits unrestricted use and redistribution provided that the original author and source are credited.</license-p></license></permissions><self-uri content-type="pdf" xlink:href="elife03952.pdf"/><abstract><object-id pub-id-type="doi">10.7554/eLife.03952.001</object-id><p>The human visual system can be divided into over two-dozen distinct areas, each of which contains a topographic map of the visual field. A fundamental question in vision neuroscience is how the visual system integrates information from the environment across different areas. Using neuroimaging, we investigated the spatial pattern of correlated BOLD signal across eight visual areas on data collected during rest conditions and during naturalistic movie viewing. The correlation pattern between areas reflected the underlying receptive field organization with higher correlations between cortical sites containing overlapping representations of visual space. In addition, the correlation pattern reflected the underlying widespread eccentricity organization of visual cortex, in which the highest correlations were observed for cortical sites with iso-eccentricity representations including regions with non-overlapping representations of visual space. This eccentricity-based correlation pattern appears to be part of an intrinsic functional architecture that supports the integration of information across functionally specialized visual areas.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.001">http://dx.doi.org/10.7554/eLife.03952.001</ext-link></p></abstract><abstract abstract-type="executive-summary"><object-id pub-id-type="doi">10.7554/eLife.03952.002</object-id><title>eLife digest</title><p>Imagine you are looking out over a scenic landscape. The image you perceive is actually made up of many different visual components—for example color and movement—that are processed across many different areas in a region of the brain called the visual cortex. An important question for neuroscience is how the visual system combines information from so many different areas to create a coherent picture of the world around us.</p><p>Many areas of the visual cortex have their own map of what we see (the visual field). These maps allow the brain to maintain its representation of the visual field as the information passes from one processing area to the next. Areas that process corresponding parts of the visual field are physically interconnected, and tend to be active at the same time, which suggests that they are working together in some way. In addition, areas of the visual cortex that process different sections of the visual field can be activated at the same time, but it is not clear how this works.</p><p>Here, Arcaro et al. used a technique called functional magnetic resonance imaging (fMRI) to image the brains of people as they watched movies and while they rested. The images showed that seemingly unrelated areas of the visual cortex respond in similar ways if they are processing sections of the visual field that are the same distance from the center of the person's gaze. For example, if you look directly at the center of a computer screen parts of the brain that process the top of the screen are active at the same time as parts that process the bottom.</p><p>Arcaro et al.'s findings suggest that the brain uses the distance from the center of our gaze to bring together information from different areas of the visual cortex. This offers a new insight into how the brain assembles the many pieces of the visual jigsaw to make a complete picture. Future work will investigate how the architecture of the visual cortex is able to support this coupling of different areas, and how it might influence our perception of the visual world.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.002">http://dx.doi.org/10.7554/eLife.03952.002</ext-link></p></abstract><kwd-group kwd-group-type="author-keywords"><title>Author keywords</title><kwd>fMRI</kwd><kwd>visual cortex</kwd><kwd>connectivity</kwd><kwd>topography</kwd><kwd>eccentricity</kwd></kwd-group><kwd-group kwd-group-type="research-organism"><title>Research organism</title><kwd>human</kwd></kwd-group><funding-group><award-group id="par-1"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100000001</institution-id><institution>National Science Foundation (NSF)</institution></institution-wrap></funding-source><principal-award-recipient><name><surname>Arcaro</surname><given-names>Michael J</given-names></name><name><surname>Mruczek</surname><given-names>Ryan EB</given-names></name><name><surname>Kastner</surname><given-names>Sabine</given-names></name></principal-award-recipient></award-group><award-group id="par-2"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100000025</institution-id><institution>National Institute of Mental Health (NIMH)</institution></institution-wrap></funding-source><principal-award-recipient><name><surname>Arcaro</surname><given-names>Michael J</given-names></name><name><surname>Honey</surname><given-names>Christopher J</given-names></name><name><surname>Mruczek</surname><given-names>Ryan EB</given-names></name><name><surname>Kastner</surname><given-names>Sabine</given-names></name><name><surname>Hasson</surname><given-names>Uri</given-names></name></principal-award-recipient></award-group><award-group id="par-3"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100000065</institution-id><institution>National Institute of Neurological Disorders and Stroke (NINDS)</institution></institution-wrap></funding-source><principal-award-recipient><name><surname>Arcaro</surname><given-names>Michael J</given-names></name><name><surname>Mruczek</surname><given-names>Ryan EB</given-names></name><name><surname>Kastner</surname><given-names>Sabine</given-names></name></principal-award-recipient></award-group><award-group id="par-4"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100000053</institution-id><institution>National Eye Institute (NEI)</institution></institution-wrap></funding-source><principal-award-recipient><name><surname>Arcaro</surname><given-names>Michael J</given-names></name><name><surname>Mruczek</surname><given-names>Ryan EB</given-names></name><name><surname>Kastner</surname><given-names>Sabine</given-names></name></principal-award-recipient></award-group><funding-statement>The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.</funding-statement></funding-group><custom-meta-group><custom-meta><meta-name>elife-xml-version</meta-name><meta-value>2.0</meta-value></custom-meta><custom-meta specific-use="meta-only"><meta-name>Author impact statement</meta-name><meta-value>Functional coupling between visual areas reflects supra-areal eccentricity organization distinct from overlapping receptive field connectivity during both rest conditions and naturalistic movie viewing.</meta-value></custom-meta></custom-meta-group></article-meta></front><body><sec sec-type="intro" id="s1"><title>Introduction</title><p>Within the visual system, a prominent organizational principle is that of retinotopy: adjacent neurons along the cortical surface typically receive input from adjacent points on the surface of the retina. Each retinotopic map can be divided along two orthogonal axes: polar angle (i.e., angular distance) and eccentricity (i.e., radial distance). Early electrophysiological recordings in monkeys and cats identified several representations of the contralateral visual field in and around the calcarine sulcus (<xref ref-type="bibr" rid="bib25">Daniel and Whitteridge, 1961</xref>; <xref ref-type="bibr" rid="bib95">Zeki, 1969</xref>; <xref ref-type="bibr" rid="bib3">Allman and Kaas, 1971</xref>; <xref ref-type="bibr" rid="bib86">Van Essen et al., 1984</xref>). Through the use of functional magnetic resonance imaging (fMRI), it has now become evident that the human visual system contains over two-dozen visual maps (for review of original mapping studies, see <xref ref-type="bibr" rid="bib78">Silver and Kastner, 2009</xref>; <xref ref-type="bibr" rid="bib89">Wandell and Winawer, 2011</xref>; <xref ref-type="bibr" rid="bib1">Abdollahi et al., 2014</xref>; <xref ref-type="bibr" rid="bib90">Wang et al., 2014</xref>).</p><p>The topographic organization of individual areas is thought to provide an infrastructure for the integration of information across areas and along the visual hierarchy (<xref ref-type="bibr" rid="bib53">Kaas, 1997</xref>; <xref ref-type="bibr" rid="bib88">Wandell et al., 2007</xref>). Anatomical studies in primates have demonstrated that neurons with overlapping receptive fields (RFs) are interconnected (<xref ref-type="bibr" rid="bib22">Cragg, 1969</xref>; <xref ref-type="bibr" rid="bib95">Zeki, 1969</xref>; <xref ref-type="bibr" rid="bib30">Van Essen and Zeki, 1978</xref>; <xref ref-type="bibr" rid="bib60">Maunsell and Van Essen, 1983</xref>). Similarly, fMRI connectivity studies in humans have demonstrated topographically-local correlations between regions with overlapping visual field representations (<xref ref-type="bibr" rid="bib46">Heinzle et al., 2011</xref>; <xref ref-type="bibr" rid="bib42">Haak et al., 2012</xref>; <xref ref-type="bibr" rid="bib18">Butt et al., 2013</xref>; <xref ref-type="bibr" rid="bib27">Donner et al., 2013</xref>; <xref ref-type="bibr" rid="bib41">Gravel et al., 2014</xref>; <xref ref-type="bibr" rid="bib70">Raemaekers et al., 2014</xref>). In addition, widespread functional correlation patterns have been observed across visual cortex in macaques (<xref ref-type="bibr" rid="bib57">Leopold et al., 2003</xref>; <xref ref-type="bibr" rid="bib87">Vincent et al., 2007</xref>) and humans (<xref ref-type="bibr" rid="bib65">Nir et al., 2006</xref>, <xref ref-type="bibr" rid="bib66">2008</xref>; <xref ref-type="bibr" rid="bib94">Yeo et al., 2011</xref>; <xref ref-type="bibr" rid="bib27">Donner et al., 2013</xref>). These patterns contain broad differences between foveal and peripheral cortex (<xref ref-type="bibr" rid="bib70">Raemaekers et al., 2014</xref>), though may also be tied to the fine-scale organization of individual retinotopic maps.</p><p>In this study, we used fMRI to investigate the relationship between the spatial pattern of correlated BOLD signal and the retinotopic organization of eight visual areas (V1–hV4, VO1–2, V3A–B). Correlation analyses were performed on data collected during task-free conditions (eyes-closed and fixation), and during movie viewing. Correlation patterns were consistent across subjects and experiments. In addition to finding patterns that support well-established anatomical connectivity between areas with overlapping RFs, our analyses revealed a widespread correlation pattern based on eccentricity representation, in which the BOLD signal was correlated in areas with non-overlapping visual field representations, but with matching eccentricity representations. This eccentricity-based correlation pattern was observed between upper and lower visual field representations, within and across visual areas, and between hemispheres. Moreover, correlation patterns were similar in the presence and absence of bottom-up visual input. Finally, the measured correlation pattern could not be accounted for by overlapping RFs, inter-hemisphere homotopic connections, anatomical distance, eye movements, subject motion, or physiological noise. Our results demonstrate that functional coupling between visual areas reflect both local and widespread topographical patterns. We propose that this widespread pattern is part of an intrinsic functional architecture of the visual system that could reflect eccentricity-dependent processing.</p></sec><sec sec-type="results" id="s2"><title>Results</title><p>Retinotopy and correlation patterns were characterized and compared within the visual systems of 14 participants. Retinotopic organization of the visual system was examined within the central 15° of visual space using a conventional travelling wave paradigm in which eccentricity and polar angle maps were collected to define visual areas V1, V2, V3, V3A–B, hV4, VO1–2 (see ‘Materials and methods’). The organization of correlation patterns was probed in two resting conditions in which participants were instructed to either (1) keep eyes open and maintain fixation on a centrally presented dot or (2) keep eyes closed for the duration of the run. In addition, we assessed the correlation patterns in 11 of these participants during a naturalistic viewing condition in which participants were instructed to attend to a movie, but maintain fixation on a centrally presented dot.</p><sec id="s2-1"><title>Exploratory seed analysis</title><p>To investigate the topography of functional correlations across visual cortex, Pearson correlation coefficients were computed between the timeseries of all surface data points (nodes) within the left and right visual cortices. For any individual node, correlation coefficients with all other nodes within visual cortex typically ranged between −0.10 and 0.75 in individual subjects (after cerebrospinal fluid and white matter signal regression). For illustration of the raw functional correlation results, we present fixation resting-state correlation maps for four example seed locations in subject S1 (<xref ref-type="fig" rid="fig1">Figure 1A</xref>), eyes shut resting-state correlation maps for four example seed locations in subject S2 (<xref ref-type="fig" rid="fig1">Figure 1B</xref>), and group average resting-state maps for four example seed locations (<xref ref-type="fig" rid="fig2">Figure 2A</xref>). In each of the four seed locations, the BOLD timeseries was sampled from a single node (black dot) within right dorsal V2, and the seed location was gradually shifted from foveal (<1.0°) to peripheral-most (∼11.5°) representations as defined from a separate eccentricity localizer experiment (rightmost panel).<fig-group><fig id="fig1" position="float"><object-id pub-id-type="doi">10.7554/eLife.03952.003</object-id><label>Figure 1.</label><caption><title>Seed-based correlations on resting state in individual subjects.</title><p>Correlation maps in both hemispheres of (<bold>A</bold>) subject S1 for resting fixation and (<bold>B</bold>) subject S2 for resting eyes shut at four seed locations (<1.0°, ∼2.5°, ∼5.5°, ∼11.5°; left to right) in dorsal V2 of the right hemisphere. For each seed, the strongest correlations (red / yellow) span several visuotopic areas within an eccentricity range roughly corresponding to that of the seed area (black dot) in both the ipsilateral and contralateral hemispheres. The correlations have a similar organization to eccentricity maps (far right). To facilitate visual comparisons between hemispheres, the left hemisphere images have been horizontally reflected. Solid white bars mark borders between visual field maps. White dashed bars outline three bands of iso-eccentricity.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.003">http://dx.doi.org/10.7554/eLife.03952.003</ext-link></p></caption><graphic xlink:href="elife03952f001"/></fig><fig id="fig1s1" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.03952.004</object-id><label>Figure 1—figure supplement 1.</label><caption><title>Seed-based correlations on resting state eyes shut in individual subjects.</title><p>Correlation maps in both hemispheres of (<bold>A</bold>) subject S3 at four seed locations (<1.0°, ∼2.5°, ∼5.5°, ∼11.5°; left to right) in ventral V3 of the right hemisphere and (<bold>B</bold>) of subject S4 in ventral V2 of the right hemisphere. For each seed, the strongest correlations (red / yellow) span several visuotopic areas within an eccentricity range roughly corresponding to that of the seed area (black dot). The correlations have a similar organization to eccentricity maps (far right). To enable comparisons between hemispheres, the left hemisphere images have been horizontally reflected. Conventions the same as <xref ref-type="fig" rid="fig1">Figure 1</xref>.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.004">http://dx.doi.org/10.7554/eLife.03952.004</ext-link></p></caption><graphic xlink:href="elife03952fs001"/></fig><fig id="fig1s2" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.03952.005</object-id><label>Figure 1—figure supplement 2.</label><caption><title>Seed-based correlations on resting state eyes shut in individual subjects.</title><p>Correlation maps in both hemispheres of (<bold>A</bold>) subject S5 at four seed locations (<1.0°, ∼2.5°, ∼5.5°, ∼11.5°; left to right) in dorsal V3 of the right hemisphere and (<bold>B</bold>)of subject S6 in ventral V2 of the right hemisphere . For each seed, the strongest correlations (red / yellow) span several visuotopic areas within an eccentricity range roughly corresponding to that of the seed area (black dot). The correlations have a similar organization to eccentricity maps (far right). To enable comparisons between hemispheres, the left hemisphere images have been horizontally reflected. Conventions the same as <xref ref-type="fig" rid="fig1">Figure 1</xref>.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.005">http://dx.doi.org/10.7554/eLife.03952.005</ext-link></p></caption><graphic xlink:href="elife03952fs002"/></fig></fig-group><fig id="fig2" position="float"><object-id pub-id-type="doi">10.7554/eLife.03952.006</object-id><label>Figure 2.</label><caption><title>Group average seed-based correlations on resting state data.</title><p>(<bold>A</bold>) Correlation maps in both hemispheres of group average data for resting fixation at four seed locations (<1.0°, ∼2.5°, ∼5.5°, ∼11.5°; left to right) in dorsal V2. For each seed, the strongest correlations (red / yellow) span several visuotopic areas within an eccentricity range roughly corresponding to that of the seed area (black dot) in both the ipsilateral and contralateral hemispheres. The correlations have a similar organization to eccentricity maps (far right). To facilitate visual comparisons between hemispheres, the left hemisphere images have been horizontally reflected. Solid white bars mark borders between visual field maps. White dashed bars outline three bands of iso-eccentricity. (<bold>B</bold>) Seed-based correlations plotted as a function of visual field representation for four seed locations. Eccentricity values are derived from a log-scaled stimulus (see ‘Materials and methods’). Black, dashed circles denote distance from fixation in visual degrees for each seed location.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.006">http://dx.doi.org/10.7554/eLife.03952.006</ext-link></p></caption><graphic xlink:href="elife03952f002"/></fig></p><p>In all cases, we observed a combination of topographically local and widespread correlation patterns (with respect to visual field representations). Each of the four seed locations was strongly correlated (red / yellow) with adjacent cortex. In addition, the strongest correlations with each seed extended across visual cortex and spanned several visual areas, from V1 to V3A–B, dorsally, and to VO1–2 ventrally. Strong correlations with nodes adjacent to the seed and within ipsilateral dorsal cortex likely reflect well-established connectivity based on overlapping visual field representations (see <xref ref-type="bibr" rid="bib46">Heinzle et al., 2011</xref>; <xref ref-type="bibr" rid="bib42">Haak et al., 2012</xref>; <xref ref-type="bibr" rid="bib18">Butt et al., 2013</xref>), but may also reflect the intrinsic spatial spread of the BOLD signal (<xref ref-type="bibr" rid="bib29">Engel et al., 1997</xref>; <xref ref-type="bibr" rid="bib69">Parkes et al., 2005</xref>). Interestingly, correlations were seen within <italic>both</italic> dorsal and ventral occipital cortex, comprised of lower and upper visual field representations respectively (<xref ref-type="fig" rid="fig1">Figure 1</xref>). Despite anatomical distance and representing a different part of visual space, the eccentricity representation (distance from fovea) of peak correlations within ventral occipital cortex corresponded to that of the seed location. These four seed locations also yielded comparable correlation patterns in the contralateral (left) hemisphere, comprised of right visual field representations. Local and widespread correlation patterns were observed in most individual subjects and the group average data for dorsal and ventral cortex seeds in V2 and V3, regardless of seeding near the horizontal or vertical meridians (see <xref ref-type="fig" rid="fig1s1 fig1s2">Figure 1—figure supplements 1, 2</xref> for additional individual subject data).</p><p>The BOLD signals in areas with eccentricity preferences similar to the seed were correlated in the presence and absence of visual input, even in cases where the spatial receptive fields (RFs) were non-overlapping (i.e., across lower and upper or right and left visual field representations). To summarize the group average V2-seeds correlation results, we projected the correlation maps in <xref ref-type="fig" rid="fig2">Figure 2A</xref> into visual field coordinates, and averaged across areas V1, V2, V3, V3A–B, hV4, and VO1–2 (<xref ref-type="fig" rid="fig2">Figure 2B</xref>). The correlation patterns highlight both visuotopically local and widespread correlation patterns (<xref ref-type="fig" rid="fig2">Figure 2B</xref>). Peak correlations (red) were evident in parts of the visual field around each seed location with strong correlations (red / yellow) also extending across the visual field along an eccentricity ring corresponding to that of the seed location. Similar local and widespread eccentricity-based correlation patterns were observed in data from the movie viewing experiment (<xref ref-type="fig" rid="fig3">Figure 3</xref>). Individual subject and group average correlation patterns were similar to previously reported group average correlation patterns (<xref ref-type="bibr" rid="bib94">Yeo et al., 2011</xref>). Below, we formally tested the relation of eccentricity representations to the spatial pattern of correlated BOLD signal between individual brain areas and across tasks.<fig id="fig3" position="float"><object-id pub-id-type="doi">10.7554/eLife.03952.007</object-id><label>Figure 3.</label><caption><title>Group average seed-based correlations on movie viewing data.</title><p>(<bold>A</bold>) Correlation maps in both hemispheres of group average data for movie viewing at four seed locations (<1.0°, ∼2.5°, ∼5.5°, ∼11.5°; left to right) in dorsal V2. For each seed, the strongest correlations (red / yellow) span several visuotopic areas within an eccentricity range roughly corresponding to that of the seed areas (black dot). The correlations have a similar organization to the eccentricity maps (far right column of <xref ref-type="fig" rid="fig2">Fig. 2</xref>). Conventions the same as <xref ref-type="fig" rid="fig2">Figure 2</xref>. (<bold>B</bold>) Seed-based correlations plotted as a function of visual field representation for the four seed locations. Black, dashed circles denote the seed area.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.007">http://dx.doi.org/10.7554/eLife.03952.007</ext-link></p></caption><graphic xlink:href="elife03952f003"/></fig></p></sec><sec id="s2-2"><title>Eccentricity binning</title><p>To characterize the widespread eccentricity-based correlation pattern that was observed in raw correlation maps, individual subject timeseries data were grouped by visual area and then partitioned into 12 bins between 0.50° and 12.50° of eccentricity. Data binning was used as a form of averaging to increase signal-to-noise within bins, while preventing the spread of signal between bins. Within-subject pairwise correlations were calculated between the mean timeseries of all bins for visual areas V1, V2, V3, hV4, V3A–B (combined), and VO1–2 (combined), each of which has sufficient surface area to allow for fine-scale binning of eccentricity data. Areas V1–V3 were separated into quadrants for most analyses. No additional extrastriate areas were included in these analyses.</p><p>Correlations between dorsal and ventral bins of V1, V2, and V3 were strongest for iso-eccentricity representations. These correlations were seen across the vertical and horizontal meridians at both foveal and peripheral-most bins. We first illustrate the binning analysis by examining the raw correlations between the ventral and dorsal quadrants in V3 (<xref ref-type="fig" rid="fig4">Figure 4</xref>). This pair was chosen for illustration because correlations computed across quadrants are less susceptible to the influence of overlapping receptive fields (RFs) or cortical proximity. For regions that represent the same visual quadrant (e.g., the dorsal portions of visual areas V1, V2, and V3), it is difficult to dissociate effects due to shared eccentricity representations from overlapping receptive fields. Similarly, for adjacent dorsal regions, the shortest cortical distances are typically at corresponding eccentricity representations, making it difficult to dissociate eccentricity-related correlations from cortical (and volumetric) distance-based correlations. The dorsal and ventral portions of visual area V3 (as well as V2), however, only anatomically border each other at the fovea, represent different parts of the visual field (lower and upper, respectively), and thereby minimize both the overlapping receptive field and anatomical adjacency concerns. Thus, correlation analyses between these areas allowed us to test for widespread eccentricity-based correlation patterns in cases where effects of cortical distance and overlapping RFs are minimal. As seen for subject S4, the exploratory seed-based analyses showed strong correlations (red/yellow) between dorsal and ventral V3 at corresponding eccentricity representations (<xref ref-type="fig" rid="fig4">Figure 4A</xref>). Binned data showed the same correlation pattern. For each of the selected dorsal V3 bins, correlations with ventral V3 bins were strongest at (and around) corresponding iso-eccentricities (<xref ref-type="fig" rid="fig4">Figure 4B</xref>). For example, peripheral-most V3d bins correlated most with peripheral-most V3v bins, and the correlations gradually decreased moving towards the foveal bin (<xref ref-type="fig" rid="fig4">Figure 4B</xref>, right). The entire correlation matrix for all eccentricity locations between dorsal and ventral V3 (<xref ref-type="fig" rid="fig4">Figure 4C</xref>, center panel), revealed a similar pattern, where correlations were strongest among ventral and dorsal bins with iso-eccentricity representations (i.e., the diagonal), and were weaker for bins with large radial distances (e.g., foveal vs peripheral-most).<fig id="fig4" position="float"><object-id pub-id-type="doi">10.7554/eLife.03952.008</object-id><label>Figure 4.</label><caption><title>Illustration of eccentricity binning correlations on resting state.</title><p>(<bold>A</bold>) Correlation maps in left hemisphere of subject S4 for 3 seed bin locations (<0.5–0.84°, 3.71–4.63°, 10.36–12.50°; left to right) in dorsal V3 and subject S4's eccentricity map (right). Grayscale dots mark approximate seed bin locations. (<bold>B</bold>) Correlations with all ventral V3 bins are plotted for the three dorsal seed locations. The strongest correlations between dorsal and ventral V3 were at, and around, iso-eccentricity representations. (<bold>C</bold>) The entire correlation matrix for all eccentricity locations between dorsal and ventral V3 (center) revealed a similar pattern where correlations were strongest at or near iso-eccentricity (i.e., the diagonal), and weaker for bins with large radial distances (e.g., foveal vs peripheral-most). Radial distance (left) was strongly correlated with the measured group connectivity (r = 0.84) and was uncorrelated to the cortical distance (right).</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.008">http://dx.doi.org/10.7554/eLife.03952.008</ext-link></p></caption><graphic xlink:href="elife03952f004"/></fig></p><p>The pattern of correlations reflects radial distance, not anatomical distance. We computed the (ranked) radial distance between eccentricity bins. A radial bin distance of zero corresponds to bins with the same (iso-) eccentricity representation, while a radial bin distance of 11 corresponds to bins furthest from each other on the eccentricity axis (i.e., foveal vs peripheral-most). <xref ref-type="fig" rid="fig4">Figure 4C</xref> provides the spatial pattern of correlations between ventral and dorsal quadrants in area V3 as well as the predicted patterns based on radial distance (left gray panel) and based on cortical distance (right gray panel). Note that both the radial bin distance and the cortical distance (as well as an overlapping RF model) predict strong correlations between foveal bins (<xref ref-type="fig" rid="fig4">Figure 4C</xref>). However, whereas cortical distance predicts that correlations will be weaker between cortically–distant iso-eccentricity bins (e.g., ventral and dorsal V3 peripheral-most), the radial bin distance predicts a strong correlation between iso-eccentricities (<xref ref-type="fig" rid="fig4">Figure 4C</xref>). Indeed, the group average bin data was strongly correlated with the predicted eccentricity pattern (<italic>r</italic> = 0.84) and was not positively correlated with cortical distance (<italic>r</italic> = −0.04) or volume distance (<italic>r</italic> = −0.19) (<xref ref-type="fig" rid="fig4">Figure 4C</xref>). Individual subject matrices were also correlated with radial distance (mean <italic>r</italic> = 0.45), and the Fisher-transformed correlations significantly differed from zero across subjects (one-sample t-test, t(13) = 6.37, p < 0.0001).</p><p>Correlations with radial bin distance were apparent within and across all visual areas tested, including areas with overlapping and non-overlapping visual field representations (both within and across hemispheres). We computed intra- and inter-hemisphere correlations as a function of eccentricity bin within and across visual areas V1, V2, V3, hV4, V3A–B (combined), and VO1–2 (combined) in each condition (resting eyes closed, resting eyes open, and movie fixation; <xref ref-type="fig" rid="fig5">Figure 5</xref>). In <xref ref-type="fig" rid="fig5">Figure 5</xref>, gray frames denote comparisons between regions with overlapping visual field representations while black frames denote comparisons between regions with minimal or no overlap in visual field representations (i.e., across ventral and dorsal visual fields and across hemispheres). For all comparisons in each condition the strongest correlations (red/yellow) were consistently between bins at comparable eccentricity locations (i.e., along the diagonal of each sub-matrix in <xref ref-type="fig" rid="fig5">Figure 5</xref>). This was the case even for correlations between dorsal and ventral visual portions of V1, V2, and V3, and between hemispheres, which represent mostly discrete parts of visual space. Across area pairs, the average individual subject correlation coefficient with radial distance ranged between 0.17 and 0.80 with a median of 0.39 and an interquartile range of 0.12. Fisher-transformed individual subject correlations significantly differed from zero for all pairs in each experiment (one-sample t-test, t(13) > 2.31, <italic>p</italic>s < 0.05, FDR-corrected) except for two pairs of inter-hemisphere correlations with hV4 in the movie viewing condition (<italic>p</italic>s = 0.13 and 0.07). In all experiments, individual subject correlations with radial distance were generally strongest for early visual areas (V1, V2, and V3). This could reflect the relatively smaller, more spatially focal receptive fields in early visual areas, but also the larger surface area (i.e., more data points).<fig id="fig5" position="float"><object-id pub-id-type="doi">10.7554/eLife.03952.009</object-id><label>Figure 5.</label><caption><title>Correlation matrices for resting state and movie viewing conditions.</title><p>Intra- (bottom-left matrix triangle) and inter-hemisphere (top-right matrix triangle) correlations are shown for all pairwise comparisons between visual areas V1, V2, V3, hV4, VO1-2, and V3A-B for resting fixation (left matrix), resting eyes shut (center matrix), and movie viewing (right matrix) experiments. For each pair of visual areas, the strongest correlations (red) are at corresponding eccentricity bins and its neighbors (diagonal in each sub-matrix). Grey and black boxes bound area pairs with overlapping and non-overlapping visual field representations, respectively.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.009">http://dx.doi.org/10.7554/eLife.03952.009</ext-link></p></caption><graphic xlink:href="elife03952f005"/></fig></p></sec><sec id="s2-3"><title>Effect of radial distance</title><p>For each area pair, correlation coefficients were positive at 0–radial distance, and decreased at larger radial distances across the visual field. To assess the general effect of radial distance for each area pair, eccentricity bin correlations were grouped as a function of radial bin distance and averaged. Averaged correlations were plotted relative to iso-eccentricity (<xref ref-type="fig" rid="fig6">Figure 6A</xref>, left panel). In each quadrant, the radial mid-arc (<xref ref-type="fig" rid="fig6">Figure 6A</xref>, white outline) corresponds to the average correlation of the 0-distance bin (iso-eccentricity). Radial arcs further away from the mid-arc correspond to larger radial distances. For each radial plot, for example, V2–V3, radial arcs inside of the mid-arc illustrate correlations between relatively more foveal representations of V2 and relatively more peripheral representations of V3. Radial arcs outside of the mid-arc illustrate correlations between relatively more peripheral representations of V2 and relatively more foveal representations of V3. Inner and outer radial points will be identical for within area comparisons (e.g., V2–V2), but could differ for between area comparisons (e.g., V2–V3). As illustrated by resting state correlations within and between areas V2 and V3 (<xref ref-type="fig" rid="fig6">Figure 6A</xref>), iso-eccentricity bins were positively correlated, regardless of visual field quadrant. This was observed in all experiment conditions and area comparisons, and individual subject Fisher-transformed correlations significantly differed from zero (<italic>p</italic>s < 0.05, FDR corrected). There were clear differences in overall correlation magnitude between quadrants. As expected, correlation coefficients decreased at larger radial distances for foveal and peripheral-most bins in each quadrant, regardless of this magnitude difference (<xref ref-type="fig" rid="fig6">Figure 6A</xref>).<fig-group><fig id="fig6" position="float"><object-id pub-id-type="doi">10.7554/eLife.03952.010</object-id><label>Figure 6.</label><caption><title>Intra-run eccentricity-based connectivity analyses for resting-state and movie viewing conditions.</title><p>(<bold>A</bold>) Radial bin distance correlation plots between resting-state data bins for within quadrant/within hemisphere (upper left), within quadrant/between hemisphere (upper right), between quadrant/within hemisphere (lower left), and between quadrant/between hemisphere (lower right) comparisons. In each quadrant, the mid-radial arc (white outline) corresponds to the average correlation at iso-eccentricity with outer and inner arcs corresponding to average correlations at increasingly larger radial distances. (<bold>B</bold>) Average individual subject correlations are plotted for areas V2 and V3 as a function of radial distance between the upper and lower visual fields (top), right and left visual fields (middle), as well as across both left and right and upper and lower visual fields (bottom) for all conditions. Correlations are normalized to the average correlations at iso-eccentricity (0-radial distance). All correlations were strongest at the 0-radial distance and steadily decreased at larger distances for all conditions.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.010">http://dx.doi.org/10.7554/eLife.03952.010</ext-link></p></caption><graphic xlink:href="elife03952f006"/></fig><fig id="fig6s1" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.03952.011</object-id><label>Figure 6—figure supplement 1.</label><caption><title>Intra-run eccentricity-based connectivity analyses for resting-state and movie viewing conditions without meridian data.</title><p>Average individual subject correlations are plotted for areas V2 and V3 as a function of radial distance between the upper and lower visual fields (top), right and left visual fields (middle), as well as across both left and right and upper and lower visual fields (bottom) for all conditions. Meridian data were removed prior to binning. Data corresponding to 60° of polar angle centered on the horizontal meridian was removed for comparisons between dorsal and ventral regions. Data corresponding to 90° of polar angle centered on each vertical meridian was removed for inter-hemisphere comparisons. All correlations were strongest at the 0-radial distance (iso-eccentricity) and steadily decreased at larger distances for all conditions.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.011">http://dx.doi.org/10.7554/eLife.03952.011</ext-link></p></caption><graphic xlink:href="elife03952fs003"/></fig><fig id="fig6s2" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.03952.012</object-id><label>Figure 6—figure supplement 2.</label><caption><title>Intra-run angular-based connectivity analyses for resting-state and movie viewing conditions.</title><p>(<bold>A</bold>) Average individual subject correlations are plotted for areas V2 and V3 as a function of angular distance within an area (top), between the upper and lower visual field (second row), right and left visual fields (third row), as well as across both left and right and upper and lower visual fields (bottom) for all conditions. (<bold>B</bold>) Average individual subject correlations are plotted for mirror symmetric connections between upper and lower visual fields (top) and between right and left visual fields (bottom). See <xref ref-type="fig" rid="fig6">Figure 6</xref> for conventions.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.012">http://dx.doi.org/10.7554/eLife.03952.012</ext-link></p></caption><graphic xlink:href="elife03952fs004"/></fig><fig id="fig6s3" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.03952.013</object-id><label>Figure 6—figure supplement 3.</label><caption><title>Real data vs artificial data.</title><p>Comparison of average individual subject correlations for real (blue) and artificial (red) data. Data are plotted for areas V1, V2, and V3 as a function of radial distance between the upper and lower visual fields (top), right and left visual fields (middle), as well as across both left and right and upper and lower visual fields (bottom) for all conditions. The pattern of correlations between real and artificial data differed for all comparisons. For the real data, correlations decreased as a function of radial distance in every comparison. For the artificial data, correlations decreased as a function of radial distance only for within quadrant and (partially) for between anatomically adjacent areas. For comparisons between dorsal and ventral regions and between hemispheres, artificial data correlations did not vary as a function of radial distance. Even for within quadrant, the slope of the correlations across radial distances was markedly different between real and artificial data. These plots demonstrate that the effect of radial distance on correlation strength cannot be explained by spatial correlations due to intrinsic properties of BOLD imaging or our preprocessing as these were preserved in the artificial data.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.013">http://dx.doi.org/10.7554/eLife.03952.013</ext-link></p></caption><graphic xlink:href="elife03952fs005"/></fig></fig-group></p><p>The relative decrease in correlation strength as a function of radial distance was similar across the visual field and across experiments (<xref ref-type="fig" rid="fig6">Figure 6B</xref>). We plotted the average correlation as a function of radial bin distance (relative to iso-eccentricity; see ‘Materials and methods’) between the upper and lower visual fields, right and left visual fields, as well as across both left and right and upper and lower visual fields separately (see graphic illustrations in <xref ref-type="fig" rid="fig6">Figure 6B</xref> left panels). Data are presented for comparisons between quadrants within and between areas V2 and V3. As discussed in the eccentricity binning section, these comparisons were chosen for illustration because overlap in visual field representation is minimal. Consistent with the radial arc plots (<xref ref-type="fig" rid="fig6">Figure 6A</xref>), correlations between dorsal and ventral portions of V2 and V3 were strongest at 0–radial distance, and steadily decreased at larger radial distances for all conditions (<xref ref-type="fig" rid="fig6">Figure 6</xref>; top row). We observed similar correlation patterns as a function of radial bin distance across hemispheres, both along the horizontal plane (e.g., RH V2v and LH V2v; middle row) and diagonally across the upper and lower visual fields (e.g., RH V2v and LH V2d; bottom row). The patterns were observed even after removal of horizontal and vertical meridian data, further demonstrating that these effects are unlikely to be driven by local, overlapping visual field representations (<xref ref-type="boxed-text" rid="box1">Box 1</xref>; <xref ref-type="fig" rid="fig6s1">Figure 6—figure supplement 1</xref>). The decrease in correlation coefficients at larger radial bin distances yielded negative slopes from linear fits in each subject for all V2 and V3 pairs. For any given pair, negative slopes were similar across condition. Across V2 and V3 pairs, the average individual subject slope of the linear fit ranged between −0.22 and −0.49. Further, the average individual subject slope for each area pair was greater than 97.5% of a permuted distribution (mean 97.5% across areas, conditions = −0.06) where the labels of radial distances were scrambled before deriving individual subject radial distance correlations and slopes. As seen in the radial plots (<xref ref-type="fig" rid="fig6">Figure 6B</xref>), the slopes were similar across regional comparisons, though were slightly shallower for comparisons that spanned both quadrants and hemispheres (i.e., diagonal; bottom row). Finally, we observed these effects during resting fixation and eyes closed conditions, as well as during the processing of the movie, attesting to the robustness of the effect, and excluding many potential confounds (see ‘Control analyses’). Similar results were found in all other comparisons between visual areas. The magnitude of correlation coefficients linearly decreased as a function of radial bin distance for all area comparisons (V1, V2, V3, hV4, VO1–2, V3A–B), though slopes were generally steeper for comparisons between V1, V2, and V3. Across all areas, average individual subject slopes ranged between −0.14 and −0.80, and each was greater than 97.5% of the permuted distribution (mean 97.5% across areas, conditions = −0.09).</p><boxed-text id="box1"><object-id pub-id-type="doi">10.7554/eLife.03952.014</object-id><label>Box 1.</label><caption><title>Widespread connectivity is not dependent on meridian representations.</title><p>Though we divided data into quadrants that represent distinct parts of visual space, monosynaptic connections have been observed between neurons in dorsal and ventral visual cortex with receptive fields overlapping at the horizontal meridian (<xref ref-type="bibr" rid="bib52">Jeffs et al., 2009</xref>; see also; <xref ref-type="bibr" rid="bib96">Zeki, 1971</xref>; <xref ref-type="bibr" rid="bib81">Stepniewska and Kaas, 1996</xref>; <xref ref-type="bibr" rid="bib31">Felleman et al., 1997</xref>; <xref ref-type="bibr" rid="bib37">Gattass et al., 1997</xref>) and between hemispheres at the vertical meridian (<xref ref-type="bibr" rid="bib30">Essen and Zeki, 1978</xref>; <xref ref-type="bibr" rid="bib64">Newsome and Allman, 1980</xref>; <xref ref-type="bibr" rid="bib23">Cusick et al., 1984</xref>; <xref ref-type="bibr" rid="bib55">Kennedy et al., 1986</xref>; <xref ref-type="bibr" rid="bib2">Abel et al., 2000</xref>). Despite constituting (at most) a minimal portion of the signal in our bin data, we tested whether the observed eccentricity-based correlations between dorsal and ventral cortex and between the hemispheres were driven by the inclusion of these overlapping representations at the horizontal and vertical meridians, respectively, by removing data from the meridians and re-running the binning analyses. For correlations between upper and lower visual fields (dorsal and ventral cortex), we cut out 60° of polar angle centered on the horizontal meridian, which spared 60° of polar angle in the upper and lower visual fields near the vertical meridians. For correlations between the right and left visual fields (right and left hemispheres), we cut out 90° of polar angle centered on each vertical meridian, which sparred 90° of polar angle centered on the horizontal meridian in each hemisphere. Consistent with the original analyses, average individual subject correlations between dorsal and ventral portions of V2 and V3 as well as between right and left hemispheres were strongest at the 0–radial distances (iso-eccentricities), and linearly decreased at larger distances for all conditions (<xref ref-type="fig" rid="fig6s1">Figure 6—figure supplement 1</xref>). In general, the slopes of correlations were marginally shallower, but did not significantly differ from the full data reported in <xref ref-type="fig" rid="fig5">Figure 5</xref> (<italic>p</italic>s > 0.05, FDR corrected). If overlapping meridian representations facilitated the correlation patterns, removal of this data should have eliminated the correlation effects. These data further support the interpretation that the observed correlations with radial distance are not due to overlapping representations of visual space.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952">http://dx.doi.org/10.7554/eLife.03952.014</ext-link></p></caption></boxed-text></sec><sec id="s2-4"><title>Effect of angular distance</title><p>For comparisons between dorsal areas, between ventral areas, and between mirror-symmetric (across vertical meridian) areas, correlations were positive at 0–angular distance, and decreased at larger angular distances (<xref ref-type="fig" rid="fig6s2">Figure 6—figure supplement 2</xref>). To assess the general effect of angular distance for each area pair, data were grouped into 12 equally spaced polar angle bins. Correlations between bins were computed and grouped as a function of angular bin distance. For within-quadrant, within-hemisphere and mirror symmetric (across vertical meridian) within-quadrant, between-hemisphere comparisons, correlations steadily decreased at larger angular distances. Taken together, these results suggest that angular connectivity reflects overlapping RF between areas and mirror-symmetric connections between hemispheres. No effect of angular distance was observed between dorsal and ventral comparisons based on actual angular distance or when reflecting across the horizontal meridian (i.e., mirror symmetry). As with any negative finding, we cannot conclude with certainty that such angular-based connectivity does not exist, though, to our knowledge, no study has shown angular connectivity between regions with non-overlapping RFs.</p></sec><sec id="s2-5"><title>Stimulus-dependent and independent correlations</title><p>For each area pair, the spatial pattern of correlations (across eccentricity bins) was similar between experimental conditions. Across all areas, the average individual subject correlation of these spatial patterns between experimental conditions was 0.86 for intra-hemisphere comparisons and 0.75 for inter-hemisphere comparisons. The average individual subject correlation between resting-state and movie viewing conditions was 0.84 for intra-hemisphere comparisons and 0.71 for inter-hemisphere comparisons. Overall, these data are consistent with the exploratory seed based analyses, and suggest that the spatial pattern of correlations between visual areas tested were similar in the presence and absence of a strong bottom-up input.</p><p>Given the similarity of correlation matrices across all three conditions, this widespread eccentricity-based correlation pattern appears to reflect an eccentricity bias that is inherent to the organization of the visual system (i.e., stable during rest) and may support processing during active perception of the visual environment (e.g., during movie viewing). We conducted correlations between individual runs to directly test whether the patterns observed during the movie viewing condition reflected processing of the stimulus input. Intrinsic neural dynamics during the resting and movie conditions that are not related to the processing of visual stimuli, as well as non-neuronal artifacts (e.g., respiratory rate, motion), can only influence the pattern of correlations within each run, but should not induce correlations between runs. Indeed, inter-run Fisher-transformed correlations on resting state data showed no effect of radial distance (all <italic>p</italic>s > 0.05), validating the assumption that noise correlations should not be reliable across runs (<xref ref-type="fig" rid="fig7">Figure 7</xref>, blue and black lines). In contrast, inter-run correlations during movie watching showed radial distance effects similar to intra-run correlations (<xref ref-type="fig" rid="fig7">Figure 7</xref>, red lines). Inter-run correlations were statistically significant for 96% of all comparisons (<italic>p</italic>s < 0.05, FDR corrected). The slopes of coefficients as a function of radial distance for movie data were, generally, weaker for the inter-run relative to intra-run analyses (compare <xref ref-type="fig" rid="fig6 fig7">Figures 6B, 7</xref>). However, the inter-run analyses inherently had less data and spatial attention was not controlled, making the differences between intra- and inter-run analyses difficult to interpret. The inter-run results indicate that the observed widespread eccentricity-based correlations can reflect the processing of the incoming information during viewing of real life stimuli, and further suggest that these correlations patterns are unlikely to be driven by non-neuronal artifacts.<fig id="fig7" position="float"><object-id pub-id-type="doi">10.7554/eLife.03952.015</object-id><label>Figure 7.</label><caption><title>Inter-run eccentricity-based connectivity analyses for resting-state and movie viewing conditions.</title><p>For the movie viewing condition, average individual subject inter-run correlations between hemispheres as well as between dorsal and ventral portions of V2 and V3 were strongest at the 0-radial distance (iso-eccentricity) and steadily decreased at larger distances. For resting state data, correlations did not vary as a function of radial distance. See <xref ref-type="fig" rid="fig6">Figure 6</xref> for conventions.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.015">http://dx.doi.org/10.7554/eLife.03952.015</ext-link></p></caption><graphic xlink:href="elife03952f007"/></fig></p></sec><sec id="s2-6"><title>Control analyses</title><p>We performed a variety of control analyses to rule out possible confounds. The observed widespread eccentricity-based correlation pattern could not be attributed to non-neuronal artifacts (e.g., physiological noise, motion, BOLD spatial autocorrelation, eye movements). Global artifacts did not drive our results as correlation patterns were strengthened by the removal of non-neuronal signals (i.e., motion and white matter). The eccentricity-based correlation pattern was not due to inherent spread of the BOLD signal, anatomical distance, or any biases in our analyses, as artificial data that preserved such biases were not correlated with radial distance (see <xref ref-type="boxed-text" rid="box2">Box 2</xref>; <xref ref-type="fig" rid="fig6s3">Figure 6—figure supplement 3</xref>). Subject motion also could not explain the observed correlation patterns. We minimized the overall influence of movement in our data by using highly trained MRI subjects and by removing signal correlated with subject motion via regression. Further, the strength of eccentricity correlation did not significantly co-vary with degree of subject movement. We performed Spearman rank correlations between each subject's total amount of movement and the slope of his or her radial bin distance coefficients (i.e., effect of eccentricity on correlation strength). There was no significant correlation across subjects between the mean slope (averaged across all area pairs) and any of the six motion parameters (all rs < 0.24, <italic>p</italic>s > 0.05). In addition, there were no significant correlations across subjects between the slope and total amount of movement for any individual area pair (<italic>p</italic>s > 0.05, FDR-corrected). These analyses strongly suggest that the observed correlation patterns were not driven by subject motion. It is unlikely that eye movements drove the eccentricity-based correlation pattern as similar results were observed at rest during both fixation and eyes closed conditions. Finally, significant inter-run eccentricity-based correlation patterns were observed during movie viewing, but not during rest, further ruling out the possible influence of non-neuronal intrinsic artifacts (e.g., respiration rate, cardiac rate, motion), which should not be correlated across runs.</p><boxed-text id="box2"><object-id pub-id-type="doi">10.7554/eLife.03952.016</object-id><label>Box 2.</label><caption><title>Widespread connectivity is not dependent on instrumentation and analyses.</title><p>The eccentricity-based correlation could not be accounted for by overlapping receptive fields or anatomical distance. There is an intrinsic spatio-temporal point-spread function in BOLD imaging (<xref ref-type="bibr" rid="bib29">Engel et al., 1997</xref>; <xref ref-type="bibr" rid="bib69">Parkes et al., 2005</xref>; <xref ref-type="bibr" rid="bib77">Shmuel et al., 2007</xref>). The point-spread function of BOLD imaging at 3T had been estimated to be about 3.5 mm (<xref ref-type="bibr" rid="bib29">Engel et al., 1997</xref>). We tested whether the observed correlation patterns were due to inherent spatial autocorrelation of the BOLD signal or pre-processing steps in our analysis pipeline by generating artificial data that maintained these non-neuronal spatial correlation patterns. We replaced the timeseries in each voxel with a randomly generated timeseries to preserve the anatomical position of each voxel in each subject and then applied a Gaussian spatial filter with a FWHM of 3.5 mm. We passed these data through the same processing as the real data (using motion correction parameters from each subject's real data) to derive estimated ‘instrumental’ correlations for each subject. These data capture the inherent spatial blur of the BOLD signal and preserve (any) volumetric and surface-based spatial correlations that were inherent in the data or introduced via the preprocessing and binning analyses. We found significant effects of radial distance for intra-areal correlations (e.g., V2d RH with itself; <italic>p</italic>s < 0.05) and weaker, though non-significant, effects for adjacent areas (e.g., V1d RH with V2d RH) (<xref ref-type="fig" rid="fig6s3">Figure 6—figure supplement 3</xref>). As discussed in the exploratory seed analysis section above, this was to be expected since the organization of eccentricity representations in these areas is correlated with (anatomical) spatial distance, and the spatial blurring in this analysis will introduce an anatomically local correlation structure to the data. Interestingly, the slopes of coefficients across eccentricity disparities did not reflect the slopes observed in our real data, suggesting that this cannot account for the entire signal measured between these areas in our real data (<xref ref-type="fig" rid="fig6s3">Figure 6—figure supplement 3</xref>). For intra-area correlations, coefficient values sharply decreased as a function of radial distance, and resembled the exponential decay reported by Butt and colleagues in their simulations of instrumental correlations for visual area V1 (<xref ref-type="bibr" rid="bib18">Butt et al., 2013</xref>). For correlations between adjacent areas, coefficient values decreased gradually for the three smallest distance values then leveled off at larger disparities. Again, this small decrease likely reflects anatomical distance, and the profile of distance correlations differed from the correlation pattern measured in our real data. Critically, there was no significant difference in correlation strength across distance values between dorsal and ventral regions (e.g., V2d RH to V3v RH; <italic>p</italic>s > 0.05) or across hemispheres (e.g., V2d RH to V2d LH; <italic>p</italic>s > 0.05) (<xref ref-type="fig" rid="fig6s3">Figure 6—figure supplement 3</xref>). We also scanned a phantom (using the same acquisition parameters as the resting data) to generate artificial data with the BOLD spatial autocorrelation, and found the same effects as with the randomly generated timeseries data. This control analysis demonstrates that the inherent (spatial) spread of BOLD imaging and anatomical distance cannot account for the observed correlations between areas with non-overlapping receptive fields (e.g., V2d RH to V2v RH or V2d RH to V2d LH), which we have focused on thus far (<xref ref-type="fig" rid="fig4 fig6 fig7">Figures 4, 6, 7</xref>). If anything, this control analysis suggests that it is important to account for BOLD spatial autocorrelation when evaluating correlation patterns within individual areas (e.g., V1d RH to V1d RH) and between (anatomically) adjacent areas (e.g., V1d RH to V2d RH).</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952">http://dx.doi.org/10.7554/eLife.03952.016</ext-link></p></caption></boxed-text></sec><sec id="s2-7"><title>Topographic model regression analyses</title><p>Both local and widespread connectivity patterns contributed to the spatial pattern of observed correlations. Each visual hemifield map was separated into 36 bins: six divisions of eccentricity (E1–E6) that each contained six divisions of polar angle (A1–A6) such that each bin represented a unique part of visual space (<xref ref-type="fig" rid="fig8s1">Figure 8—figure supplement 1</xref>). To visualize the pattern of correlations across all bins with respect to local (overlapping RF) and widespread (eccentricity and polar angle) connectivity patterns, data were plotted as a function of radial and angular distance (<xref ref-type="fig" rid="fig8">Figure 8</xref>). Correlations were qualitatively strongest at the intersection of iso-polar angle and iso-eccentricity, and were generally elevated along iso-eccentricity representations, irrespective of polar angle distance (<xref ref-type="fig" rid="fig8">Figure 8</xref>). To quantify this, we assessed the relationship between the measured correlations across visual areas V1, V2, V3, and hV4 and several possible connectivity patterns using linear, least squares regression (see ‘Materials and methods’). For any pair of visual areas, we modeled correlations between all bin pairs (36 × 36) as the linear weighted sum of four sources of connectivity. Two sources reflect topographically local connectivity between regions that are in close anatomical proximity or contain overlapping receptive fields: (1) instrumental ‘noise’ connectivity: correlation pattern based on an assessment of the spatial auto-correlation and preprocessing in our data; (2) overlapping RF connectivity: correlation pattern based on the overlap of estimated population receptive fields (pRF; <xref ref-type="bibr" rid="bib28">Dumoulin and Wandell, 2008</xref>; <xref ref-type="bibr" rid="bib4">Amano et al., 2009</xref>; <xref ref-type="bibr" rid="bib44">Harvey and Dumoulin, 2011</xref>). The two other sources reflect topographically widespread connectivity that span much of the visual field and are not specifically tied to receptive fields or anatomical proximity: (1) eccentricity connectivity: correlation pattern based on radial distance with correlations strongest at iso-eccentricity representations; (2) polar angle connectivity: correlation pattern based on angular distance with correlations strongest at iso-polar angle representations.<fig-group><fig id="fig8" position="float"><object-id pub-id-type="doi">10.7554/eLife.03952.017</object-id><label>Figure 8.</label><caption><title>Radial and angular distance plots.</title><p>Resting state correlation bins plotted as a function of average radial and angular distance (ΔE and ΔA, respectively) for areas V1, V2, V3, and hV4. Intra- and inter- hemisphere correlations are plotted in the left and right hemifields, respectively. For intra-hemisphere correlations (left hemifield), the mid arc (oval; see legend) corresponds to the average correlation at the intersection of iso-eccentricity and iso-polar angle. There is no iso-polar angle for inter-hemisphere comparisons (right hemifield), so plotted data begin off the horizontal meridian at an iso-polar distance of one, and correspond to the correlations between bins adjacent to the vertical meridian. Outer and inner arcs correspond to larger radial distance correlations (same conventions as symmetry plots in <xref ref-type="fig" rid="fig6">Figure 6A</xref>). Arcs closer to the vertical meridian correspond to larger angular distances (see legend). For each plot, for example, V2–V3, arcs above the mid arc (+ΔA) illustrate correlations between relatively more upper visual field representations of V2 and relatively lower visual field representations of V3 (reverse for below mid arc, −ΔA). For comparisons with area hV4, there was no data for some radial and angular distances (black arcs).</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.017">http://dx.doi.org/10.7554/eLife.03952.017</ext-link></p></caption><graphic xlink:href="elife03952f008"/></fig><fig id="fig8s1" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.03952.018</object-id><label>Figure 8—figure supplement 1.</label><caption><title>Connectivity patterns for model fitting.</title><p>(<bold>A</bold>) For areas V1, V2, V3, and hV4, data were divided into six bins of eccentricity (E1–E6), each containing six bins of polar angle (A1–A6). (<bold>B</bold>) Illustration of the conversion from visual field representation to matrix representation for all pairwise (36 × 36) bin correlations. (<bold>C</bold>) Illustrations of four predicted patterns of intra-hemisphere connectivity between areas V2 and V3 (from left to right): (1) Noise, (2) RF overlap, (3) Eccentricity, and (4) Polar angle.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.018">http://dx.doi.org/10.7554/eLife.03952.018</ext-link></p></caption><graphic xlink:href="elife03952fs006"/></fig><fig id="fig8s2" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.03952.019</object-id><label>Figure 8—figure supplement 2.</label><caption><title>Homotopic radial and angular distance plots.</title><p>Correlation bins are plotted as a function of average radial and angular distance for V1, V2, V3, and hV4. Intra- and inter- hemisphere correlations are plotted in the left and right hemifields, respectively. For inter-hemisphere comparisons, the visual field position of bins in one hemifield is mirror reflected across the vertical meridian (homotopic angular distances denoted as A′ and eccentricity distances as E′, see top-right legend). In each hemifield, the mid arc (oval; see legend) corresponds to the average correlation at the intersection of iso-eccentricity and iso-polar angle. All other conventions the same as <xref ref-type="fig" rid="fig8">Figure 8</xref>.</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.019">http://dx.doi.org/10.7554/eLife.03952.019</ext-link></p></caption><graphic xlink:href="elife03952fs007"/></fig></fig-group></p><p>In general, the strongest overlapping RF effects were observed for intra-areal comparisons whereas eccentricity effects were consistently observed across all comparisons and were much larger than overlapping RF and polar angle effects for inter-hemisphere comparisons. To quantify the apparent connectivity patterns, we conducted an initial regression analysis using a model that included all four sources of connectivity as predictors. For individual subjects, the combination of local and widespread predictors well fit intra-area correlation patterns, and moderately fit inter-area and inter-hemisphere patterns. The average individual subject variance explained by the full model ranged between 40% and 68% for intra-areal comparisons and between 11% and 28% for inter-areal comparisons. Across area pairs, there were consistent effects of both local and widespread predictors (<xref ref-type="fig" rid="fig9s1">Figure 9—figure supplement 1</xref>). Across subjects, the coefficients for the overlapping RF predictor were significantly positive for all intra-hemisphere, intra-areal comparisons as well as for most other comparisons (<italic>p</italic>s < 0.05, FDR-corrected). Coefficients for the eccentricity predictor were significantly positive for all but one comparison (V1–hV4 inter-hemisphere) (<italic>p</italic>s < 0.05, FDR corrected). Coefficients for the polar angle predictor were significantly positive for most intra-hemisphere comparisons, but only for a few inter-hemisphere comparisons (<italic>p</italic>s < 0.05, FDR corrected). Though these results suggest some effect of both local and widespread connectivity, we found that both eccentricity and polar angle predictors were strongly correlated with the overlapping RF predictor for most intra-hemisphere area pairs (mean r = 0.54; STD = 0.08). Since the local and widespread predictors have a non-trivial degree of collinearity, the coefficient estimates for individual predictors from the full model may not be accurate for intra-hemisphere comparisons, which had the strongest correlations between local and widespread predictors. From these results, it is possible that some of the observed eccentricity and polar angle effects are due to shared variance with the local, overlapping RF model. However, it is very unlikely that all of the widespread effects can generically be explained by the overlapping receptive field model since significant correlations were observed in the eccentricity bin analyses between iso-eccentricity representations in distinct parts of visual space (<xref ref-type="fig" rid="fig6">Figure 6</xref>; e.g., between V2d RH and V2v LH).</p><p>When removing the shared variance between local and widespread connectivity predictors, significant effects of eccentricity connectivity were still observed. To assess the contribution of widespread connectivity on the measured correlation patterns while controlling for the shared variance with the local connectivity model, we first removed the variance explained by local connectivity (i.e., overlapping RF and instrumental predictors), then calculated correlations between the residuals and the widespread connectivity models in each subject. The average correlation coefficient between the residual pattern and the eccentricity predictor was similar across most areas for both within and across hemispheres (<xref ref-type="fig" rid="fig9">Figure 9</xref>, medium gray bars). The Fisher-transformed individual subject eccentricity residual correlations were reliably different from zero for 56/60 visual area pairs across rest and movie viewing experiments (<italic>p</italic>s < 0.05, FDR corrected). Consistent with the binning analyses, residual correlations that included hV4 were weaker than residual correlations between areas V1, V2, and V3. In contrast, there was no clear, consistent effect of the polar angle predictor across areas (<xref ref-type="fig" rid="fig9">Figure 9</xref>, lightest gray bars). The polar angle residual correlations were only significantly positive for 4/60 pairs (<italic>p</italic>s < 0.05, FDR corrected). For all other comparisons, residual polar angle correlations tended to be negative. The residual pattern was more correlated with the eccentricity predictor than with the polar angle predictor for 49/60 pairs (<italic>p</italic>s < 0.05, FDR corrected) (<xref ref-type="fig" rid="fig9">Figure 9</xref>).<fig-group><fig id="fig9" position="float"><object-id pub-id-type="doi">10.7554/eLife.03952.020</object-id><label>Figure 9.</label><caption><title>Residual correlations with eccentricity and polar angle predictors.</title><p>(<bold>A</bold>) Intra- and (<bold>B</bold>) inter- hemisphere average individual subject correlations between the unexplained variance from an overlapping RF model fit and eccentricity (medium gray bars) and polar angle (light gray bars) predictors are plotted for all pairs of visual areas V1, V2, V3, and hV4. Residual correlations were significantly above 0 for 56/60 eccentricity comparisons (<italic>p</italic>s < 0.05; FDR corrected; one-sample t-test) and were significantly greater than polar angle correlations for 49/60 comparisons (<italic>p</italic>s < 0.05, FDR corrected; paired t-test). Correlations between the residuals and the eccentricity predictor were comparable within and across areas, as well as within and across hemispheres, though were generally weaker for comparisons with hV4. Notations above each bar denote significance relative to the null hypothesis (one-sample t-test) and brackets denote significant differences between conditions (two-sample t-test). * <italic>p</italic>s < 0.05; ∼<italic>p</italic>s < 0.10 (FDR corrected).</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.020">http://dx.doi.org/10.7554/eLife.03952.020</ext-link></p></caption><graphic xlink:href="elife03952f009"/></fig><fig id="fig9s1" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.03952.021</object-id><label>Figure 9—figure supplement 1.</label><caption><title>Intra- and inter- hemisphere effects of local and widespread connectivity.</title><p>(<bold>A</bold>) Intra- and (<bold>B</bold>) inter- hemisphere group mean coefficients plotted from the linear regression with a model that included two predictors of local connectivity (Noise and overlapping RF) and two predictors of widespread connectivity (Eccentricity and polar angle). Significant effects of overlapping RF (dark gray) and polar angle (white) were generally observed for intra-hemisphere comparisons. Significant effects of eccentricity (light gray) were generally observed for both intra- and inter- hemisphere comparisons. *<italic>p</italic>s < 0.05 (FDR corrected).</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.021">http://dx.doi.org/10.7554/eLife.03952.021</ext-link></p></caption><graphic xlink:href="elife03952fs008"/></fig><fig id="fig9s2" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.03952.022</object-id><label>Figure 9—figure supplement 2.</label><caption><title>Residual correlations with eccentricity and polar angle predictors after accounting for potential effects of BOLD signal spread.</title><p>(<bold>A</bold>) Intra- and (<bold>B</bold>) inter- hemisphere group mean correlations between the unexplained variance after simulating the spread of the BOLD signal across RFs (via Cholesky analysis) and eccentricity (medium gray bars) and polar angle (light gray bars) predictors are plotted for all pairs of visual areas V1, V2, V3, and hV4. The Cholesky analysis accounts for any additional effects of the point spread function on overlapping RF connectivity pattern. Residual correlations were significantly above 0 for 58/60 eccentricity comparisons (<italic>p</italic>s < 0.05; FDR corrected; one-sample t-test) and were significantly greater than polar angle correlations for 46/60 comparisons (<italic>p</italic>s < 0.05; FDR corrected; paired t-test). Overall, the results were little changed from the initial residual correlations. Notation conventions the same as <xref ref-type="fig" rid="fig9">Figure 9</xref>. *<italic>p</italic>s < 0.05; ∼<italic>p</italic>s < 0.10 (FDR corrected).</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.022">http://dx.doi.org/10.7554/eLife.03952.022</ext-link></p></caption><graphic xlink:href="elife03952fs009"/></fig><fig id="fig9s3" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.03952.023</object-id><label>Figure 9—figure supplement 3.</label><caption><title>Effect of homotopic connectivity.</title><p>(<bold>A</bold>) Coefficients are plotted for inter-hemisphere comparisons from a linear regression with a model that included homotopic RF connectivity, eccentricity, and polar angle predictors. In contrast to the overlapping RF predictor, significant effects of homotopic RF connectivity were observed for some inter-hemisphere comparisons. Conventions the same as <xref ref-type="fig" rid="fig9s1">Figure 9—figure supplement 1</xref>. (<bold>B</bold>) Inter- hemisphere group mean correlations between the unexplained variance from a homotopic RF connectivity model fit and eccentricity (medium gray bars) and polar angle (lightest gray bars) predictors are plotted for all area pairs of visual areas V1, V2, V3, and hV4. Residual correlations were significantly above 0 for 26/30 eccentricity comparisons (<italic>p</italic>s < 0.05; FDR corrected; one-sample t-test) and were significantly greater than polar angle correlations for 28/30 comparisons (<italic>p</italic>s < 0.05; FDR corrected; paired t-test). Overall, the results were similar to the initial residual correlations. Notation conventions the same as <xref ref-type="fig" rid="fig9">Figure 9</xref>. *<italic>p</italic>s < 0.05; ∼<italic>p</italic>s < 0.10 (FDR corrected).</p><p><bold>DOI:</bold> <ext-link ext-link-type="doi" xlink:href="10.7554/eLife.03952.023">http://dx.doi.org/10.7554/eLife.03952.023</ext-link></p></caption><graphic xlink:href="elife03952fs010"/></fig></fig-group></p><p>These residual widespread effects were robust to scaling of pRF sizes in the local connectivity model. We used the average pRF values reported across several studies to derive our model of overlapping RF connectivity, however the size of pRFs for some areas varied as much as 2× across these studies. To make sure that the variance in pRF estimates across studies didn't significantly affect the results, we re-ran the model analyses using pRF estimates from individual studies to generate the model of overlapping RFs (see ‘Materials and methods’). Using these parameters to derive the overlapping RF model had no appreciable effect on the magnitude or significance of residual correlations. Further, the magnitude and significance of the eccentricity-based effects were also largely unchanged when we scaled pRF estimates to account for any additional effects of the point spread function in the BOLD signal not captured in the original pRF estimate (see ‘Materials and methods’; <xref ref-type="fig" rid="fig9s2">Figure 9—figure supplement 2</xref>). Overall, these control analyses provide additional evidence for a widespread eccentricity-based connectivity that cannot be accounted for by local connectivity patterns.</p></sec><sec id="s2-8"><title>Eccentricity effect does not reflect asymmetry in local connectivity patterns</title><p>Widespread eccentricity-based connectivity effects were consistently observed before and after removal of local connectivity variance. Even when attributing all shared variance between local and widespread predictors to the local predictor, eccentricity-based effects were still observed. In contrast, consistent polar angle effects were generally only observed prior to removal of local connectivity variance. Eccentricity effects were generally stronger than polar angle both before and after removal of local connectivity variance, suggesting that this difference is driven by asymmetries in widespread, not local, connectivity patterns. In further support of this, the variance explained by the local connectivity model was strongly correlated with both eccentricity (mean r across areas = 0.50; STD = 0.08) and polar angle (mean r across areas = 0.49; STD = 0.07) predictors, and there was little difference between them (mean <italic>r</italic> difference across areas = 0.01; STD = 0.06). Taken together, these data suggest that connectivity effects based on overlapping RFs are comparable along eccentricity and polar angle axes (i.e., isotropic with respect to visual field), which is consistent with anatomical studies in macaques that suggest visual field coverage of local, lateral connections is generally isotropic (<xref ref-type="bibr" rid="bib5">Angelucci et al., 2002</xref>). Importantly, the pattern of widespread correlated BOLD signal beyond overlapping receptive fields reflects eccentricity organization.</p><p>Our data suggests little effect of polar angle connectivity beyond overlapping RFs. As with any negative finding, we can only report that we failed to see an effect of angular connectivity between regions with non-overlapping RFs, but do not conclude with certainty that it does not exist. It is important to note that significant polar angle correlations were observed in the full model analysis for several area pairs. However, these correlations were almost entirely explained by the overlapping RF connectivity model. As discussed above, the variance explained by the local connectivity model was strongly correlated with both the eccentricity and polar angle models, and there was little difference between them. Second, the residual widespread effects were not biased by any imbalance in the number of data-points (nodes) across data bins. We conducted a control analysis where we equalized the number of nodes in each bin of the topographic model analysis. We used bins with 20, 30, and 50 surface nodes. For bins that contained more than the maximum number of nodes, we subsampled the data. In some subjects, a few bins contained less than the set number of nodes. We conducted the control analyses both including and excluding these bins with few nodes. Eccentricity and (lack of) polar angle residual correlations were observed in these analyses and were consistent with the original results shown in <xref ref-type="fig" rid="fig9">Figure 9</xref>, and the interpretation of the results did not change. Thus, our analyses were sensitive enough to reveal polar angle connectivity effects, and the lack of consistent residual polar angle correlations between regions with non-overlapping RFs cannot be explained by such biases in our data.</p></sec><sec id="s2-9"><title>Homotopic RF connectivity</title><p>One of the more robust effects consistently observed in resting state correlations is that of homotopic connectivity; the BOLD signals of homologous cortical regions between hemispheres are highly correlated (<xref ref-type="bibr" rid="bib13">Biswal et al., 1995</xref>). Recent studies have demonstrated homotopic connectivity in visual cortex with respect to visual field representation (<xref ref-type="bibr" rid="bib46">Heinzle et al., 2011</xref>; <xref ref-type="bibr" rid="bib18">Butt et al., 2013</xref>). Homotopic connectivity was evident in our data when plotting inter-hemisphere correlations with respect to the radial and angular distance after reflecting the position coordinates of one hemisphere across the vertical meridian onto the other hemifield (<xref ref-type="fig" rid="fig8s2">Figure 8—figure supplement 2</xref>). Indeed, for inter-hemisphere comparisons, the homotopic RF predictor generally captured more variance than the overlapping RF predictor. We tested whether homotopic connectivity (with respect to the vertical meridian) could account for the inter-hemisphere eccentricity-based connectivity effects. We removed all variance explained by the homotopic RF predictor then recomputed the residual correlations. The eccentricity effects were still present in these residual correlations (<xref ref-type="fig" rid="fig9s3">Figure 9—figure supplement 3</xref>), though intra-areal hV4 correlations were notably weaker. This was to be expected as the eccentricity and overlapping RF predictors were highly correlated for intra-areal hV4, and removal of the local connectivity variance also removed a large portion of the variance attributable to the eccentricity predictor. Overall, these data show that the eccentricity-based connectivity effects were not driven by homotopic RF connectivity.</p></sec><sec id="s2-10"><title>Comparison of local and widespread connectivity influences</title><p>The variance explained by the overlapping RF model (after accounting for patterns attributable to the ‘noise’ model) was generally greater than the variance explained by the model of residual widespread connectivity (after removal of overlapping RF connectivity effects) for within hemisphere comparisons (mean ratio across areas, conditions = 1.69:1.00). For between hemisphere comparisons, variance explained by the overlapping RF model (after accounting for patterns attributable to the ‘noise’ model) was weaker than the variance explained by the model of residual widespread connectivity without accounting for mirror symmetrical connections (mean ratio across areas, condition = 1.00:5.65), but was generally greater when accounting for these connections (mean ratio across areas, condition = 1.27:1.00).</p><p>In summary, these data demonstrate that both topographically local and widespread connectivity patterns significantly contributed to the observed correlation patterns across V1, V2, V3 and hV4. Critically, the eccentricity correlation patterns explained an additional, significant amount of variance in our data, which was not accounted for by the local connectivity model. For most inter-hemispheric comparisons, adding the eccentricity predictor more than doubled the variance explained by the local connectivity model, suggesting that it was the major component of the correlation patterns between hemispheres. Notably, after accounting for overlapping pRF connectivity, the strength of eccentricity correlations were similar within and across hemispheres for most comparisons between V1, V2, and V3, suggesting that this widespread connectivity pattern uniformly spans the whole visual field.</p></sec></sec><sec sec-type="discussion" id="s3"><title>Discussion</title><p>We investigated the spatial pattern of functional correlations across eight human early visual and extrastriate cortical areas using fMRI during conditions in which there was little or no external input (resting-state) and during conditions in which there was a dynamic external input (movie viewing). Local and widespread (spatial) patterns of correlated BOLD signal were observed in all experiments, both in the presence and absence of visual input. In agreement with prior reports, we observed topographically-local correlation patterns based on overlapping representations of visual space. In addition, we found strong evidence for topographically-widespread correlation patterns based on eccentricity within and across hemispheres, which spanned the entire visual field. The eccentricity-based correlation patterns were strongest between early visual areas (V1–V3). The effects observed in extrastriate areas (hV4, V3A–B, and VO1–2) were generally weaker than early visual areas, consistent with the coarser topographic organization of visual space and smaller surface area in these areas.</p><p>Our data extend recent findings of topographic connectivity within visual cortex using fMRI. At a fine-scale, correlation patterns reflect overlapping RFs within hemispheres and homotopic connections between hemispheres (<xref ref-type="bibr" rid="bib46">Heinzle et al., 2011</xref>; <xref ref-type="bibr" rid="bib42">Haak et al., 2012</xref>; <xref ref-type="bibr" rid="bib18">Butt et al., 2013</xref>). By modeling V1 inter-areal connections based on overlapping RFs, Gravel and colleagues (2014) generated polar angle and eccentricity maps for V2 and V3. However, Raemaekers and colleagues (2014) reported that such fine-scaled connectivity was only observable after filtering coarse-scale components. At a coarse scale, there has been evidence for a general foveal-peripheral distinction (<xref ref-type="bibr" rid="bib87">Vincent et al., 2007</xref>; <xref ref-type="bibr" rid="bib63">Nallasamy and Tsao, 2011</xref>; <xref ref-type="bibr" rid="bib79">Smith et al., 2012</xref>; <xref ref-type="bibr" rid="bib70">Raemaekers et al., 2014</xref>), but these studies did not report a systematic <italic>topography</italic> of correlation patterns that reflects the underlying eccentricity organization, nor were effects of overlapping RFs and cortical distance controlled. Evidence for a more systematic relation to eccentricity was reported between V1 and ventral V3 (<xref ref-type="bibr" rid="bib94">Yeo et al., 2011</xref>), though such data are also consistent with overlapping RF connectivity as well as cortical distance since only ventral V3 was probed. Here, we explicitly and quantitatively combined these connectivity phenomena in a fine-grained analysis across a wider range of brain regions, replicating previous findings of overlapping RF and homotopic correlation patterns without filtering coarse spatial components (<xref ref-type="bibr" rid="bib46">Heinzle et al., 2011</xref>; <xref ref-type="bibr" rid="bib42">Haak et al., 2012</xref>; <xref ref-type="bibr" rid="bib18">Butt et al., 2013</xref>; <xref ref-type="bibr" rid="bib41">Gravel et al., 2014</xref>). Even after accounting for these patterns, we found a robust widespread correlation pattern that reflects the eccentricity organization across much of visual cortex. Such systematic correlation patterns suggest orderly integration processes across the whole visual field at multiple levels of the processing hierarchy, not just a foveal-peripheral dichotomy.</p><p>Eccentricity-based correlation patterns may reflect an intrinsic functional organization of visual cortex. Our results during rest demonstrate that regions with iso-eccentricity representations are likely to be co-active, even in the absence of visual input. Our results during the movie viewing condition demonstrate that the temporal dynamics yielding these eccentricity-based correlation patterns are also present during strong bottom-up input, and indicate that this organization is relevant for the processing of incoming visual input. In agreement with these findings, studies in macaques and cats have shown that the activity of neurons with similar response properties are correlated in the presence and absence of external input, suggesting that spontaneous neuronal activity is tightly linked to intrinsic cortical networks (<xref ref-type="bibr" rid="bib9">Arieli et al., 1995</xref>; <xref ref-type="bibr" rid="bib83">Tsodyks et al., 1999</xref>; <xref ref-type="bibr" rid="bib40">Goldberg et al., 2004</xref>; <xref ref-type="bibr" rid="bib38">Ghuman et al., 2013</xref>). Further, both overlapping RF and eccentricity-based connectivity patterns were observed in the presence and absence of external input, suggesting that topographically-local and widespread patterns are both part of this intrinsic functional architecture.</p><p>The observed patterns of functional connectivity may reflect both direct and indirect anatomical connectivity (<xref ref-type="bibr" rid="bib87">Vincent et al., 2007</xref>; <xref ref-type="bibr" rid="bib48">Honey et al., 2009</xref>). Local connectivity is likely supported by direct anatomical connections between overlapping RFs (<xref ref-type="bibr" rid="bib22">Cragg, 1969</xref>; <xref ref-type="bibr" rid="bib30">Essen and Zeki, 1978</xref>; <xref ref-type="bibr" rid="bib60">Maunsell and Van Essen, 1983</xref>). Such wiring is necessary for the integration of information within focal points of our visual environment. Direct intra-areal anatomical connections between dorsal and ventral visual cortex at the horizontal meridians (<xref ref-type="bibr" rid="bib52">Jeffs et al., 2009</xref>) and between both hemispheres at the vertical meridians (<xref ref-type="bibr" rid="bib50">Hubel and Wiesel, 1967</xref>; <xref ref-type="bibr" rid="bib30">Essen and Zeki, 1978</xref>; <xref ref-type="bibr" rid="bib64">Newsome and Allman, 1980</xref>; <xref ref-type="bibr" rid="bib23">Cusick et al., 1984</xref>; <xref ref-type="bibr" rid="bib55">Kennedy et al., 1986</xref>) could support widespread functional connectivity across the visual field, though we are not aware of anatomical studies explicitly reporting eccentricity-based patterns of intra-areal connectivity. Further, while labeled cells of lateral connections in macaque striate and extrastriate cortex exhibit some anisotropy with respect to the cortical surface, this is thought to reflect cortical magnification factor, and yield isotropic visual field coverage (<xref ref-type="bibr" rid="bib5">Angelucci et al., 2002</xref>). Consistent with the anatomical connectivity, we found that correlation patterns between regions with overlapping RFs were comparable along eccentricity and polar angle dimensions. Beyond overlapping RFs, correlation patterns were <italic>anisotropic</italic> (with respect to visual field coverage) and reflected the underlying eccentricity organization. Alternatively, the observed eccentricity-based correlation patterns may actually reflect a broader-scale anatomical organization of direct (and indirect) connections, facilitated via differences in intra-areal projections between cortical sites representing central and peripheral space (<xref ref-type="bibr" rid="bib20">Colby et al., 1988</xref>; <xref ref-type="bibr" rid="bib62">Nakamura et al., 1993</xref>; <xref ref-type="bibr" rid="bib36">Gattass et al., 2005</xref>; <xref ref-type="bibr" rid="bib84">Ungerleider et al., 2008</xref>, <xref ref-type="bibr" rid="bib85">2014</xref>). Such a distinction has been characterized in the patterns of supra-areal anatomical connections between early visual and extrastriate cortex in non-human primates (<xref ref-type="bibr" rid="bib71">Rosa, 2002</xref>; <xref ref-type="bibr" rid="bib36">Gattass et al., 2005</xref>; <xref ref-type="bibr" rid="bib73">Rosa and Tweedale, 2005</xref>; <xref ref-type="bibr" rid="bib72">Rosa et al., 2009</xref>; <xref ref-type="bibr" rid="bib16">Buckner and Yeo, 2014</xref>). It is not known whether these anatomical connectivity patterns are ‘bi-modal’, and only distinguish central and peripheral space, or reflect a finer-scale organization where connectivity patterns with intermediate eccentricity representations are distinguishable from central and peripheral connectivity profiles. Our results predict that these anatomical connections should reflect a gradient, though this remains to be explored. In particular, feedback projections from extrastriate areas with receptive fields covering wide swaths of the visual field to early and intermediate visual areas could facilitate such widespread, eccentricity-dependent correlation patterns. It is interesting to note that when comparing the profile of anatomical connectivity between V2/V4 and higher order cortex (e.g., <xref ref-type="fig" rid="fig7">Figure 7</xref>, <xref ref-type="bibr" rid="bib36">Gattass et al., 2005</xref>) with the organization of eccentricity across visual cortex in macaques (<xref ref-type="bibr" rid="bib15">Brewer et al., 2002</xref>; <xref ref-type="bibr" rid="bib56">Kolster et al., 2009</xref>; <xref ref-type="bibr" rid="bib7">Arcaro et al., 2011</xref>), it is clear that higher order areas connected with peripheral parts of V2 and V4 (e.g., PO, PIP, LIP, DP, TF) have a large representation of the periphery, and higher order areas connected with foveal parts of V2 and V4 (e.g., TEO and TE) have a large representation of the fovea. The exact relationship between the observed correlation patterns and anatomical pathways will need to be further investigated.</p><p>Our data link the functional organization of early and higher order visual cortex. Previous studies have proposed eccentricity as a large-scale functional organizing principle for higher order visual cortex (<xref ref-type="bibr" rid="bib58">Levy et al., 2001</xref>; <xref ref-type="bibr" rid="bib45">Hasson et al., 2002</xref>; <xref ref-type="bibr" rid="bib59">Malach et al., 2002</xref>). Higher order areas with foveal biases tend to be specialized in face and object recognition, and areas with peripheral biases tend to be involved in scene analysis (<xref ref-type="bibr" rid="bib58">Levy et al., 2001</xref>; <xref ref-type="bibr" rid="bib45">Hasson et al., 2002</xref>; <xref ref-type="bibr" rid="bib59">Malach et al., 2002</xref>). Perceptually, these recognition processes require different visual acuities. For example, while fine acuity is needed for the featural discrimination among similar face and object exemplars (<xref ref-type="bibr" rid="bib32">Fiorentini et al., 1983</xref>; <xref ref-type="bibr" rid="bib39">Goffaux et al., 2005</xref>; <xref ref-type="bibr" rid="bib54">Keil, 2008</xref>), a coarser acuity is needed for mapping the surrounding layout necessary for navigation in space (<xref ref-type="bibr" rid="bib67">Oliva and Schyns, 1997</xref>; <xref ref-type="bibr" rid="bib68">Oliva and Torralba, 2006</xref>). Further, eye movement patterns during scene perception are related to the types of information within a scene (<xref ref-type="bibr" rid="bib17">Buswell, 1935</xref>; <xref ref-type="bibr" rid="bib47">Henderson and Hollingworth, 1999</xref>). People tend to foveate on faces while orienting their peripheral vision at landscape features and room contours (<xref ref-type="bibr" rid="bib93">Yarbus, 1967</xref>). These eccentricity biases are also reflected in the connectivity patterns between face and place category-selective regions in ventral temporal cortex with extrastriate visual area hV4 (<xref ref-type="bibr" rid="bib10">Baldassano et al., 2012</xref>). Our data show that this divergence in the computational processes necessary for foveal and peripheral recognition is evident even in early visual cortex.</p><p>Our results underscore the importance of relating functional connectivity data to known functional (or anatomical) organization (<xref ref-type="bibr" rid="bib51">Jbabdi et al., 2013</xref>; <xref ref-type="bibr" rid="bib80">Sporns and Honey, 2013</xref>; <xref ref-type="bibr" rid="bib91">Wang et al., 2013</xref>). The detailed retinotopic organization of visual cortex allowed for a unique opportunity to systematically compare patterns of correlated BOLD activity with the known underlying functional organization of the visual system. Across subjects and experiments, correlations were stronger between areas at matched eccentricities than within areas at large eccentricity distances, suggesting that functional connectivity analyses on BOLD data are more sensitive at revealing widespread, inter-areal connectivity patterns than the localization of individual retinotopic areas (see also <xref ref-type="bibr" rid="bib94">Yeo et al., 2011</xref>). Alternative connectivity approaches not based on similarity may prove useful at revealing area boundaries (<xref ref-type="bibr" rid="bib92">Wig et al., 2014</xref>), though this remains to be tested more thoroughly beyond the V1/V2 border. Thus, we propose that relating correlation patterns to known functional and anatomical data will prove important for identifying the functional pathways for the integration of information across individual, functionally specialized areas.</p></sec><sec sec-type="materials|methods" id="s4"><title>Materials and methods</title><sec id="s4-1"><title>Participants</title><p>14 subjects (aged 24–34 years, six females) participated in the study, which was approved by the Institutional Review Board of Princeton University (Resting State & Retinotopy Experiments: IRB#4616, Movie Viewing Experiments: IRB#5516). All participants were in good health without history of psychiatric or neurological disorders and gave their informed written consent to participate in the study and consent to publish in accordance with ethical standards set out by the Federal Policy for the Protection of Human Subjects (or ‘Common Rule’, U.S. Department of Health and Human Services Title 45 DFR 46). Subjects had normal or corrected-to-normal visual acuity. All participants were experienced MRI subjects that were well trained to maintain central fixation for several minutes at a time while lying still during scans.</p></sec><sec id="s4-2"><title>General procedure</title><p>All subjects participated in three scanning sessions, during which resting state scans were collected, high-resolution structural images were acquired for cortical surface reconstructions, and polar angle and eccentricity measurements were obtained to delineate retinotopic areas. 11 of these subjects viewed movie clips in a single additional scanning session.</p></sec><sec id="s4-3"><title>Resting state</title><p>Each subject participated in two versions of resting state: (1) fixation and (2) eyes closed. During the fixation scans, subjects were instructed to maintain fixation on a centrally presented dot (0.3° diameter) overlaid on a mean grey luminance screen background for 10 min. During the eyes closed scans, the projector was turned off and subjects were instructed to keep their eyes closed for 10 min. Two runs were collected per resting condition.</p></sec><sec id="s4-4"><title>Movie condition</title><p>11 subjects viewed an audiovisual movie clip from the film <italic>Dog Day Afternoon</italic>. Subjects were instructed to attend to the movie, but maintain fixation on a centrally presented dot (0.3° diameter). Movie stimuli subtended 20° horizontally and 16° vertically. Two runs were collected per condition with each run lasting 5 min 45 s.</p></sec><sec id="s4-5"><title>Retinotopic mapping</title><p>Polar angle and eccentricity representations were measured using a standard traveling wave paradigm consisting of a colored checkerboard wedge or annulus, respectively (<xref ref-type="bibr" rid="bib82">Swisher et al., 2007</xref>; <xref ref-type="bibr" rid="bib6">Arcaro et al., 2009</xref>, <xref ref-type="bibr" rid="bib7">2011</xref>). For eccentricity mapping, the annulus increased on a logarithmic scale over time in size and rate of expansion to approximately match the human cortical magnification function in early visual cortex (<xref ref-type="bibr" rid="bib49">Horton and Hoyt, 1991</xref>; <xref ref-type="bibr" rid="bib82">Swisher et al., 2007</xref>). Using a logarithmic scale yields a roughly even distribution of eccentricity phases across the cortical surface for early visual areas V1 and V2 (<xref ref-type="bibr" rid="bib43">Hansen et al., 2007</xref>; <xref ref-type="bibr" rid="bib82">Swisher et al., 2007</xref>; <xref ref-type="bibr" rid="bib75">Schira et al., 2009</xref>). Stimuli mapped the central 15° of the visual field. Due to limitations of the scanner bore size and viewing angle, peripheral representations beyond 15° were not mapped nor included in any analyses. Each run consisted of eight 40 s cycles. For each subject, 4–5 polar angle runs and 2–3 eccentricity runs were collected. Early visual and extrastriate areas V1, V2, V3, hV4, V3A–B, VO1–2 were defined using standard criteria reported previously (<xref ref-type="bibr" rid="bib76">Sereno et al., 1995</xref>; <xref ref-type="bibr" rid="bib26">DeYoe et al., 1996</xref>; <xref ref-type="bibr" rid="bib29">Engel et al., 1997</xref>; <xref ref-type="bibr" rid="bib14">Brewer et al., 2005</xref>; <xref ref-type="bibr" rid="bib88">Wandell et al., 2007</xref>; <xref ref-type="bibr" rid="bib6">Arcaro et al., 2009</xref>). For more details, see <xref ref-type="bibr" rid="bib6 bib7">Arcaro et al. (2009, 2011)</xref>.</p></sec><sec id="s4-6"><title>Data acquisition and preprocessing</title><p>Data were acquired with a 3T Skyra magnetic resonance imaging (MRI) scanner (Siemens, Munich, Germany) using a 16-channel head coil. All functional acquisitions used a gradient echo, echo planar sequence with a 64 square matrix (slice thickness of 4 mm, interleaved acquisition) leading to an in-plane resolution of 3 × 3 mm<sup>2</sup> (field of view [FOV], 192 × 192 mm<sup>2</sup>; GRAPPA iPAT = 2; 32 slices per volume for resting state and 27 for movie stimuli; repetition time [TR] = 1.8 s for resting state and 1.5 s for movie scans; echo time [TE] = 30 ms; flip angle = 72°). High-resolution structural scans were acquired in each scan session for registration to surface anatomical images (MPRAGE sequence; 256 matrix; 240 × 240 mm<sup>2</sup> FOV; TR, 1.9 s; TE 2.1 ms; flip angle 9°, 0.9375 × 0.9375 × 0.9375 mm<sup>3</sup> resolution).</p></sec><sec id="s4-7"><title>Data preprocessing</title><p>Data were analyzed using AFNI (<xref ref-type="bibr" rid="bib21">Cox, 1996</xref>) (<ext-link ext-link-type="uri" xlink:href="http://afni.nimh.nih.gov/afni/">http://afni.nimh.nih.gov/afni/</ext-link>), SUMA (<ext-link ext-link-type="uri" xlink:href="http://afni.nimh.nih.gov/afni/suma/">http://afni.nimh.nih.gov/afni/suma/</ext-link>), MATLAB (The MathWorks Inc., Natick, MA), and FreeSurfer (<xref ref-type="bibr" rid="bib24">Dale et al., 1999</xref>; <xref ref-type="bibr" rid="bib33">Fischl et al., 1999a</xref>) (<ext-link ext-link-type="uri" xlink:href="http://surfer.nmr.mgh.harvard.edu/">http://surfer.nmr.mgh.harvard.edu/</ext-link>). Functional data were slice-time and motion corrected. Motion distance (estimated by AFNI's 3dvolreg) did not exceed 1.0 mm (relative to starting head position) in any of the six motion parameter estimates (three translation and three rotation) during any run for any subject. In preparation for correlation analyses, several additional steps were performed on the data: (1) removal of signal deviation >2.5 SDs from the mean (AFNI's 3dDespike); (2) temporal filtering retaining frequencies in the 0.01–0.1 Hz band; (3) linear and quadratic detrending; and (4) removal by linear regression of several sources of variance: (i) the six motion parameter estimates (three translation and three rotation) and their temporal derivatives, (ii) the signal from a ventricular region, and (iii) the signal from a white matter region. Removal of ventricular and white matter signal resulted in a general, broad decrease in the raw correlation values by about 0.2, though subsequent eccentricity-specific effects were slightly increased. These are standard preprocessing steps for resting-state correlation analyses (e.g., <xref ref-type="bibr" rid="bib87">Vincent et al., 2007</xref>; <xref ref-type="bibr" rid="bib94">Yeo et al., 2011</xref>), though our results were not dependent on these preprocessing steps, and correlation analyses on the raw data yielded qualitatively and statistically similar results. Global mean signal (GMS) removal was not included in the analysis reported here given concerns about negative correlations (<xref ref-type="bibr" rid="bib35">Fox et al., 2009</xref>; <xref ref-type="bibr" rid="bib61">Murphy et al., 2009</xref>; <xref ref-type="bibr" rid="bib74">Saad et al., 2012</xref>), though inclusion of GMS removal yielded statistically similar results. To minimize the effect of any evoked response due to the scanner onset, the initial 21.6 s and 19.5 s were removed from each rest and movie scan, respectively. All voxels that fell between the gray and white matter boundaries were mapped to surface model units (nodes). Only for figure illustrations from single seed correlation analyses, data were spatially filtered using a Gaussian filter to a maximum smoothness of 4 mm full-width at half-max (FWHM) (by estimating the FWHM before spatial filtering), ensuring uniformity across the surface and maintaining spatial specificity while increasing the signal-to-noise ratio (SNR) (<xref ref-type="bibr" rid="bib19">Chung et al., 2005</xref>). No such spatial filtering was applied on data used for eccentricity bin or topographic model regression analyses. The timeseries from all surface nodes spanning early visual and extrastriate areas V1, V2, V3, hV4, VO1–2 (combined), V3A–B (combined), were extracted into MATLAB for correlation analyses. Eccentricity measurements were coarse for the surrounding visual cortex, so no additional extrastriate visual areas (e.g., LO1/2, TO1/2, PHC1/2, IPS0-5) were included in the analyses.</p></sec><sec id="s4-8"><title>Group seed analysis</title><p>Each subject's reconstructed cortical surface was warped to the Buckner40 template in Freesurfer (<xref ref-type="bibr" rid="bib34">Fischl et al., 1999b</xref>) and then resampled in SUMA using an icosahedral shape to generate a standard mesh with a constant number of co-registered nodes (<xref ref-type="bibr" rid="bib8">Argall et al., 2006</xref>). Phase and correlation maps were converted from individual surface space to the standard-mesh surface to generate group average data. Co-registered correlation maps were averaged across subjects to derive group average maps for the four seed locations. To visualize correlations as a function of visual field representation, individual subject radial and angular position data were converted to a 30 × 30 Cartesian grid space. Correlations were grouped as a function of visual field position on the grid (rounded to the nearest whole number). Multiple correlation values for the same visual position in individual subjects were averaged, and then correlation maps were averaged across subjects.</p></sec><sec id="s4-9"><title>Eccentricity bin analysis</title><p>For each subject, nodes were grouped by visual area (V1, V2, V3, hV4, V3A, V3B, VO1, VO2) for right and left hemispheres separately. Visual areas V1, V2, and V3 were separated into dorsal and ventral parts. Given the smaller surface area relative to V1, V2, and V3, visual areas V3A and V3B (as well as VO1 and VO2) were grouped together to increase the total number of samples (nodes). Due to the cortical magnification factor of early visual cortex, there were a limited number of voxels representing the periphery beyond 12.50°. Therefore, we restricted our analyses to the central 12.50°, and we refer to representations >10° eccentricity as peripheral-most. To increase signal-to-noise and control the extent of spatial signal blur, nodes were subdivided into 12 bins spanning 0.50°–12.50° eccentricity for each visual area. Eccentricity values from a log-scaled stimulus were used for the current analyses because the cortical magnification factor in early visual cortex is accounted for, yielding an approximately uniform distribution of nodes (i.e., data points) across eccentricity bins. The boundaries between bins corresponded to: 0.50°, 0.84°, 1.24°, 1.71°, 2.27°, 2.93°, 3.71°, 4.63°, 5.73°, 7.02°, 8.55°, 10.36°, and 12.50°. For each visual area, the timeseries of all nodes within each eccentricity bin were averaged to derive a mean timeseries for each eccentricity bin. In each subject, Pearson correlation coefficients were calculated between the mean timeseries of all eccentricity bins within as well as between visual areas, both within and between hemispheres. For each pair of visual areas, matrices were created containing all possible correlations between eccentricity bins. For each subject, these correlation matrices were created for each run separately and then averaged. Group average correlation matrices were also calculated for each area pair in each task (resting-fixation, resting-eyes shut, and movie-fixation). The magnitude of coefficients varied considerably between area pairs (e.g., correlation coefficients between V1 and V2 were larger than between V1 and hV4 at matched eccentricity bins). In order to illustrate the consistency in the pattern of eccentricity bin correlations across matrices (<xref ref-type="fig" rid="fig2 fig3">Figures 2C, 3</xref> only), coefficients were z-score normalized for each area pair separately. This preserved the relative differences in correlations between eccentricity bins within each matrix, but removed large magnitude differences between matrices. Non-normalized correlation matrices were used for all subsequent analyses and statistics. To ensure that the log-scaled eccentricity stimulus was not confounding the results, analyses were also run using eccentricity values that were converted into visual degrees. Comparable eccentricity-based effects were observed in this control analysis, though larger eccentricities had relatively fewer nodes per bin, and correlations with these bins were therefore more variable.</p></sec><sec id="s4-10"><title>Radial distance</title><p>Next, a ranked radial distance was calculated for each bin pair such that a radial distance of 0 corresponded to bin pairs with the same eccentricity value (iso-eccentricity) and 11 corresponded to pairs containing the foveal and peripheral-most bins. A radial distance matrix was created containing the differences between all eccentricity bin pairs (<xref ref-type="fig" rid="fig2">Figure 2C</xref>, left). To assess the relationship between the eccentricity structure and correlation patterns within the visual system, individual subject correlation matrices were correlated with this radial distance matrix.</p><p>Correlation coefficients within each matrix were then grouped as a function of radial distance. This yielded several correlation estimates for each radial distance value. Grouped correlation coefficients were then averaged to yield a single, mean correlation coefficient for each radial distance from 0–11. Correlations were Fisher z-transformed for statistical tests. For each visual area pair, two-tailed t-tests were performed on the Fisher-transformed 0–distance correlations to assess whether the subject population reliably differed from 0. False Discovery Rate (FDR) corrections were applied for each condition separately (<xref ref-type="bibr" rid="bib12">Benjamini and Hochberg, 1995</xref>). A linear regression was performed across all distance correlations (0–11) for each visual area pair in each subject. For within area correlations (e.g., V1 to V1), 0 distance coefficients were excluded from the regression since the coefficients reflected correlations between identical timeseries (i.e., mean coefficients were always a value of 1). The slopes were used to evaluate the strength of eccentricity-based correlations between area pairs and across conditions. Correlation coefficients and slopes were then averaged across subjects to derive group mean distance correlations and group mean slopes for each visual area pair. Statistical significance of the group mean slopes was tested using a non-parametric permutation test in which the radial distance values were shuffled prior to grouping of individual correlations. For each iteration, the same label shuffling was applied to all subjects. Linear regression analyses were performed on each subject's permuted data and the mean slopes were calculated. This permutation was run 10,000 times. For all area pairs, the mean slopes (averaged across subjects) from the non-permuted data were larger than 97.5% of slopes from the mean permuted data (i.e., significant for a two-tailed test with α = 0.05).</p><p>As observed with the binning analysis, the overall magnitude of mean radial distance coefficients broadly varied across visual area pairs (e.g., coefficients between V1 and V2 were much larger than between V1 and hV4 at matched eccentricity differences) and across conditions. Such variability in correlation strength was orthogonal to the focus of the current study. To minimize this magnitude variability, but preserve the relation of correlations across radial distances for illustration purposes (<xref ref-type="fig" rid="fig6 fig7">Figures 6, 7</xref>), correlation coefficients were normalized to the mean correlation coefficient at 0–distance (i.e., iso-eccentricity) in individual subjects as follows:<disp-formula id="equ1"><mml:math id="m1"><mml:mrow><mml:msub><mml:mtext>d</mml:mtext><mml:mtext>X</mml:mtext></mml:msub><mml:mo>=</mml:mo><mml:mn>1</mml:mn><mml:mo>−</mml:mo><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:msub><mml:mtext>r</mml:mtext><mml:mrow><mml:mn>0</mml:mn><mml:mtext>distance</mml:mtext></mml:mrow></mml:msub><mml:mo>−</mml:mo><mml:msub><mml:mtext>r</mml:mtext><mml:mrow><mml:mtext>Xdistance</mml:mtext></mml:mrow></mml:msub></mml:mrow><mml:mo>)</mml:mo></mml:mrow><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>where r<sub>0</sub> is the average correlation value at iso-eccentricity and d<sub>X</sub> is the correlation normalized to the average correlation at iso-eccentricity. This yielded a scale where 1 equals the correlation value of 0–distance. Values smaller (or larger) than 1 indicate that the correlations decrease (or increase) with greater radial distance. Since this was a simple subtraction, the relative coefficient differences between radial distances were identical to the non-normalized coefficients (i.e., preserves the slope). Importantly, the non-normalized coefficient values at and near the 0–distance were always significantly positive. Slight negative coefficients (between 0.00 and −0.15) were only observed for a few visual area pairs at large radial distance (i.e., between foveal and peripheral-most).</p></sec><sec id="s4-11"><title>Topographic model regression</title><p>The relation of local and widespread connectivity models to the observed correlation patterns was quantified across visual areas V1, V2, V3, and hh using regression. V3A–B and VO1–2 were excluded from these analyses due to the lack of published data on their population receptive fields (pRFs; <xref ref-type="bibr" rid="bib28">Dumoulin and Wandell, 2008</xref>). To compare the widespread eccentricity connectivity with other models of connectivity, each hemifield map was separated into six divisions of eccentricity, each containing six divisions of polar angle, yielding a total of 36 bins for each hemifield representation. Two spatial patterns of topographically local connectivity and two patterns of topographically widespread connectivity were generated:<list list-type="simple"><list-item><p>A. Topographically local connectivity:</p><list list-type="order"><list-item><p>Instrumental ‘noise’ connectivity (NSE): subject-specific predicted ‘noise’ correlations that are assumed to be non-neuronal in nature, and could result from any biases introduced from data acquisition and analyses, as well as the intrinsic spatial signal spread in the BOLD imaging (<xref ref-type="bibr" rid="bib11">Bandettini, 2009</xref>). The point spread function at 3T has been estimated to be about 3.5 mm (<xref ref-type="bibr" rid="bib29">Engel et al., 1997</xref>). To simulate the spatial autocorrelation in BOLD imaging at 3T, the timeseries of each voxel was replaced with a randomly generated, un-correlated timeseries for each subject's data, and a Gaussian filter with a kernel of 3.5 mm was applied to the data. These artificial data were passed through the same processing steps as the real data to derive estimated ‘noise’ correlations for each subject.</p></list-item><list-item><p>Point-to-point RF connectivity (RF): predicted correlations based on overlap of receptive fields in visual space. Due to the log-scaling used for eccentricity mapping, phase measurements were converted to visual degrees for this analysis. The receptive field of each node was calculated using a two-dimensional circular Gaussian spread (<xref ref-type="bibr" rid="bib28">Dumoulin and Wandell, 2008</xref>):</p></list-item></list></list-item></list><disp-formula id="equ2"><mml:math id="m2"><mml:mrow><mml:mtext>g</mml:mtext><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mtext>x,y</mml:mtext></mml:mrow><mml:mo>)</mml:mo></mml:mrow><mml:mo>=</mml:mo><mml:mi>A</mml:mi><mml:mtext>exp</mml:mtext><mml:mo>−</mml:mo><mml:mrow><mml:mo>[</mml:mo><mml:mrow><mml:msup><mml:mrow><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mtext>x</mml:mtext><mml:mo>−</mml:mo><mml:msub><mml:mtext>x</mml:mtext><mml:mtext>0</mml:mtext></mml:msub></mml:mrow><mml:mo>)</mml:mo></mml:mrow></mml:mrow><mml:mn>2</mml:mn></mml:msup><mml:mo>+</mml:mo><mml:msup><mml:mrow><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mtext>y</mml:mtext><mml:mo>−</mml:mo><mml:msub><mml:mtext>y</mml:mtext><mml:mtext>0</mml:mtext></mml:msub></mml:mrow><mml:mo>)</mml:mo></mml:mrow></mml:mrow><mml:mn>2</mml:mn></mml:msup></mml:mrow><mml:mo>]</mml:mo></mml:mrow><mml:mo>/</mml:mo><mml:mn>2</mml:mn><mml:msup><mml:mi>σ</mml:mi><mml:mn>2</mml:mn></mml:msup><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>where (x<sub>0</sub>, y<sub>0</sub>) is the visual field representation of a given node (in Cartesian coordinates), <italic>A</italic> is normalization constant to ensure integration unity, and σ is the Gaussian spread inferred from previously published pRF measurements (<xref ref-type="bibr" rid="bib28">Dumoulin and Wandell, 2008</xref>; <xref ref-type="bibr" rid="bib4">Amano et al., 2009</xref>; <xref ref-type="bibr" rid="bib44">Harvey and Dumoulin, 2011</xref>; <xref ref-type="bibr" rid="bib46">Heinzle et al., 2011</xref>) such that the integral of g(x, y) is 1. Specifically, a linear relationship was estimated between pRFs and eccentricity from the minimum and maximum eccentricities reported for each area. For any given area, pRF sizes varied across studies. To best approximate the pRF sizes from these prior reports, the average slope and intercept across reports was used, though analyses using individual slopes and intercepts from each of the prior reports yielded qualitatively and statistically similar results. On average, the size of the pRFs for 0.5° and 12.5° eccentricities were calculated as 0.4° and 1.6° for V1, 0.48° and 2.3° for V2, 1.0° and 4.15° for V3. For hV4, the size of the pRFs for 0.5° and 12.5° eccentricities were calculated solely from <xref ref-type="bibr" rid="bib44">Harvey and Dumoulin (2011)</xref> as 1.2° and 5.8°. pRFs were constructed for each node. Individual node pRFs were binned and averaged to construct a response field for each bin. Signal spread between bins was then calculated as the amount of response field overlap between bins relative to the total response field area of the bin pair.<list list-type="simple"><list-item><p>B. Topographically widespread connectivity:</p><list list-type="order"><list-item><p>Eccentricity connectivity (Ecc): predicted correlations based on radial distance. Bin pairs were assigned a value between 0 and 1; correlations were then assumed to be linearly proportional to the difference in eccentricity representations with iso-eccentricity representations assigned a value of 1.</p></list-item><list-item><p>Polar angle connectivity (Pol): predicted correlations based on angular distance. Bin pairs were assigned a value between 0 and 1; correlations were then assumed to be linearly proportional to the difference in polar angle representations with iso-polar angle representations assigned a value of 1.</p></list-item></list></list-item></list></p><p>The contribution of these four spatial patterns on the measured correlations was assessed using linear least-squares regression:<disp-formula id="equ3"><mml:math id="m3"><mml:mrow><mml:mtext>C</mml:mtext><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mtext>x,y</mml:mtext></mml:mrow><mml:mo>)</mml:mo></mml:mrow><mml:mo>=</mml:mo><mml:mi>A</mml:mi><mml:mo>+</mml:mo><mml:msub><mml:mi>β</mml:mi><mml:mn>1</mml:mn></mml:msub><mml:mo>∗</mml:mo><mml:mtext>NSE</mml:mtext><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mtext>x,y</mml:mtext></mml:mrow><mml:mo>)</mml:mo></mml:mrow><mml:mo>+</mml:mo><mml:msub><mml:mi>β</mml:mi><mml:mn>2</mml:mn></mml:msub><mml:mo>∗</mml:mo><mml:mtext>RF</mml:mtext><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mtext>x,y</mml:mtext></mml:mrow><mml:mo>)</mml:mo></mml:mrow><mml:mo>+</mml:mo><mml:msub><mml:mi>β</mml:mi><mml:mn>3</mml:mn></mml:msub><mml:mo>∗</mml:mo><mml:mtext>Ecc</mml:mtext><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mtext>x,y</mml:mtext></mml:mrow><mml:mo>)</mml:mo></mml:mrow><mml:mo>+</mml:mo><mml:msub><mml:mi>β</mml:mi><mml:mn>4</mml:mn></mml:msub><mml:mo>∗</mml:mo><mml:mtext>Pol</mml:mtext><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mtext>x,y</mml:mtext></mml:mrow><mml:mo>)</mml:mo></mml:mrow><mml:mo>+</mml:mo><mml:mi>ε</mml:mi><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mi>x</mml:mi><mml:mo>,</mml:mo><mml:mi>y</mml:mi></mml:mrow><mml:mo>)</mml:mo></mml:mrow><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>such that for any two area pairs (x, y), the correlation pattern C(x, y) is the linear weighted sum of four modeled sources, NSE(x, y), RF(x, y), Ecc(x, y) & Pol(x, y), with separate parameter coefficients <italic>ß</italic><sub><italic>x</italic></sub>, a constant <italic>A</italic>, and some measured error <italic>ε</italic>. Intra- and inter-hemisphere patterns were assessed separately. The results did not statistically differ using iterative reweighted least squares regression (bi-square), and so we only report the ordinary least squares regression results.</p></sec><sec id="s4-12"><title>Residual correlation analysis</title><p>The contribution of widespread predictors (eccentricity and polar angle) on the measured correlation patterns was re-assessed after first removing any shared variance with the topographically local connectivity. In each subject, topographically local connectivity (overlapping RF and instrumental ‘noise’ correlation) was removed from the data (via linear least-squares regression), and then the unexplained variance in the data (residuals) was correlated with eccentricity and polar angle predicted patterns, separately. Two-tailed t-tests were performed on the Fisher-transformed residual correlations to assess whether the subject population reliably differed from 0. Paired t-tests were performed between eccentricity and polar angle residual correlations. False Discovery Rate (FDR) corrections were applied to each condition separately (<xref ref-type="bibr" rid="bib12">Benjamini and Hochberg, 1995</xref>).</p></sec><sec id="s4-13"><title>Controlling for effects of spatial autocorrelation</title><p>Previously published pRF measurements were used to create our local connectivity model. Though these pRF size estimates were likely influenced by the point spread function of BOLD imaging, any unaccounted effect of the point spread function could lead to an underestimation of pRF coverage and thus of overlapping RF connectivity. We tested whether the observed connectivity patterns could be explained by such an underestimation of overlapping pRF effects. For each area pair, artificial data were generated with the correlation structure of the local connectivity via Cholesky decomposition. Artificial data were simulated for each run of resting and movie scans in individual subjects. These artificial data were then spatially smoothed with a 3.5 mm kernel Gaussian filter to approximate the point-spread function (<xref ref-type="bibr" rid="bib29">Engel et al., 1997</xref>), and correlations were computed between all 36 bins to generate a new local connectivity model. Though the instrumental noise predictor (NSE) in the local connectivity model accounts for the same point spread function, this connectivity model extends the effects of increased spatial blur for inter-areal, overlapping RF connectivity.</p></sec><sec id="s4-14"><title>Inter-hemisphere homotopic RF connectivity</title><p>To create a model of homotopic RF connectivity (with respect to the vertical meridian), the sign of the x coordinate was flipped for the RFs in one hemisphere. Overlap was calculated in the same manner as for local overlapping RF connectivity. Topographically local connectivity and homotopic effects were removed from the data (via linear least-squares regression). The unexplained variance in the data (residuals) was then correlated with eccentricity and polar angle predictors, and statistical tests were performed on the Fisher-transformed residual correlation coefficients.</p></sec></sec></body><back><ack id="ack"><title>Acknowledgements</title><p>We thank A Schapiro and L Wang for helpful comments. Supported by NSF: BCS 1025149, NIH (NINDS): F32-NS063619, NIH (NIMH): RO1-MH64043, NIH (NEI): R21-EY021078, NIH (NEI) R01-EY017699, NIH (NIMH): RO1-MH094480.</p></ack><sec sec-type="additional-information"><title>Additional information</title><fn-group content-type="competing-interest"><title>Competing interests</title><fn fn-type="conflict" id="conf1"><p>The authors declare that no competing interests exist.</p></fn></fn-group><fn-group content-type="author-contribution"><title>Author contributions</title><fn fn-type="con" id="con1"><p>MJA, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article</p></fn><fn fn-type="con" id="con2"><p>CJH, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article</p></fn><fn fn-type="con" id="con3"><p>REBM, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article</p></fn><fn fn-type="con" id="con4"><p>SK, Conception and design, Analysis and interpretation of data, Drafting or revising the article</p></fn><fn fn-type="con" id="con5"><p>UH, Conception and design, Analysis and interpretation of data, Drafting or revising the article</p></fn></fn-group><fn-group content-type="ethics-information"><title>Ethics</title><fn fn-type="other"><p>Human subjects: This study was approved by the Institutional Review Board of Princeton University (Resting State & Retinotopy Experiments: IRB#4616, Movie Viewing Experiments: IRB#5516). All participants were in good health without history of psychiatric or neurological disorders and gave their informed written consent to participate in the study and consent to publish in accordance with ethical standards set out by the Federal Policy for the Protection of Human Subjects (or ‘Common Rule’, U.S. Department of Health and Human Services Title 45 DFR 46).</p></fn></fn-group></sec><ref-list><title>References</title><ref id="bib1"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Abdollahi</surname><given-names>RO</given-names></name><name><surname>Kolster</surname><given-names>H</given-names></name><name><surname>Glasser</surname><given-names>MF</given-names></name><name><surname>Robinson</surname><given-names>EC</given-names></name><name><surname>Coalson</surname><given-names>TS</given-names></name><name><surname>Dierker</surname><given-names>D</given-names></name><name><surname>Jenkinson</surname><given-names>M</given-names></name><name><surname>Van Essen</surname><given-names>DC</given-names></name><name><surname>Orban</surname><given-names>GA</given-names></name></person-group><year>2014</year><article-title>Correspondences between retinotopic areas and myelin maps in human visual cortex</article-title><source>Neuroimage</source><volume>99</volume><fpage>509</fpage><lpage>524</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2014.06.042</pub-id></element-citation></ref><ref id="bib2"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Abel</surname><given-names>PL</given-names></name><name><surname>O'brien</surname><given-names>BJ</given-names></name><name><surname>Olavarria</surname><given-names>JF</given-names></name></person-group><year>2000</year><article-title>Organization of callosal linkages in visual area V2 of macaque monkey</article-title><source>The Journal of Comparative Neurology</source><volume>428</volume><fpage>278</fpage><lpage>293</lpage><pub-id pub-id-type="doi">10.1002/1096-9861(20001211)428:2<278::AID-CNE7>3.0.CO;2-R</pub-id></element-citation></ref><ref id="bib3"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Allman</surname><given-names>JM</given-names></name><name><surname>Kaas</surname><given-names>JH</given-names></name></person-group><year>1971</year><article-title>Representation of the visual field in striate and adjoining cortex of the owl monkey (<italic>Aotus trivirgatus</italic>)</article-title><source>Brain Research</source><volume>35</volume><fpage>89</fpage><lpage>106</lpage><pub-id pub-id-type="doi">10.1016/0006-8993(71)90596-8</pub-id></element-citation></ref><ref id="bib4"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Amano</surname><given-names>K</given-names></name><name><surname>Wandell</surname><given-names>BA</given-names></name><name><surname>Dumoulin</surname><given-names>SO</given-names></name></person-group><year>2009</year><article-title>Visual field maps, population receptive field sizes, and visual field coverage in the human MT+ complex</article-title><source>Journal of Neurophysiology</source><volume>102</volume><fpage>2704</fpage><lpage>2718</lpage><pub-id pub-id-type="doi">10.1152/jn.00102.2009</pub-id></element-citation></ref><ref id="bib5"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Angelucci</surname><given-names>A</given-names></name><name><surname>Levitt</surname><given-names>JB</given-names></name><name><surname>Walton</surname><given-names>EJ</given-names></name><name><surname>Hupe</surname><given-names>JM</given-names></name><name><surname>Bullier</surname><given-names>J</given-names></name><name><surname>Lund</surname><given-names>JS</given-names></name></person-group><year>2002</year><article-title>Circuits for local and global signal integration in primary visual cortex</article-title><source>The Journal of Neuroscience</source><volume>22</volume><fpage>8633</fpage><lpage>8646</lpage></element-citation></ref><ref id="bib6"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Arcaro</surname><given-names>MJ</given-names></name><name><surname>Mcmains</surname><given-names>SA</given-names></name><name><surname>Singer</surname><given-names>BD</given-names></name><name><surname>Kastner</surname><given-names>S</given-names></name></person-group><year>2009</year><article-title>Retinotopic organization of human ventral visual cortex</article-title><source>The Journal of Neuroscience</source><volume>29</volume><fpage>10638</fpage><lpage>10652</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.2807-09.2009</pub-id></element-citation></ref><ref id="bib7"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Arcaro</surname><given-names>MJ</given-names></name><name><surname>Pinsk</surname><given-names>MA</given-names></name><name><surname>Li</surname><given-names>X</given-names></name><name><surname>Kastner</surname><given-names>S</given-names></name></person-group><year>2011</year><article-title>Visuotopic organization of macaque posterior parietal cortex: a functional magnetic resonance imaging study</article-title><source>The Journal of Neuroscience</source><volume>31</volume><fpage>2064</fpage><lpage>2078</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.3334-10.2011</pub-id></element-citation></ref><ref id="bib8"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Argall</surname><given-names>BD</given-names></name><name><surname>Saad</surname><given-names>ZS</given-names></name><name><surname>Beauchamp</surname><given-names>MS</given-names></name></person-group><year>2006</year><article-title>Simplified intersubject averaging on the cortical surface using SUMA</article-title><source>Human Brain Mapping</source><volume>27</volume><fpage>14</fpage><lpage>27</lpage><pub-id pub-id-type="doi">10.1002/hbm.20158</pub-id></element-citation></ref><ref id="bib9"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Arieli</surname><given-names>A</given-names></name><name><surname>Shoham</surname><given-names>D</given-names></name><name><surname>Hildesheim</surname><given-names>R</given-names></name><name><surname>Grinvald</surname><given-names>A</given-names></name></person-group><year>1995</year><article-title>Coherent spatiotemporal patterns of ongoing activity revealed by real-time optical imaging coupled with single-unit recording in the cat visual cortex</article-title><source>Journal of Neurophysiology</source><volume>73</volume><fpage>2072</fpage><lpage>2093</lpage></element-citation></ref><ref id="bib10"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Baldassano</surname><given-names>C</given-names></name><name><surname>Iordan</surname><given-names>MC</given-names></name><name><surname>Beck</surname><given-names>DM</given-names></name><name><surname>Fei-Fei</surname><given-names>L</given-names></name></person-group><year>2012</year><article-title>Voxel-level functional connectivity using spatial regularization</article-title><source>Neuroimage</source><volume>63</volume><fpage>1099</fpage><lpage>1106</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2012.07.046</pub-id></element-citation></ref><ref id="bib11"><element-citation publication-type="book"><person-group person-group-type="author"><name><surname>Bandettini</surname><given-names>P</given-names></name></person-group><year>2009</year><article-title>Functional mri limitations and aspirations</article-title><source>Neural correlates of thinking</source><publisher-loc>Berlin, Heidelberg</publisher-loc><publisher-name>Springer</publisher-name></element-citation></ref><ref id="bib12"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Benjamini</surname><given-names>Y</given-names></name><name><surname>Hochberg</surname><given-names>Y</given-names></name></person-group><year>1995</year><article-title>Controlling the false discovery rate: a practical and powerful approach to multiple testing</article-title><source>Journal of Royal Statistical Society, Series B</source><volume>57</volume><fpage>289</fpage><lpage>300</lpage></element-citation></ref><ref id="bib13"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Biswal</surname><given-names>B</given-names></name><name><surname>Yetkin</surname><given-names>FZ</given-names></name><name><surname>Haughton</surname><given-names>VM</given-names></name><name><surname>Hyde</surname><given-names>JS</given-names></name></person-group><year>1995</year><article-title>Functional connectivity in the motor cortex of resting human brain using echo-planar MRI</article-title><source>Magnetic Resonance in Medicine</source><volume>34</volume><fpage>537</fpage><lpage>541</lpage><pub-id pub-id-type="doi">10.1002/mrm.1910340409</pub-id></element-citation></ref><ref id="bib14"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Brewer</surname><given-names>AA</given-names></name><name><surname>Liu</surname><given-names>J</given-names></name><name><surname>Wade</surname><given-names>AR</given-names></name><name><surname>Wandell</surname><given-names>BA</given-names></name></person-group><year>2005</year><article-title>Visual field maps and stimulus selectivity in human ventral occipital cortex</article-title><source>Nature Neuroscience</source><volume>8</volume><fpage>1102</fpage><lpage>1109</lpage><pub-id pub-id-type="doi">10.1038/nn1507</pub-id></element-citation></ref><ref id="bib15"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Brewer</surname><given-names>AA</given-names></name><name><surname>Press</surname><given-names>WA</given-names></name><name><surname>Logothetis</surname><given-names>NK</given-names></name><name><surname>Wandell</surname><given-names>BA</given-names></name></person-group><year>2002</year><article-title>Visual areas in macaque cortex measured using functional magnetic resonance imaging</article-title><source>The Journal of Neuroscience</source><volume>22</volume><fpage>10416</fpage><lpage>10426</lpage></element-citation></ref><ref id="bib16"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Buckner</surname><given-names>RL</given-names></name><name><surname>Yeo</surname><given-names>BT</given-names></name></person-group><year>2014</year><article-title>Borders, map clusters, and supra-areal organization in visual cortex</article-title><source>Neuroimage</source><volume>93</volume><fpage>292</fpage><lpage>297</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2013.12.036</pub-id></element-citation></ref><ref id="bib17"><element-citation publication-type="book"><person-group person-group-type="author"><name><surname>Buswell</surname><given-names>GT</given-names></name></person-group><year>1935</year><source>How people look at pictures: a study of the psychology and perception in art</source><publisher-loc>Oxford, England</publisher-loc><publisher-name>Chicago Press</publisher-name></element-citation></ref><ref id="bib18"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Butt</surname><given-names>OH</given-names></name><name><surname>Benson</surname><given-names>NC</given-names></name><name><surname>Datta</surname><given-names>R</given-names></name><name><surname>Aguirre</surname><given-names>GK</given-names></name></person-group><year>2013</year><article-title>The fine-scale functional correlation of striate cortex in sighted and blind people</article-title><source>The Journal of Neuroscience</source><volume>33</volume><fpage>16209</fpage><lpage>16219</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.0363-13.2013</pub-id></element-citation></ref><ref id="bib19"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Chung</surname><given-names>MK</given-names></name><name><surname>Robbins</surname><given-names>SM</given-names></name><name><surname>Dalton</surname><given-names>KM</given-names></name><name><surname>Davidson</surname><given-names>RJ</given-names></name><name><surname>Alexander</surname><given-names>AL</given-names></name><name><surname>Evans</surname><given-names>AC</given-names></name></person-group><year>2005</year><article-title>Cortical thickness analysis in autism with heat kernel smoothing</article-title><source>Neuroimage</source><volume>25</volume><fpage>1256</fpage><lpage>1265</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2004.12.052</pub-id></element-citation></ref><ref id="bib20"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Colby</surname><given-names>CL</given-names></name><name><surname>Gattass</surname><given-names>R</given-names></name><name><surname>Olson</surname><given-names>CR</given-names></name><name><surname>Gross</surname><given-names>CG</given-names></name></person-group><year>1988</year><article-title>Topographical organization of cortical afferents to extrastriate visual area PO in the macaque: a dual tracer study</article-title><source>The Journal of Comparative Neurology</source><volume>269</volume><fpage>392</fpage><lpage>413</lpage><pub-id pub-id-type="doi">10.1002/cne.902690307</pub-id></element-citation></ref><ref id="bib21"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cox</surname><given-names>RW</given-names></name></person-group><year>1996</year><article-title>AFNI: software for analysis and visualization of functional magnetic resonance neuroimages</article-title><source>Computers and Biomedical Research</source><volume>29</volume><fpage>162</fpage><lpage>173</lpage><pub-id pub-id-type="doi">10.1006/cbmr.1996.0014</pub-id></element-citation></ref><ref id="bib22"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cragg</surname><given-names>BG</given-names></name></person-group><year>1969</year><article-title>The topography of the afferent projections in the circumstriate visual cortex of the monkey studied by the Nauta method</article-title><source>Vision Research</source><volume>9</volume><fpage>733</fpage><lpage>747</lpage><pub-id pub-id-type="doi">10.1016/0042-6989(69)90011-X</pub-id></element-citation></ref><ref id="bib23"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cusick</surname><given-names>CG</given-names></name><name><surname>Gould</surname><given-names>HJ</given-names><suffix>III</suffix></name><name><surname>Kaas</surname><given-names>JH</given-names></name></person-group><year>1984</year><article-title>Interhemispheric connections of visual cortex of owl monkeys (<italic>Aotus trivirgatus</italic>), marmosets (<italic>Callithrix jacchus</italic>), and galagos (<italic>Galago crassicaudatus</italic>)</article-title><source>The Journal of Comparative Neurology</source><volume>230</volume><fpage>311</fpage><lpage>336</lpage><pub-id pub-id-type="doi">10.1002/cne.902300302</pub-id></element-citation></ref><ref id="bib24"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Dale</surname><given-names>AM</given-names></name><name><surname>Fischl</surname><given-names>B</given-names></name><name><surname>Sereno</surname><given-names>MI</given-names></name></person-group><year>1999</year><article-title>Cortical surface-based analysis. I. Segmentation and surface reconstruction</article-title><source>Neuroimage</source><volume>9</volume><fpage>179</fpage><lpage>194</lpage><pub-id pub-id-type="doi">10.1006/nimg.1998.0395</pub-id></element-citation></ref><ref id="bib25"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Daniel</surname><given-names>PM</given-names></name><name><surname>Whitteridge</surname><given-names>D</given-names></name></person-group><year>1961</year><article-title>The representation of the visual field on the cerebral cortex in monkeys</article-title><source>The Journal of Physiology</source><volume>159</volume><fpage>203</fpage><lpage>221</lpage><pub-id pub-id-type="doi">10.1113/jphysiol.1961.sp006803</pub-id></element-citation></ref><ref id="bib26"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>DeYoe</surname><given-names>EA</given-names></name><name><surname>Carman</surname><given-names>GJ</given-names></name><name><surname>Bandettini</surname><given-names>P</given-names></name><name><surname>Glickman</surname><given-names>S</given-names></name><name><surname>Wieser</surname><given-names>J</given-names></name><name><surname>Cox</surname><given-names>R</given-names></name><name><surname>Miller</surname><given-names>D</given-names></name><name><surname>Neitz</surname><given-names>J</given-names></name></person-group><year>1996</year><article-title>Mapping striate and extrastriate visual areas in human cerebral cortex</article-title><source>Proceedings of the National Academy of Sciences of USA</source><volume>93</volume><fpage>2382</fpage><lpage>2386</lpage><pub-id pub-id-type="doi">10.1073/pnas.93.6.2382</pub-id></element-citation></ref><ref id="bib27"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Donner</surname><given-names>TH</given-names></name><name><surname>Sagi</surname><given-names>D</given-names></name><name><surname>Bonneh</surname><given-names>YS</given-names></name><name><surname>Heeger</surname><given-names>DJ</given-names></name></person-group><year>2013</year><article-title>Retinotopic patterns of correlated fluctuations in visual cortex reflect the dynamics of spontaneous perceptual suppression</article-title><source>The Journal of Neuroscience</source><volume>33</volume><fpage>2188</fpage><lpage>2198</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.3388-12.2013</pub-id></element-citation></ref><ref id="bib28"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Dumoulin</surname><given-names>SO</given-names></name><name><surname>Wandell</surname><given-names>BA</given-names></name></person-group><year>2008</year><article-title>Population receptive field estimates in human visual cortex</article-title><source>Neuroimage</source><volume>39</volume><fpage>647</fpage><lpage>660</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2007.09.034</pub-id></element-citation></ref><ref id="bib29"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Engel</surname><given-names>SA</given-names></name><name><surname>Glover</surname><given-names>GH</given-names></name><name><surname>Wandell</surname><given-names>BA</given-names></name></person-group><year>1997</year><article-title>Retinotopic organization in human visual cortex and the spatial precision of functional MRI</article-title><source>Cerebral Cortex</source><volume>7</volume><fpage>181</fpage><lpage>192</lpage><pub-id pub-id-type="doi">10.1093/cercor/7.2.181</pub-id></element-citation></ref><ref id="bib31"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Felleman</surname><given-names>DJ</given-names></name><name><surname>Burkhalter</surname><given-names>A</given-names></name><name><surname>Van Essen</surname><given-names>DC</given-names></name></person-group><year>1997</year><article-title>Cortical connections of areas V3 and VP of macaque monkey extrastriate visual cortex</article-title><source>The Journal of Comparative Neurology</source><volume>379</volume><fpage>21</fpage><lpage>47</lpage><pub-id pub-id-type="doi">10.1002/(SICI)1096-9861(19970303)379:13.0.CO;2-K</pub-id></element-citation></ref><ref id="bib32"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Fiorentini</surname><given-names>A</given-names></name><name><surname>Maffei</surname><given-names>L</given-names></name><name><surname>Sandini</surname><given-names>G</given-names></name></person-group><year>1983</year><article-title>The role of high spatial frequencies in face perception</article-title><source>Perception</source><volume>12</volume><fpage>195</fpage><lpage>201</lpage><pub-id pub-id-type="doi">10.1068/p120195</pub-id></element-citation></ref><ref id="bib33"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Fischl</surname><given-names>B</given-names></name><name><surname>Sereno</surname><given-names>MI</given-names></name><name><surname>Dale</surname><given-names>AM</given-names></name></person-group><year>1999a</year><article-title>Cortical surface-based analysis. II: Inflation, flattening, and a surface-based coordinate system</article-title><source>Neuroimage</source><volume>9</volume><fpage>195</fpage><lpage>207</lpage><pub-id pub-id-type="doi">10.1006/nimg.1998.0396</pub-id></element-citation></ref><ref id="bib34"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Fischl</surname><given-names>B</given-names></name><name><surname>Sereno</surname><given-names>MI</given-names></name><name><surname>Tootell</surname><given-names>RB</given-names></name><name><surname>Dale</surname><given-names>AM</given-names></name></person-group><year>1999b</year><article-title>High-resolution intersubject averaging and a coordinate system for the cortical surface</article-title><source>Human Brain Mapping</source><volume>8</volume><fpage>272</fpage><lpage>284</lpage><pub-id pub-id-type="doi">10.1002/(SICI)1097-0193(1999)8:43.0.CO;2-4</pub-id></element-citation></ref><ref id="bib35"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Fox</surname><given-names>MD</given-names></name><name><surname>Zhang</surname><given-names>D</given-names></name><name><surname>Snyder</surname><given-names>AZ</given-names></name><name><surname>Raichle</surname><given-names>ME</given-names></name></person-group><year>2009</year><article-title>The global signal and observed anticorrelated resting state brain networks</article-title><source>Journal of Neurophysiology</source><volume>101</volume><fpage>3270</fpage><lpage>3283</lpage><pub-id pub-id-type="doi">10.1152/jn.90777.2008</pub-id></element-citation></ref><ref id="bib36"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Gattass</surname><given-names>R</given-names></name><name><surname>Nascimento-Silva</surname><given-names>S</given-names></name><name><surname>Soares</surname><given-names>JG</given-names></name><name><surname>Lima</surname><given-names>B</given-names></name><name><surname>Jansen</surname><given-names>AK</given-names></name><name><surname>Diogo</surname><given-names>AC</given-names></name><name><surname>Farias</surname><given-names>MF</given-names></name><name><surname>Botelho</surname><given-names>MM</given-names></name><name><surname>Mariani</surname><given-names>OS</given-names></name><name><surname>Azzi</surname><given-names>J</given-names></name><name><surname>Fiorani</surname><given-names>M</given-names></name></person-group><year>2005</year><article-title>Cortical visual areas in monkeys: location, topography, connections, columns, plasticity and cortical dynamics</article-title><source>Philosophical Transactions of the Royal Society of London Series B, Biological Sciences</source><volume>360</volume><fpage>709</fpage><lpage>731</lpage><pub-id pub-id-type="doi">10.1098/rstb.2005.1629</pub-id></element-citation></ref><ref id="bib37"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Gattass</surname><given-names>R</given-names></name><name><surname>Sousa</surname><given-names>AP</given-names></name><name><surname>Mishkin</surname><given-names>M</given-names></name><name><surname>Ungerleider</surname><given-names>LG</given-names></name></person-group><year>1997</year><article-title>Cortical projections of area V2 in the macaque</article-title><source>Cerebral Cortex</source><volume>7</volume><fpage>110</fpage><lpage>129</lpage><pub-id pub-id-type="doi">10.1093/cercor/7.2.110</pub-id></element-citation></ref><ref id="bib38"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Ghuman</surname><given-names>AS</given-names></name><name><surname>van den Honert</surname><given-names>RN</given-names></name><name><surname>Martin</surname><given-names>A</given-names></name></person-group><year>2013</year><article-title>Interregional neural synchrony has similar dynamics during spontaneous and stimulus-driven states</article-title><source>Scientific Reports</source><volume>3</volume><fpage>1481</fpage><pub-id pub-id-type="doi">10.1038/srep01481</pub-id></element-citation></ref><ref id="bib39"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Goffaux</surname><given-names>V</given-names></name><name><surname>Hault</surname><given-names>B</given-names></name><name><surname>Michel</surname><given-names>C</given-names></name><name><surname>Vuong</surname><given-names>QC</given-names></name><name><surname>Rossion</surname><given-names>B</given-names></name></person-group><year>2005</year><article-title>The respective role of low and high spatial frequencies in supporting configural and featural processing of faces</article-title><source>Perception</source><volume>34</volume><fpage>77</fpage><lpage>86</lpage><pub-id pub-id-type="doi">10.1068/p5370</pub-id></element-citation></ref><ref id="bib40"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Goldberg</surname><given-names>JA</given-names></name><name><surname>Rokni</surname><given-names>U</given-names></name><name><surname>Sompolinsky</surname><given-names>H</given-names></name></person-group><year>2004</year><article-title>Patterns of ongoing activity and the functional architecture of the primary visual cortex</article-title><source>Neuron</source><volume>42</volume><fpage>489</fpage><lpage>500</lpage><pub-id pub-id-type="doi">10.1016/S0896-6273(04)00197-7</pub-id></element-citation></ref><ref id="bib41"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Gravel</surname><given-names>N</given-names></name><name><surname>Harvey</surname><given-names>B</given-names></name><name><surname>Nordhjem</surname><given-names>B</given-names></name><name><surname>Haak</surname><given-names>KV</given-names></name><name><surname>Dumoulin</surname><given-names>SO</given-names></name><name><surname>Renken</surname><given-names>R</given-names></name><name><surname>Curčić-Blake</surname><given-names>B</given-names></name><name><surname>Cornelissen</surname><given-names>FW</given-names></name></person-group><year>2014</year><article-title>Cortical connective field estimates from resting state fMRI activity</article-title><source>Frontiers in Neuroscience</source><volume>8</volume><fpage>339</fpage><pub-id pub-id-type="doi">10.3389/fnins.2014.00339</pub-id></element-citation></ref><ref id="bib42"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Haak</surname><given-names>KV</given-names></name><name><surname>Winawer</surname><given-names>J</given-names></name><name><surname>Harvey</surname><given-names>BM</given-names></name><name><surname>Renken</surname><given-names>R</given-names></name><name><surname>Dumoulin</surname><given-names>SO</given-names></name><name><surname>Wandell</surname><given-names>BA</given-names></name><name><surname>Cornelissen</surname><given-names>FW</given-names></name></person-group><year>2012</year><article-title>Connective field modeling</article-title><source>Neuroimage</source><volume>66</volume><fpage>376</fpage><lpage>384</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2012.10.037</pub-id></element-citation></ref><ref id="bib43"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Hansen</surname><given-names>KA</given-names></name><name><surname>Kay</surname><given-names>KN</given-names></name><name><surname>Gallant</surname><given-names>JL</given-names></name></person-group><year>2007</year><article-title>Topographic organization in and near human visual area V4</article-title><source>The Journal of Neuroscience</source><volume>27</volume><fpage>11896</fpage><lpage>11911</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.2991-07.2007</pub-id></element-citation></ref><ref id="bib44"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Harvey</surname><given-names>BM</given-names></name><name><surname>Dumoulin</surname><given-names>SO</given-names></name></person-group><year>2011</year><article-title>The relationship between cortical magnification factor and population receptive field size in human visual cortex: constancies in cortical architecture</article-title><source>The Journal of Neuroscience</source><volume>31</volume><fpage>13604</fpage><lpage>13612</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.2572-11.2011</pub-id></element-citation></ref><ref id="bib45"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Hasson</surname><given-names>U</given-names></name><name><surname>Levy</surname><given-names>I</given-names></name><name><surname>Behrmann</surname><given-names>M</given-names></name><name><surname>Hendler</surname><given-names>T</given-names></name><name><surname>Malach</surname><given-names>R</given-names></name></person-group><year>2002</year><article-title>Eccentricity bias as an organizing principle for human high-order object areas</article-title><source>Neuron</source><volume>34</volume><fpage>479</fpage><lpage>490</lpage><pub-id pub-id-type="doi">10.1016/S0896-6273(02)00662-1</pub-id></element-citation></ref><ref id="bib46"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Heinzle</surname><given-names>J</given-names></name><name><surname>Kahnt</surname><given-names>T</given-names></name><name><surname>Haynes</surname><given-names>JD</given-names></name></person-group><year>2011</year><article-title>Topographically specific functional connectivity between visual field maps in the human brain</article-title><source>Neuroimage</source><volume>56</volume><fpage>1426</fpage><lpage>1436</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2011.02.077</pub-id></element-citation></ref><ref id="bib47"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Henderson</surname><given-names>JM</given-names></name><name><surname>Hollingworth</surname><given-names>A</given-names></name></person-group><year>1999</year><article-title>High-level scene perception</article-title><source>Annual Review of Psychology</source><volume>50</volume><fpage>243</fpage><lpage>271</lpage><pub-id pub-id-type="doi">10.1146/annurev.psych.50.1.243</pub-id></element-citation></ref><ref id="bib48"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Honey</surname><given-names>CJ</given-names></name><name><surname>Sporns</surname><given-names>O</given-names></name><name><surname>Cammoun</surname><given-names>L</given-names></name><name><surname>Gigandet</surname><given-names>X</given-names></name><name><surname>Thiran</surname><given-names>JP</given-names></name><name><surname>Meuli</surname><given-names>R</given-names></name><name><surname>Hagmann</surname><given-names>P</given-names></name></person-group><year>2009</year><article-title>Predicting human resting-state functional connectivity from structural connectivity</article-title><source>Proceedings of the National Academy of Sciences of USA</source><volume>106</volume><fpage>2035</fpage><lpage>2040</lpage><pub-id pub-id-type="doi">10.1073/pnas.0811168106</pub-id></element-citation></ref><ref id="bib49"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Horton</surname><given-names>JC</given-names></name><name><surname>Hoyt</surname><given-names>WF</given-names></name></person-group><year>1991</year><article-title>The representation of the visual field in human striate cortex. A revision of the classic Holmes map</article-title><source>Archives of Ophthalmology</source><volume>109</volume><fpage>816</fpage><lpage>824</lpage><pub-id pub-id-type="doi">10.1001/archopht.1991.01080060080030</pub-id></element-citation></ref><ref id="bib50"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Hubel</surname><given-names>DH</given-names></name><name><surname>Wiesel</surname><given-names>TN</given-names></name></person-group><year>1967</year><article-title>Cortical and callosal connections concerned with the vertical meridian of visual fields in the cat</article-title><source>Journal of Neurophysiology</source><volume>30</volume><fpage>1561</fpage><lpage>1573</lpage></element-citation></ref><ref id="bib51"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Jbabdi</surname><given-names>S</given-names></name><name><surname>Sotiropoulos</surname><given-names>SN</given-names></name><name><surname>Behrens</surname><given-names>TE</given-names></name></person-group><year>2013</year><article-title>The topographic connectome</article-title><source>Current Opinion in Neurobiology</source><volume>23</volume><fpage>207</fpage><lpage>215</lpage><pub-id pub-id-type="doi">10.1016/j.conb.2012.12.004</pub-id></element-citation></ref><ref id="bib52"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Jeffs</surname><given-names>J</given-names></name><name><surname>Ichida</surname><given-names>JM</given-names></name><name><surname>Federer</surname><given-names>F</given-names></name><name><surname>Angelucci</surname><given-names>A</given-names></name></person-group><year>2009</year><article-title>Anatomical evidence for classical and extra-classical receptive field completion across the discontinuous horizontal meridian representation of primate area V2</article-title><source>Cerebral Cortex</source><volume>19</volume><fpage>963</fpage><lpage>981</lpage><pub-id pub-id-type="doi">10.1093/cercor/bhn142</pub-id></element-citation></ref><ref id="bib53"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Kaas</surname><given-names>JH</given-names></name></person-group><year>1997</year><article-title>Topographic maps are fundamental to sensory processing</article-title><source>Brain Research Bulletin</source><volume>44</volume><fpage>107</fpage><lpage>112</lpage><pub-id pub-id-type="doi">10.1016/S0361-9230(97)00094-4</pub-id></element-citation></ref><ref id="bib54"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Keil</surname><given-names>MS</given-names></name></person-group><year>2008</year><article-title>Does face image statistics predict a preferred spatial frequency for human face processing?</article-title><source>Proceedings Biological Sciences/ The Royal Society</source><volume>275</volume><fpage>2095</fpage><lpage>2100</lpage><pub-id pub-id-type="doi">10.1098/rspb.2008.0486</pub-id></element-citation></ref><ref id="bib55"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Kennedy</surname><given-names>H</given-names></name><name><surname>Dehay</surname><given-names>C</given-names></name><name><surname>Bullier</surname><given-names>J</given-names></name></person-group><year>1986</year><article-title>Organization of the callosal connections of visual areas V1 and V2 in the macaque monkey</article-title><source>The Journal of Comparative Neurology</source><volume>247</volume><fpage>398</fpage><lpage>415</lpage><pub-id pub-id-type="doi">10.1002/cne.902470309</pub-id></element-citation></ref><ref id="bib56"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Kolster</surname><given-names>H</given-names></name><name><surname>Mandeville</surname><given-names>JB</given-names></name><name><surname>Arsenault</surname><given-names>JT</given-names></name><name><surname>Ekstrom</surname><given-names>LB</given-names></name><name><surname>Wald</surname><given-names>LL</given-names></name><name><surname>Vanduffel</surname><given-names>W</given-names></name></person-group><year>2009</year><article-title>Visual field map clusters in macaque extrastriate visual cortex</article-title><source>The Journal of Neuroscience</source><volume>29</volume><fpage>7031</fpage><lpage>7039</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.0518-09.2009</pub-id></element-citation></ref><ref id="bib57"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Leopold</surname><given-names>DA</given-names></name><name><surname>Murayama</surname><given-names>Y</given-names></name><name><surname>Logothetis</surname><given-names>NK</given-names></name></person-group><year>2003</year><article-title>Very slow activity fluctuations in monkey visual cortex: implications for functional brain imaging</article-title><source>Cerebral Cortex</source><volume>13</volume><fpage>422</fpage><lpage>433</lpage><pub-id pub-id-type="doi">10.1093/cercor/13.4.422</pub-id></element-citation></ref><ref id="bib58"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Levy</surname><given-names>I</given-names></name><name><surname>Hasson</surname><given-names>U</given-names></name><name><surname>Avidan</surname><given-names>G</given-names></name><name><surname>Hendler</surname><given-names>T</given-names></name><name><surname>Malach</surname><given-names>R</given-names></name></person-group><year>2001</year><article-title>Center-periphery organization of human object areas</article-title><source>Nature Neuroscience</source><volume>4</volume><fpage>533</fpage><lpage>539</lpage></element-citation></ref><ref id="bib59"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Malach</surname><given-names>R</given-names></name><name><surname>Levy</surname><given-names>I</given-names></name><name><surname>Hasson</surname><given-names>U</given-names></name></person-group><year>2002</year><article-title>The topography of high-order human object areas</article-title><source>Trends in Cognitive Sciences</source><volume>6</volume><fpage>176</fpage><lpage>184</lpage><pub-id pub-id-type="doi">10.1016/S1364-6613(02)01870-3</pub-id></element-citation></ref><ref id="bib60"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Maunsell</surname><given-names>JH</given-names></name><name><surname>Van Essen</surname><given-names>DC</given-names></name></person-group><year>1983</year><article-title>The connections of the middle temporal visual area (MT) and their relationship to a cortical hierarchy in the macaque monkey</article-title><source>The Journal of Neuroscience</source><volume>3</volume><fpage>2563</fpage><lpage>2586</lpage></element-citation></ref><ref id="bib61"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Murphy</surname><given-names>K</given-names></name><name><surname>Birn</surname><given-names>RM</given-names></name><name><surname>Handwerker</surname><given-names>DA</given-names></name><name><surname>Jones</surname><given-names>TB</given-names></name><name><surname>Bandettini</surname><given-names>PA</given-names></name></person-group><year>2009</year><article-title>The impact of global signal regression on resting state correlations: are anti-correlated networks introduced?</article-title><source>Neuroimage</source><volume>44</volume><fpage>893</fpage><lpage>905</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2008.09.036</pub-id></element-citation></ref><ref id="bib62"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Nakamura</surname><given-names>H</given-names></name><name><surname>Gattass</surname><given-names>R</given-names></name><name><surname>Desimone</surname><given-names>R</given-names></name><name><surname>Ungerleider</surname><given-names>LG</given-names></name></person-group><year>1993</year><article-title>The modular organization of projections from areas V1 and V2 to areas V4 and TEO in macaques</article-title><source>The Journal of Neuroscience</source><volume>13</volume><fpage>3681</fpage><lpage>3691</lpage></element-citation></ref><ref id="bib63"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Nallasamy</surname><given-names>N</given-names></name><name><surname>Tsao</surname><given-names>DY</given-names></name></person-group><year>2011</year><article-title>Functional connectivity in the brain: effects of anesthesia</article-title><source>The Neuroscientist</source><volume>17</volume><fpage>94</fpage><lpage>106</lpage><pub-id pub-id-type="doi">10.1177/1073858410374126</pub-id></element-citation></ref><ref id="bib64"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Newsome</surname><given-names>WT</given-names></name><name><surname>Allman</surname><given-names>JM</given-names></name></person-group><year>1980</year><article-title>Interhemispheric connections of visual cortex in the owl monkey, Aotus trivirgatus, and the bushbaby, Galago senegalensis</article-title><source>The Journal of Comparative Neurology</source><volume>194</volume><fpage>209</fpage><lpage>233</lpage><pub-id pub-id-type="doi">10.1002/cne.901940111</pub-id></element-citation></ref><ref id="bib65"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Nir</surname><given-names>Y</given-names></name><name><surname>Hasson</surname><given-names>U</given-names></name><name><surname>Levy</surname><given-names>I</given-names></name><name><surname>Yeshurun</surname><given-names>Y</given-names></name><name><surname>Malach</surname><given-names>R</given-names></name></person-group><year>2006</year><article-title>Widespread functional connectivity and fMRI fluctuations in human visual cortex in the absence of visual stimulation</article-title><source>Neuroimage</source><volume>30</volume><fpage>1313</fpage><lpage>1324</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2005.11.018</pub-id></element-citation></ref><ref id="bib66"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Nir</surname><given-names>Y</given-names></name><name><surname>Mukamel</surname><given-names>R</given-names></name><name><surname>Dinstein</surname><given-names>I</given-names></name><name><surname>Privman</surname><given-names>E</given-names></name><name><surname>Harel</surname><given-names>M</given-names></name><name><surname>Fisch</surname><given-names>L</given-names></name><name><surname>Gelbard-Sagiv</surname><given-names>H</given-names></name><name><surname>Kipervasser</surname><given-names>S</given-names></name><name><surname>Andelman</surname><given-names>F</given-names></name><name><surname>Neufeld</surname><given-names>MY</given-names></name><name><surname>Kramer</surname><given-names>U</given-names></name><name><surname>Arieli</surname><given-names>A</given-names></name><name><surname>Fried</surname><given-names>I</given-names></name><name><surname>Malach</surname><given-names>R</given-names></name></person-group><year>2008</year><article-title>Interhemispheric correlations of slow spontaneous neuronal fluctuations revealed in human sensory cortex</article-title><source>Nature Neuroscience</source><volume>11</volume><fpage>1100</fpage><lpage>1108</lpage><pub-id pub-id-type="doi">10.1038/nn.2177</pub-id></element-citation></ref><ref id="bib67"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Oliva</surname><given-names>A</given-names></name><name><surname>Schyns</surname><given-names>PG</given-names></name></person-group><year>1997</year><article-title>Coarse blobs or fine edges? Evidence that information diagnosticity changes the perception of complex visual stimuli</article-title><source>Cognitive Psychology</source><volume>34</volume><fpage>72</fpage><lpage>107</lpage><pub-id pub-id-type="doi">10.1006/cogp.1997.0667</pub-id></element-citation></ref><ref id="bib68"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Oliva</surname><given-names>A</given-names></name><name><surname>Torralba</surname><given-names>A</given-names></name></person-group><year>2006</year><article-title>Building the gist of a scene: the role of global image features in recognition</article-title><source>Progress in Brain Research</source><volume>155</volume><fpage>23</fpage><lpage>36</lpage><pub-id pub-id-type="doi">10.1016/S0079-6123(06)55002-2</pub-id></element-citation></ref><ref id="bib69"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Parkes</surname><given-names>LM</given-names></name><name><surname>Schwarzbach</surname><given-names>JV</given-names></name><name><surname>Bouts</surname><given-names>AA</given-names></name><name><surname>Deckers</surname><given-names>RH</given-names></name><name><surname>Pullens</surname><given-names>P</given-names></name><name><surname>Kerskens</surname><given-names>CM</given-names></name><name><surname>Norris</surname><given-names>DG</given-names></name></person-group><year>2005</year><article-title>Quantifying the spatial resolution of the gradient echo and spin echo BOLD response at 3 Tesla</article-title><source>Magnetic Resonance in Medicine</source><volume>54</volume><fpage>1465</fpage><lpage>1472</lpage><pub-id pub-id-type="doi">10.1002/mrm.20712</pub-id></element-citation></ref><ref id="bib70"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Raemaekers</surname><given-names>M</given-names></name><name><surname>Schellekens</surname><given-names>W</given-names></name><name><surname>Van Wezel</surname><given-names>RJ</given-names></name><name><surname>Petridou</surname><given-names>N</given-names></name><name><surname>Kristo</surname><given-names>G</given-names></name><name><surname>Ramsey</surname><given-names>NF</given-names></name></person-group><year>2014</year><article-title>Patterns of resting state connectivity in human primary visual cortical areas: a 7T fMRI study</article-title><source>Neuroimage</source><volume>84</volume><fpage>911</fpage><lpage>921</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2013.09.060</pub-id></element-citation></ref><ref id="bib71"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Rosa</surname><given-names>MG</given-names></name></person-group><year>2002</year><article-title>Visual maps in the adult primate cerebral cortex: some implications for brain development and evolution</article-title><source>Brazilian Journal of Medical and Biological Research</source><volume>35</volume><fpage>1485</fpage><lpage>1498</lpage><pub-id pub-id-type="doi">10.1590/S0100-879X2002001200008</pub-id></element-citation></ref><ref id="bib72"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Rosa</surname><given-names>MG</given-names></name><name><surname>Palmer</surname><given-names>SM</given-names></name><name><surname>Gamberini</surname><given-names>M</given-names></name><name><surname>Burman</surname><given-names>KJ</given-names></name><name><surname>Yu</surname><given-names>HH</given-names></name><name><surname>Reser</surname><given-names>DH</given-names></name><name><surname>Bourne</surname><given-names>JA</given-names></name><name><surname>Tweedale</surname><given-names>R</given-names></name><name><surname>Galletti</surname><given-names>C</given-names></name></person-group><year>2009</year><article-title>Connections of the dorsomedial visual area: pathways for early integration of dorsal and ventral streams in extrastriate cortex</article-title><source>The Journal of Neuroscience</source><volume>29</volume><fpage>4548</fpage><lpage>4563</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.0529-09.2009</pub-id></element-citation></ref><ref id="bib73"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Rosa</surname><given-names>MG</given-names></name><name><surname>Tweedale</surname><given-names>R</given-names></name></person-group><year>2005</year><article-title>Brain maps, great and small: lessons from comparative studies of primate visual cortical organization</article-title><source>Philosophical Transactions of the Royal Society of London Series B, Biological Sciences</source><volume>360</volume><fpage>665</fpage><lpage>691</lpage><pub-id pub-id-type="doi">10.1098/rstb.2005.1626</pub-id></element-citation></ref><ref id="bib74"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Saad</surname><given-names>ZS</given-names></name><name><surname>Gotts</surname><given-names>SJ</given-names></name><name><surname>Murphy</surname><given-names>K</given-names></name><name><surname>Chen</surname><given-names>G</given-names></name><name><surname>Jo</surname><given-names>HJ</given-names></name><name><surname>Martin</surname><given-names>A</given-names></name><name><surname>Cox</surname><given-names>RW</given-names></name></person-group><year>2012</year><article-title>Trouble at rest: how correlation patterns and group differences become distorted after global signal regression</article-title><source>Brain Connectivity</source><volume>2</volume><fpage>25</fpage><lpage>32</lpage><pub-id pub-id-type="doi">10.1089/brain.2012.0080</pub-id></element-citation></ref><ref id="bib75"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Schira</surname><given-names>MM</given-names></name><name><surname>Tyler</surname><given-names>CW</given-names></name><name><surname>Breakspear</surname><given-names>M</given-names></name><name><surname>Spehar</surname><given-names>B</given-names></name></person-group><year>2009</year><article-title>The foveal confluence in human visual cortex</article-title><source>The Journal of Neuroscience</source><volume>29</volume><fpage>9050</fpage><lpage>9058</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.1760-09.2009</pub-id></element-citation></ref><ref id="bib76"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Sereno</surname><given-names>MI</given-names></name><name><surname>Dale</surname><given-names>AM</given-names></name><name><surname>Reppas</surname><given-names>JB</given-names></name><name><surname>Kwong</surname><given-names>KK</given-names></name><name><surname>Belliveau</surname><given-names>JW</given-names></name><name><surname>Brady</surname><given-names>TJ</given-names></name><name><surname>Rosen</surname><given-names>BR</given-names></name><name><surname>Tootell</surname><given-names>RB</given-names></name></person-group><year>1995</year><article-title>Borders of multiple visual areas in humans revealed by functional magnetic resonance imaging</article-title><source>Science</source><volume>268</volume><fpage>889</fpage><lpage>893</lpage><pub-id pub-id-type="doi">10.1126/science.7754376</pub-id></element-citation></ref><ref id="bib77"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Shmuel</surname><given-names>A</given-names></name><name><surname>Yacoub</surname><given-names>E</given-names></name><name><surname>Chaimow</surname><given-names>D</given-names></name><name><surname>Logothetis</surname><given-names>NK</given-names></name><name><surname>Ugurbil</surname><given-names>K</given-names></name></person-group><year>2007</year><article-title>Spatio-temporal point-spread function of fMRI signal in human gray matter at 7 Tesla</article-title><source>Neuroimage</source><volume>35</volume><fpage>539</fpage><lpage>552</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2006.12.030</pub-id></element-citation></ref><ref id="bib78"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Silver</surname><given-names>MA</given-names></name><name><surname>Kastner</surname><given-names>S</given-names></name></person-group><year>2009</year><article-title>Topographic maps in human frontal and parietal cortex</article-title><source>Trends in Cognitive Sciences</source><volume>13</volume><fpage>488</fpage><lpage>495</lpage><pub-id pub-id-type="doi">10.1016/j.tics.2009.08.005</pub-id></element-citation></ref><ref id="bib79"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Smith</surname><given-names>SM</given-names></name><name><surname>Miller</surname><given-names>KL</given-names></name><name><surname>Moeller</surname><given-names>S</given-names></name><name><surname>XU</surname><given-names>J</given-names></name><name><surname>Auerbach</surname><given-names>EJ</given-names></name><name><surname>Woolrich</surname><given-names>MW</given-names></name><name><surname>Beckmann</surname><given-names>CF</given-names></name><name><surname>Jenkinson</surname><given-names>M</given-names></name><name><surname>Andersson</surname><given-names>J</given-names></name><name><surname>Glasser</surname><given-names>MF</given-names></name><name><surname>Van Essen</surname><given-names>DC</given-names></name><name><surname>Feinberg</surname><given-names>DA</given-names></name><name><surname>Yacoub</surname><given-names>ES</given-names></name><name><surname>Ugurbil</surname><given-names>K</given-names></name></person-group><year>2012</year><article-title>Temporally-independent functional modes of spontaneous brain activity</article-title><source>Proceedings of the National Academy of Sciences of USA</source><volume>109</volume><fpage>3131</fpage><lpage>3136</lpage><pub-id pub-id-type="doi">10.1073/pnas.1121329109</pub-id></element-citation></ref><ref id="bib80"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Sporns</surname><given-names>O</given-names></name><name><surname>Honey</surname><given-names>CJ</given-names></name></person-group><year>2013</year><article-title>Topographic dynamics in the resting brain</article-title><source>Neuron</source><volume>78</volume><fpage>955</fpage><lpage>956</lpage><pub-id pub-id-type="doi">10.1016/j.neuron.2013.05.037</pub-id></element-citation></ref><ref id="bib81"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Stepniewska</surname><given-names>I</given-names></name><name><surname>Kaas</surname><given-names>JH</given-names></name></person-group><year>1996</year><article-title>Topographic patterns of V2 cortical connections in macaque monkeys</article-title><source>The Journal of Comparative Neurology</source><volume>371</volume><fpage>129</fpage><lpage>152</lpage><pub-id pub-id-type="doi">10.1002/(SICI)1096-9861(19960715)371:13.0.CO;2-5</pub-id></element-citation></ref><ref id="bib82"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Swisher</surname><given-names>JD</given-names></name><name><surname>Halko</surname><given-names>MA</given-names></name><name><surname>Merabet</surname><given-names>LB</given-names></name><name><surname>McMains</surname><given-names>SA</given-names></name><name><surname>Somers</surname><given-names>DC</given-names></name></person-group><year>2007</year><article-title>Visual topography of human intraparietal sulcus</article-title><source>The Journal of Neuroscience</source><volume>27</volume><fpage>5326</fpage><lpage>5337</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.0991-07.2007</pub-id></element-citation></ref><ref id="bib83"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Tsodyks</surname><given-names>M</given-names></name><name><surname>Kenet</surname><given-names>T</given-names></name><name><surname>Grinvald</surname><given-names>A</given-names></name><name><surname>Arieli</surname><given-names>A</given-names></name></person-group><year>1999</year><article-title>Linking spontaneous activity of single cortical neurons and the underlying functional architecture</article-title><source>Science</source><volume>286</volume><fpage>1943</fpage><lpage>1946</lpage><pub-id pub-id-type="doi">10.1126/science.286.5446.1943</pub-id></element-citation></ref><ref id="bib84"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Ungerleider</surname><given-names>LG</given-names></name><name><surname>Galkin</surname><given-names>TW</given-names></name><name><surname>Desimone</surname><given-names>R</given-names></name><name><surname>Gattass</surname><given-names>R</given-names></name></person-group><year>2008</year><article-title>Cortical connections of area V4 in the macaque</article-title><source>Cerebral Cortex</source><volume>18</volume><fpage>477</fpage><lpage>499</lpage><pub-id pub-id-type="doi">10.1093/cercor/bhm061</pub-id></element-citation></ref><ref id="bib85"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Ungerleider</surname><given-names>LG</given-names></name><name><surname>Galkin</surname><given-names>TW</given-names></name><name><surname>Desimone</surname><given-names>R</given-names></name><name><surname>Gattass</surname><given-names>R</given-names></name></person-group><year>2014</year><article-title>Subcortical projections of area v2 in the macaque</article-title><source>Journal of Cognitive Neuroscience</source><volume>26</volume><fpage>1220</fpage><lpage>1233</lpage><pub-id pub-id-type="doi">10.1162/jocn_a_00571</pub-id></element-citation></ref><ref id="bib30"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Van Essen</surname><given-names>DC</given-names></name><name><surname>Zeki</surname><given-names>SM</given-names></name></person-group><year>1978</year><article-title>The topographic organization of rhesus monkey prestriate cortex</article-title><source>The Journal of Physiology</source><volume>277</volume><fpage>193</fpage><lpage>226</lpage><pub-id pub-id-type="doi">10.1113/jphysiol.1978.sp012269</pub-id></element-citation></ref><ref id="bib86"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Van Essen</surname><given-names>DC</given-names></name><name><surname>Newsome</surname><given-names>WT</given-names></name><name><surname>Maunsell</surname><given-names>JH</given-names></name></person-group><year>1984</year><article-title>The visual field representation in striate cortex of the macaque monkey: asymmetries, anisotropies, and individual variability</article-title><source>Vision Research</source><volume>24</volume><fpage>429</fpage><lpage>448</lpage><pub-id pub-id-type="doi">10.1016/0042-6989(84)90041-5</pub-id></element-citation></ref><ref id="bib87"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Vincent</surname><given-names>JL</given-names></name><name><surname>Patel</surname><given-names>GH</given-names></name><name><surname>Fox</surname><given-names>MD</given-names></name><name><surname>Snyder</surname><given-names>AZ</given-names></name><name><surname>Baker</surname><given-names>JT</given-names></name><name><surname>Van Essen</surname><given-names>DC</given-names></name><name><surname>Zempel</surname><given-names>JM</given-names></name><name><surname>Snyder</surname><given-names>LH</given-names></name><name><surname>Corbetta</surname><given-names>M</given-names></name><name><surname>Raichle</surname><given-names>ME</given-names></name></person-group><year>2007</year><article-title>Intrinsic functional architecture in the anaesthetized monkey brain</article-title><source>Nature</source><volume>447</volume><fpage>83</fpage><lpage>86</lpage><pub-id pub-id-type="doi">10.1038/nature05758</pub-id></element-citation></ref><ref id="bib88"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Wandell</surname><given-names>BA</given-names></name><name><surname>Dumoulin</surname><given-names>SO</given-names></name><name><surname>Brewer</surname><given-names>AA</given-names></name></person-group><year>2007</year><article-title>Visual field maps in human cortex</article-title><source>Neuron</source><volume>56</volume><fpage>366</fpage><lpage>383</lpage><pub-id pub-id-type="doi">10.1016/j.neuron.2007.10.012</pub-id></element-citation></ref><ref id="bib89"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Wandell</surname><given-names>BA</given-names></name><name><surname>Winawer</surname><given-names>J</given-names></name></person-group><year>2011</year><article-title>Imaging retinotopic maps in the human brain</article-title><source>Vision Research</source><volume>51</volume><fpage>718</fpage><lpage>737</lpage><pub-id pub-id-type="doi">10.1016/j.visres.2010.08.004</pub-id></element-citation></ref><ref id="bib90"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Wang</surname><given-names>L</given-names></name><name><surname>Mruczek</surname><given-names>RE</given-names></name><name><surname>Arcaro</surname><given-names>MJ</given-names></name><name><surname>Kastner</surname><given-names>S</given-names></name></person-group><year>2014</year><article-title>Probabilistic maps of visual topography in human cortex</article-title><source>Cerebral Cortex</source></element-citation></ref><ref id="bib91"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Wang</surname><given-names>Z</given-names></name><name><surname>Chen</surname><given-names>LM</given-names></name><name><surname>Negyessy</surname><given-names>L</given-names></name><name><surname>Friedman</surname><given-names>RM</given-names></name><name><surname>Mishra</surname><given-names>A</given-names></name><name><surname>Gore</surname><given-names>JC</given-names></name><name><surname>Roe</surname><given-names>AW</given-names></name></person-group><year>2013</year><article-title>The relationship of anatomical and functional connectivity to resting-state connectivity in primate somatosensory cortex</article-title><source>Neuron</source><volume>78</volume><fpage>1116</fpage><lpage>1126</lpage><pub-id pub-id-type="doi">10.1016/j.neuron.2013.04.023</pub-id></element-citation></ref><ref id="bib92"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Wig</surname><given-names>GS</given-names></name><name><surname>Laumann</surname><given-names>TO</given-names></name><name><surname>Petersen</surname><given-names>SE</given-names></name></person-group><year>2014</year><article-title>An approach for parcellating human cortical areas using resting-state correlations</article-title><source>Neuroimage</source><volume>93</volume><fpage>276</fpage><lpage>291</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2013.07.035</pub-id></element-citation></ref><ref id="bib93"><element-citation publication-type="book"><person-group person-group-type="author"><name><surname>Yarbus</surname><given-names>AL</given-names></name></person-group><year>1967</year><source>Eye movements and vision</source><publisher-loc>New York</publisher-loc><publisher-name>Plenum</publisher-name></element-citation></ref><ref id="bib94"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Yeo</surname><given-names>BT</given-names></name><name><surname>Krienen</surname><given-names>FM</given-names></name><name><surname>Sepulcre</surname><given-names>J</given-names></name><name><surname>Sabuncu</surname><given-names>MR</given-names></name><name><surname>Lashkari</surname><given-names>D</given-names></name><name><surname>Hollinshead</surname><given-names>M</given-names></name><name><surname>Roffman</surname><given-names>JL</given-names></name><name><surname>Smoller</surname><given-names>JW</given-names></name><name><surname>Zollei</surname><given-names>L</given-names></name><name><surname>Polimeni</surname><given-names>JR</given-names></name><name><surname>Fischl</surname><given-names>B</given-names></name><name><surname>Liu</surname><given-names>H</given-names></name><name><surname>Buckner</surname><given-names>RL</given-names></name></person-group><year>2011</year><article-title>The organization of the human cerebral cortex estimated by intrinsic functional connectivity</article-title><source>Journal of Neurophysiology</source><volume>106</volume><fpage>1125</fpage><lpage>1165</lpage><pub-id pub-id-type="doi">10.1152/jn.00338.2011</pub-id></element-citation></ref><ref id="bib95"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Zeki</surname><given-names>SM</given-names></name></person-group><year>1969</year><article-title>The secondary visual areas of the monkey</article-title><source>Brain Research</source><volume>13</volume><fpage>197</fpage><lpage>226</lpage><pub-id pub-id-type="doi">10.1016/0006-8993(69)90282-0</pub-id></element-citation></ref><ref id="bib96"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Zeki</surname><given-names>SM</given-names></name></person-group><year>1971</year><article-title>Cortical projections from two prestriate areas in the monkey</article-title><source>Brain Research</source><volume>34</volume><fpage>19</fpage><lpage>35</lpage><pub-id pub-id-type="doi">10.1016/0006-8993(71)90348-9</pub-id></element-citation></ref></ref-list></back><sub-article article-type="article-commentary" id="SA1"><front-stub><article-id pub-id-type="doi">10.7554/eLife.03952.024</article-id><title-group><article-title>Decision letter</article-title></title-group><contrib-group content-type="section"><contrib contrib-type="editor"><name><surname>Behrens</surname><given-names>Timothy</given-names></name><role>Reviewing editor</role><aff><institution>Oxford University</institution>, <country>United Kingdom</country></aff></contrib></contrib-group></front-stub><body><boxed-text><p>eLife posts the editorial decision letter and author response on a selection of the published articles (subject to the approval of the authors). An edited version of the letter sent to the authors after peer review is shown, indicating the substantive concerns or comments; minor concerns are not usually shown. Reviewers have the opportunity to discuss the decision before the letter is sent (see <ext-link ext-link-type="uri" xlink:href="http://elifesciences.org/review-process">review process</ext-link>). Similarly, the author response typically shows only responses to the major concerns raised by the reviewers.</p></boxed-text><p>Thank you for sending your work entitled “Widespread correlation patterns of fMRI signal across visual cortex reflect eccentricity organization” for consideration at <italic>eLife</italic>. Your article has been favorably evaluated by Eve Marder (Senior editor), Timothy Behrens (Reviewing editor), and 3 reviewers.</p><p>The Reviewing editor and the reviewers discussed their comments before we reached this decision, and the Reviewing editor has assembled the following comments to help you prepare a revised submission.</p><p>You can see the comments below. During our discussion, the central concerns with your manuscript are the overlap with Raemaekers et al. Neuroimage 2013 and disagreement between the two studies of the extent of the bias towards eccentricity rather than homotopic or polar connectivity.</p><p>It will be very important in your revisions to include a thorough discussion of these points. It would also be useful if there are further analyses you can try that may address these issues directly. It is of particular interest whether your data quality are adequate to disambiguate the different possibilities (see last comment R3). It was clear from the discussion among the reviewers that it is essential to address this point if <italic>eLife</italic> are to consider publishing a revision.</p><p><italic>Reviewer #1</italic>:</p><p>This is an impressively careful and thorough study of an important set of phenomena related to resting-state functional connectivity in human visual cortex. Several previous studies cited by the authors have identified resting-state correlations that include retinotopic components and other components that are spatially more widespread. The current study systematically examines the spatial pattern of correlations using two resting-state paradigms (with eyes closed and eyes open), plus a paradigm involving watching a movie while fixating. The most significant finding is that for all 3 paradigms the spatially extended correlations are stronger along iso-eccentricity bands. There is also a weaker component related to polar angle and another component related to homotopic (mirror-image retinotopy) in the opposite hemisphere.</p><p>In the Discussion, the authors attempt to provide a neurobiologically plausible interpretation of the strong eccentricity-related functional correlations analyzed in this study. In this reviewer's opinion, the results remain intriguing but still quite puzzling. However, the careful analysis done in this study represents a major advance, and it will likely stimulate further explorations aimed at better understanding the functional significance of these phenomena.</p><p>The manuscript is well reasoned and well written, and needs only minor revisions in the opinion of this reviewer.</p><p>Minor comments:</p><p>1) Both the Abstract and Introduction refer to evidence for more than two dozen retinotopic areas in human visual cortex. However, neither of the cited articles (<xref ref-type="bibr" rid="bib78">Silver and Kastner, 2009</xref>; <xref ref-type="bibr" rid="bib89">Wandell and Winawer, 2011</xref>) actually document evidence for two dozen retinotopic areas. Either the statement should be toned down (it is not critical to the study) or appropriate references should be added.</p><p>2) Results section, Materials and Methods section and <xref ref-type="fig" rid="fig2">Figure 2</xref>.</p><p>Please state the extent of visual stimuli for the retinotopic mapping (was it 30 degrees along the meridia, as in Arcaro et al., 2011?). Also, add a scale bar for the eccentricity maps in <xref ref-type="fig" rid="fig2">Figure 2B</xref>. Finally, consider clarifying that 'peripheral' really means 'near periphery', since nearly half of retinotopic cortex representing the far periphery (> 11 deg) was not mapped.</p><p>3) Results section, paragraph 12, and <xref ref-type="fig" rid="fig7">Figure 7</xref> legend. The legend states that “Average individual subject inter-run correlations between hemispheres......steadily decreased at larger distances for all conditions”, which doesn't match what the figure actually shows. The text has it right: “Indeed, inter-run Fisher-transformed correlations on resting state data showed no effect of radial distance (all ps > 0.05), validating the assumption that noise correlations should not be reliable across runs (<xref ref-type="fig" rid="fig7">Figure 7</xref>, blue and black lines).” Please rectify the confusing statement in the legend.</p><p><italic>Reviewer #2</italic>:</p><p>Overall I think this work presents a carefully conducted study. The level of careful evaluation and discussion is very prominent and this work should find an interested readership within the vision neuroscience community.</p><p>The authors use a seed-based approach to investigate the spatial distribution of connectivity patterns in rs-FRMI data. Interestingly, this was done using both eyes-closed and 'fixation' resting state fMRI. The paper further suggests that the retinotopic connectivity patterns (during rest as well as movie viewing) are largely organized along iso-eccentricity bands, which might reflect an important feature of cortical visual processing.</p><p>The findings are presented along with a number of important checks to rule out some of the more obvious confounds, such as BOLD smearing, eye-movements, subject motion and other nuisances.</p><p>The overall bottom-line finding (as e.g. revealed by the authors choice of title and impact statement) of correlation patterns reflecting eccentricity organisation in principle has been demonstrated multiple times before (see e.g. discussion of foveal vs high eccentricity representations in <xref ref-type="bibr" rid="bib79">Smith et al 2012</xref> PNAS or <xref ref-type="bibr" rid="bib46">Heinzle et al 2011</xref> Neuroimage). Particularly the latter paper already demonstrates that the retinotopic organization of visual cortex can be probed using resting state connectivity fMRI.</p><p><italic>Reviewer #3</italic>:</p><p>This manuscript describes a study into the functional connectivity across several human visual cortical areas derived from conventional fMRI resting-state measurements (under fixation and eyes-closed conditions) and measurements while subjects viewed a movie stimulus. The main finding is that the connectivity exhibits a striking organization across the visual cortical areas in which locations that retinotopically represent the same eccentricity of the visual field tend to be correlated, across adjacent cortical areas as well as across the ventral and dorsal subdivisions and across hemispheres. The consistency of this organization throughout the visual cortical areas argues that it is present both in the presence and absence of bottom-up anatomical connectivity. Steps are taken to control for effects such as receptive field size and spatial spread based on modeling their effects using previously reported measurements.</p><p>Overall this manuscript is extremely well written and the results are compelling. Methodologically the approach is well thought out and carefully executed. The intra-run correlation of the movie watching data presented in <xref ref-type="fig" rid="fig7">Figure 7</xref> and the dorsal-ventral correlations in <xref ref-type="fig" rid="fig4">Figure 4B</xref> were particularly compelling.</p><p><italic>My main concern is in the interpretation of these striking findings. As the authors mention, functional connectivity reflects both direct and indirect anatomical connections, and the authors also mention that there have been no reports of anatomical connectivity that could support the observed correlations across eccentricities. The foveal-peripheral biases that are reviewed could be consistent with these observations. While I acknowledge that this is beyond the scope of the current report to identify the origin of this organizational feature, perhaps the manuscript could address further this issue which may be very challenging to reconcile</italic>.</p><p>The issue of the effective resolution along the eccentricity coordinate versus the polar angle coordinate was not discussed, and I wonder how this might affect the findings. Because the cortical areas are elongated along the eccentricity direction there are more voxels to sample the progression of eccentricity than there are to sample the progressions of polar angles. Upon examining the plots in <xref ref-type="fig" rid="fig2">Figure 2B</xref> (and to some extent the plots in <xref ref-type="fig" rid="fig3">Figure 3B</xref>) it appears that there is some consistent correlations across all areas averaged in the retinotopically corresponding region (especially well seen in seeds centered on 1 deg, 2.5 deg, and 5 deg), but also there seems to be some symmetry in the correlation patterns across quadrants in the 2.5 deg and 5 deg maps. One wonders if it is possible that there is a polar angle component to the organization as well, i.e., the correlations may not be only a function of eccentricity but could be a function of polar angle as well. This question gets at my main concern outlined above. Given the reduced resolution in the polar angle direction this question may be difficult to address using this data, and some attempt is made in <xref ref-type="fig" rid="fig8">Figure 8</xref>, but if the resolution is not sufficient perhaps the authors could discuss the potential of an unresolved polar angle organization as an alternate possibility.</p><p><bold>[Editors' note: further revisions were requested prior to acceptance, as described below.]</bold></p><p>Thank you for resubmitting your work entitled “Widespread correlation patterns of fMRI signal across visual cortex reflect eccentricity organization” for further consideration at eLife. Your revised article has been favorably evaluated by Eve Marder (Senior editor), by Timothy Behrens (BRE), and by the original 3 reviewers.</p><p>Often the BRE will choose to handle revisions himself, but on this occasion Tim Behrens wanted to return to the referees as there were some detailed points in the revisions that required their analytical expertise.</p><p>We are now almost ready to accept for publication, but there are a few remaining issues about presentation.</p><p><italic>The main issue is that one reviewer is still concerned by the absence of any reflection of polar geometry in the data, and by the apparent new claim that this cannot be due to the quality of your data. During the course of the revision, a second paper with higher quality data has shown that it is possible to see reflections of polar geometry in resting connectivity (Gravel et al., <ext-link ext-link-type="uri" xlink:href="http://journal.frontiersin.org/Journal/10.3389/fnins.2014.00339/">http://journal.frontiersin.org/Journal/10.3389/fnins.2014.00339/</ext-link>)</italic>.</p><p>The review panel are not sure why this is difficult to find in your data, but we would like you to be clear in the Introduction and the discussion that such structure is visible in higher quality data. ”</p><p>In concert with these changes, we would like you to be slightly clearer about the novelty of your claim with respect to other papers that have shown eccentricity organisation’s – albeit with less detail and precision than you show here.</p><p>For example, we do not think that the following statement in the Introduction is a fair reflection of the literature:</p><p>“In addition, widespread functional correlation patterns have been observed across visual cortex in both macaques (<xref ref-type="bibr" rid="bib57">Leopold et al., 2003</xref>; <xref ref-type="bibr" rid="bib87">Vincent et al., 2007</xref>) and humans (<xref ref-type="bibr" rid="bib65">Nir et al., 2006</xref>; <xref ref-type="bibr" rid="bib66">Nir et al., 2008</xref>; <xref ref-type="bibr" rid="bib94">Yeo et al., 2011</xref>; <xref ref-type="bibr" rid="bib27">Donner et al., 2013</xref>). The governing principles, if any, of such widespread connectivity patterns are still unknown.”</p><p>Reviewer 3 had related concerns that I reproduce here. We do not require you to perform the analysis in point 1 below, if you are clear (as above) that the polar connectivity can be seen in higher quality data, but if you choose to do the analysis, that would be welcome. However, we would like you to address point 2.</p><p>When you resubmit your manuscript, it will not go back to review, but will be assessed by BRE member Tim Behrens.</p><p><italic>Reviewer #3</italic>:</p><p>I thank the authors for responding my concerns and for adding the analyses shown in <xref ref-type="fig" rid="fig12 fig13">Author response images 3 and 4</xref>. The new results presented in <xref ref-type="fig" rid="fig12">Author response image 3</xref>, showing that in some cases within and across V2 and V3 (although V1 is not included?) the correlation falls with angular distance, are particularly helpful.</p><p><italic>I am not sure, however, that the concern I had raised about the potential asymmetry in the resolving power along the eccentricity coordinate and polar angle coordinate has been completely addressed. First, the analyses are restricted to only eccentricities from 1 degree to about 12 degrees, so it is not clear how much a given fMRI voxel will sample along the eccentricity direction relative to the polar angle direction given the cortical magnification and therefore how marginal the effect would be. Also, the new analysis accounting for the potential statistical biases arising from re-binning the data does not address the fact that there is higher effective spatial resolution of the eccentricity coordinate than the polar angle coordinate. The plots in <xref ref-type="fig" rid="fig12">Author response image 3</xref> are the most direct measure of this, since the observation that in some regions the correlation falls off with distance argues that there may be in some cases enough voxels along the polar angle direction to properly resolve a spatial gradient of correlation. Perhaps the authors could address this issue more directly. (Smoothing the data along the eccentricity direction to yield similar resolution for both directions may be one brute-force approach, but perhaps there is another, better way.</italic>)</p><p><italic>Still, the new results provided in <xref ref-type="fig" rid="fig12">Author response image 3</xref> highlight that there is some level of “polar angle connectivity”, which was previously only indirectly demonstrated, e.g., in the second and third panels of <xref ref-type="fig" rid="fig2">Figure 2B</xref> and in <xref ref-type="fig" rid="fig8">Figure 8</xref>. As the authors point out, based on the topographic model regression analyses there is evidence for “local connectivity patterns” including overlapping RF effects and correlations between homotopic regions. Beneath these overlapping RF and homotopic correlations there are also the widespread correlations, and this pattern appears to be weaker in the polar angle direction. Perhaps the authors would consider including the results of <xref ref-type="fig" rid="fig12">Author response image 3</xref> in the main manuscript, and boiling down the results of <xref ref-type="fig" rid="fig9">Figure 9</xref> by reporting the overall variance explained by local connectivity versus widespread connectivity and what proportion of the widespread connectivity is along the eccentricity direction and what proportion is along the polar angle direction. While the authors do point out that these four spatial components are not orthogonal, some clarification of how much of the observed correlations can be explained by widespread patterns would strengthen the manuscript</italic>.</p></body></sub-article><sub-article article-type="reply" id="SA2"><front-stub><article-id pub-id-type="doi">10.7554/eLife.03952.025</article-id><title-group><article-title>Author response</article-title></title-group></front-stub><body><p>Reviewer #1:</p><p>Below, we address each of the reviewer’s specific points.</p><p><italic>Minor comments</italic>:</p><p><italic>1) Both the Abstract and Introduction refer to evidence for more than two dozen retinotopic areas in human visual cortex. However, neither of the cited articles (<xref ref-type="bibr" rid="bib78"><italic>Silver and Kastner, 2009</italic></xref>; <xref ref-type="bibr" rid="bib89"><italic>Wandell and Winawer, 2011</italic></xref>) actually document evidence for two dozen retinotopic areas. Either the statement should be toned down (it is not critical to the study) or appropriate references should be added</italic>.</p><p>We agree with the reviewer that these two articles do not discuss the organization of each area in detail; however, these reviews serve as a general reference for the extent of retinotopic areas across the visual system. These two review articles reference a combined 24 cortical retinotopic areas: V1-3, hV4, VO1-2, PHC1-2, V3A-B, V6, LO1-2, TO1-2, IPS0-5, SPL1, PreCC/SFS, PreCC/IFS and 2 sub-cortical areas: LGN and superior colliculus. References to the particular mapping papers discussing the organization of each of these areas can be found in these review articles. We initially had referenced the original mapping studies. To appropriately cite articles that document the topographic organization in detail for all of these areas, we need to cite a minimum of 12 papers. Due to space limitations of the introduction section, we were not able to include all references to these original papers, and we were only able to cite the two reviews. To address the reviewer’s concern, we added two additional review paper references and explicitly note that interested readers should use theses review papers as a resource, which can refer readers to additional references and the original mapping studies:</p><p>“Through the use of functional magnetic resonance imaging (fMRI), it has now become evident that the human visual system contains over two-dozen visual maps (for review and references to original mapping studies, see <xref ref-type="bibr" rid="bib88">Wandell et al., 2007</xref>, <xref ref-type="bibr" rid="bib78">Silver and Kastner, 2009</xref>, <xref ref-type="bibr" rid="bib89">Wandell and Winawer, 2011</xref>, <xref ref-type="bibr" rid="bib1">Abdollahi et al., 2014</xref>).” paragraph 1 in Introduction section.</p><p><italic>2) Results section, Materials and Methods section and <xref ref-type="fig" rid="fig2"><italic>Figure 2</italic></xref></italic>.</p><p><italic>Please state the extent of visual stimuli for the retinotopic mapping (was it 30 degrees along the meridia, as in Arcaro et al., 2011?). Also, add a scale bar for the eccentricity maps in <xref ref-type="fig" rid="fig2"><italic>Figure 2B</italic></xref>. Finally, consider clarifying that 'peripheral' really means 'near periphery', since nearly half of retinotopic cortex representing the far periphery (> 11 deg) was not mapped</italic>.</p><p>The extent was indeed 30°. We have added this to the text in the Results section and Materials and Methods section.</p><p>We have added outlines of 1.8°, 5.5°, and 15°. Since the eccentricity phases were derived from log-scaled stimuli, distance from the fovea in visual degrees is nonlinear. As discussed in the methods section, the log-scaled mapping stimulus accounts for the cortical magnification factor and yields an approximately uniform distribution of data points across eccentricities. These outlines should provide a good reference for correspondence to visual field position.</p><p>We agree with the reviewer that we did not evaluate the ‘far periphery’ in our binning analyses. Our eccentricity mapping covered the central 15°. By conventional standards, we agree that this is not considered the ‘periphery.’ In the text, we now specifically state the eccentricity ranges that we consider ‘peripheral’ in the Results section and Materials and Methods section. We also state that we did not map responses in the far-periphery (i.e. above 15°), and changed references of periphery to ‘peripheral-most’.</p><p>As discussed in our manuscript, the number of voxels representing 12.5° – 15° eccentricities was minimal due to cortical magnification. As such, we restricted our binning analyses to the central 12.5°, not the central 11° as stated by the reviewer. For the eccentricity binning analysis, we grouped the data into 11 bins. We assume that this confusion arose from <xref ref-type="fig" rid="fig6 fig7">Figures 6-7</xref> where we labeled the bins on the x-axis as ‘Radial Distance’, which could be inferred as visual degrees. We have changed the x-axis label to read ‘Radial Bin Distance’ and clarified the x-axis scale in the figure legend.</p><p><italic>3) Results section, paragraph 12, and <xref ref-type="fig" rid="fig7"><italic>Figure 7</italic></xref> legend. The legend states that “Average individual subject inter-run correlations between hemispheres......steadily decreased at larger distances for all conditions”, which doesn't match what the figure actually shows. The text has it right: “Indeed, inter-run Fisher-transformed correlations on resting state data showed no effect of radial distance (all ps > 0.05), validating the assumption that noise correlations should not be reliable across runs (<xref ref-type="fig" rid="fig7"><italic>Figure 7</italic></xref>, blue and black lines).” Please rectify the confusing statement in the legend</italic>.</p><p>We thank the reviewer for catching this error. We have corrected the legend to state:</p><p>“For the movie viewing condition, average individual subject inter-run correlations between hemispheres as well as between dorsal and ventral portions of V2 and V3 were strongest at the 0-radial distance (iso-eccentricity) and steadily decreased at larger distances. For resting data, correlations did not vary as a function of radial distance. See <xref ref-type="fig" rid="fig6">Figure 6</xref>.”</p><p><italic>Reviewer #2</italic>:</p><p><italic>The overall bottom-line finding (as e.g. revealed by the authors choice of title and impact statement) of correlation patterns reflecting eccentricity organisation in principle has been demonstrated multiple times before (see e.g. discussion of foveal vs high eccentricity representations in <xref ref-type="bibr" rid="bib79"><italic>Smith et al 2012</italic></xref> PNAS or <xref ref-type="bibr" rid="bib46"><italic>Heinzle et al 2011</italic></xref> Neuroimage). Particularly the latter paper already demonstrates that the retinotopic organization of visual cortex can be probed using resting state connectivity fMRI</italic>.</p><p>We agree with the reviewer that a few prior papers have demonstrated a coarse distinction in the patterns of activity during ‘rest’ across cortex representing ‘foveal’ and ‘peripheral’ visual space (<xref ref-type="bibr" rid="bib79">Smith et al. 2012</xref>, ; <xref ref-type="bibr" rid="bib87">Vincent et al. 2007</xref>; see also ; <xref ref-type="bibr" rid="bib63">Nallasamy & Tsao 2011</xref>). We briefly reviewed this in our Discussion section. In contrast to our study, however, these prior studies do not show a large-scale, topographic correlation pattern that reflects the underlying fine-scale eccentricity organization. We have added text to the Discussion to address some of these concerns. Here, we further discuss why our current study provides a substantial new understanding of visual cortical correlation patterns beyond what previously has been shown:</p><p>1) We demonstrated a topography of correlation patterns that reflects the underlying, fine-scale eccentricity organization. Prior studies found a coarse distinction between ‘foveal’ and ‘peripheral’ cortex using data-driven analyses (e.g., ICA, clustering), but did not look for / or report a systematic organization reflecting the underlying eccentricity map. As we discuss in our study, coarse center-periphery distinction not only has been found in early visual areas, but also in higher order extrastriate areas (<xref ref-type="bibr" rid="bib58">Levy et al. 2001</xref>. ; <xref ref-type="bibr" rid="bib45">Hasson et al. 2002</xref>; <xref ref-type="bibr" rid="bib59">Malach et al. 2002</xref>). Its origin may reflect two very different gross processes computed at foveal and peripheral areas (Malach, 2004 TiCS). In contrast, our findings report a novel fine-scale topography of eccentricity-based correlation patterns, in which, for example, local regions representing 2°, 5°, 10° eccentricities are most correlated, respectively, with other regions representing 2°, 5°, 10° eccentricities within and across early visual (V1-V3) and extrastriate (hV4) areas. Such a systematic correlation pattern suggests orderly integration processes across the whole visual field at multiple levels of the processing hierarchy, not just a foveal-peripheral dichotomy. As such, we feel that these eccentricity-specific correlation patterns go substantially beyond prior work.</p><p>2) As the reviewer stated, Heinzle and colleagues demonstrated that the retinotopic organization of visual cortex, specifically between areas V1 – V3, can be probed using resting state connectivity measures on fMRI data. Consistent with Heinzle and colleagues, we found retinotopically-specific correlation patterns based on overlapping response fields between V1 and V3. Importantly, however, we also found significant overlapping response field effects for other area pairs including V2 and hV4. Distinct from the Heinzle study, we found eccentricity-specific correlation patterns that cannot be accounted for by overlapping response fields. Heinzle and colleagues did not probe such separate eccentricity-based connectivity. As such, we feel that our study provides new insight into the intrinsic organization of visual cortex beyond the Heinzle study while also validating this previous report of overlapping receptive field connectivity. We have already cited the Heinzle study in our Discussion about overlapping receptive field connectivity and have discussed extensively the novel aspects of our study.</p><p><italic>Most importantly, Raemaekers et al (2014, Neuroimage) present a detailed analysis of ICA-derived spatial networks in V1-3 from high-resolution rs-FMRI at 7T. These authors argue for more complicated overlapping organisations (eccentricity as well as polar angle based), where the finer-grained organisation emerges only after filtering large-scale network fluctuations. This leads to a more extensive set of findings in Raemaekers et al. relative to this work. A discussion of their paper would be crucially important but is lacking from this manuscript. This also needs to include a thorough discussion of methodological differences, e.g. the use of multi-variate rather than uni-variate statistical methods. It appears that the inherent limitations of seed-based analysis (in terms of difficulties accounting for additional signals) may be responsible for differences between the two manuscripts</italic>.</p><p>We thank the reviewer for raising these important points.</p><p>In relation to the first point (relation between the present study and that of <xref ref-type="bibr" rid="bib70">Raemaekers et al. 2014</xref>), we have now expanded our Discussion of the <xref ref-type="bibr" rid="bib70">Raemaekers et al. (2014)</xref> paper, as suggested, and have made clearer how the present results relate to and extend that prior work:</p><p>“At a fine-scale, there have been correlation patterns reflecting overlapping RF within hemispheres and homotopic connections between hemispheres (<xref ref-type="bibr" rid="bib46">Heinzle et al., 2011</xref>; <xref ref-type="bibr" rid="bib42">Haak et al., 2012</xref>; <xref ref-type="bibr" rid="bib18">Butt et al., 2013</xref>; <xref ref-type="bibr" rid="bib70">Raemaekers et al., 2014</xref>). Raemaekers and colleagues (2014) recently reported that such fine-scaled connectivity was only observable after filtering coarse-scale components. Consistent with prior studies (<xref ref-type="bibr" rid="bib46">Heinzle et al., 2011</xref>; <xref ref-type="bibr" rid="bib42">Haak et al., 2012</xref>; <xref ref-type="bibr" rid="bib18">Butt et al., 2013</xref>), we observed overlapping RF within hemispheres and homotopic connections between hemispheres without such data filtering. At a coarse scale, there has been evidence for a general foveal-peripheral distinction (<xref ref-type="bibr" rid="bib87">Vincent et al., 2007</xref>; <xref ref-type="bibr" rid="bib63">Nallasamy and Tsao, 2011</xref>; <xref ref-type="bibr" rid="bib79">Smith et al., 2012</xref>; <xref ref-type="bibr" rid="bib70">Raemaekers et al., 2014</xref>), but these studies did not report a systematic topography of correlation patterns that reflects the underlying eccentricity organization, nor were effects of overlapping RFs and cortical distance controlled. Evidence for a more systematic relation to eccentricity was reported between V1 and ventral V3 (<xref ref-type="bibr" rid="bib94">Yeo et al., 2011</xref>), though such data are also consistent with overlapping RF connectivity as well as cortical distance since only ventral V3 was probed. Here, we explicitly and quantitatively combine all of these connectivity phenomena in a fine-grained analysis and across a wider range of brain regions, replicating previous findings of overlapping RF and homotopic correlation patterns.”</p><p>In relation to the second point (methodological differences and univariate vs. multivariate analyses), we present arguments and analyses that support our methodological approach, and buttress the novelty of our findings.</p><p>To summarize the prior relevant findings: Raemaekers and colleagues (2014) demonstrated 1) an effect of overlapping response field connectivity between visual areas V1, V2, and V3 with correlations decreasing between areas as a function of visual degree distance, 2) homotopic correlations across the right and left hemispheres within individual areas, and 3) coarse ICA components spanning various portions of visual cortex. Importantly, the study by Raemaekers and colleagues did not report fine scale eccentricity-based correlation patterns. As stated above, we have expanded our original discussion of this paper in our Discussion section. Here, we further discuss the differences between our study and the Raemaekers and colleagues (2014) study:</p><p>Overlapping response field connectivity:</p><p>1) Similar to our results and the Heinzle study, Raemaekers and colleagues found significant functional connectivity effects based on overlapping response fields. As opposed to the Heinzle paper and our paper, Raemaekers and colleagues only observed these effects after filtering their data with ICA components. Consistent with Heinzle and colleagues, we found significant overlapping response field connectivity effects without such filtering. Raemaekers and colleagues discussed this difference in regards to the Heinzle study (see <xref ref-type="bibr" rid="bib70">Raemaekers et al., 2014</xref>, p. 918-9). In their discussion of the Heinzle study, Raemaekers and colleagues noted a few differences in the methods and analyses, but ultimately concluded: “Whether this difference in methods can fully explain the difference in results is unclear…” We agree that there was considerable variability in the imaging and analytical approaches across all three studies (Heinzle’s, Raemaekers’, and ours). As such, it is unclear why filtering was needed in the Raemaekers study to observe topographic patterns of connectivity, but not in our study or the Heinzle study.</p><p>Eccentricity-specific connectivity:</p><p>1) We observed eccentricity-specific topographic correlation patterns distinguishable from overlapping response field connectivity. Raemaekers and colleagues did not find a significant difference in their topographic connectivity analyses between radial or tangential dimensions during ‘rest’ (see <xref ref-type="fig" rid="fig4">Figure 4C&F</xref> in <xref ref-type="bibr" rid="bib70">Raemaekers et al. 2014</xref>), but they also did not directly test for eccentricity-specific connectivity distinct from overlapping response fields (e.g., between dorsal and ventral cortex). Raemaekers and colleagues discuss the lack of a radial/tangential distinction in their discussion: “The results regarding a radial/tangential connectivity bias were inconclusive” (see <xref ref-type="bibr" rid="bib70">Raemaekers et al., 2014</xref>, p. 919). Given that we found strong evidence for eccentricity-specific topographic connectivity across several analyses, we feel that our current study provides novel results not previously shown (or directly tested) by Raemaekers and colleagues.</p><p>2) Raemaekers and colleagues evaluated polar angle and eccentricity distance connectivity effects during ‘rest’ only for intra-areal patterns (see <xref ref-type="fig" rid="fig6">Figure 6</xref> in <xref ref-type="bibr" rid="bib70">Raemaekers et al. 2014</xref>). We found that intra-areal correlation patterns differ from inter-areal correlation patterns, and likely contain strong anatomical distance effects (see our <xref ref-type="fig" rid="fig6s2">Figure 6–figure supplement 2</xref>). We found that eccentricity-specific correlation patterns were most apparent between regions where (anatomical) distance and overlapping response field effects were minimal (e.g., between dorsal and ventral V2 / V3). This was not directly evaluated / reported in the Raemaekers study.</p><p>3) Raemaekers and colleagues observed significant topographic connectivity effects for polar angle and eccentricity mapping experiments (see <xref ref-type="fig" rid="fig3">Figure 3B&C</xref> in <xref ref-type="bibr" rid="bib70">Raemaekers et al. 2014</xref>). Such topographic patterns were likely to be induced by the correlation in the stimulus structure, and does not necessarily reflect intrinsic polar angle or eccentricity connectivity (i.e., connectivity was driven by the stimulus systematically moving across visual space along either radial or tangential dimensions).</p><p>4) Raemaekers and colleagues reported general ‘foveal’ and ‘peripheral’ ICA components in 5 subjects (see <xref ref-type="fig" rid="fig7">Figure 7C&D</xref> in <xref ref-type="bibr" rid="bib70">Raemaekers et al. 2014</xref>). These ICA components appear consistent with previously reported foveal and peripheral distinctions (see above section discussing <xref ref-type="bibr" rid="bib79">Smith et al 2012</xref>). As discussed above, this differs from our results, which demonstrates fine-scale, topographic correlation patterns that reflect the underlying eccentricity organization. In the Raemaekers’ study, the ICA components were not directly compared with eccentricity maps. In fact, the yellow peaks of the peripheral ICA components (see their <xref ref-type="fig" rid="fig7">Figure 7D</xref>) fall outside their retinotopically-identified visual areas. Further, there was no evaluation of the consistency of these ICA components across subjects and additional ICA components orthogonal to the ‘foveal’ and ‘peripheral’ ICAs were also reported. As such, it is difficult to interpret the significance and consistency of these ICA components, and their specific relation to the retinotopic organization of visual cortex. Below in this letter, we further investigate the interpretability of ICA components and their correspondence to the effects we report in the present manuscript.</p><p>Methodological differences:</p><p>Numerous methodological differences between our study and Raemaekers and colleague’s study may account for the increased sensitivity in our study to the large-scale eccentricity organization.</p><p>1) There were significant differences in the extent of the visual field mapped. We mapped eccentricity representations out to 15° (from central fixation) and included the central 12.5° in all of our bin analyses. Raemaekers and colleagues only mapped out to 7.5°. Though we observed significant effects of eccentricity even within a few degrees of visual space (see <xref ref-type="fig" rid="fig6">Figure 6</xref>), the restricted coverage of visual space may have limited Raemaekers and colleagues’ ability to reliably detect eccentricity-specific correlation patterns.</p><p>2) The filtering applied by Raemaekers may have removed key components. In their discussion, Raemaekers and colleagues acknowledged that the data filtering might have removed a foveal / peripheral distinction in their topographic connectivity analyses (see <xref ref-type="bibr" rid="bib70">Raemaekers et al., 2014</xref>, p. 919): “There were, however, complicating factors that may have caused a lack of significant results. Some of the coarse-scale networks that were detected with ICA contained features that would either contribute to a radial or tangential connectivity bias, and this may have obscured results in the initial analysis… The fluctuations within these (radial/tangential) networks were, however, filtered from the data in the second analysis, which may have attenuated the observed effect.”</p><p>3) As the reviewer suggested, it is possible that the difference between the Raemaekers study and our study reflects differences between data-driven multivariate and seed-based correlation analyses. Multivariate methods are potentially more powerful than univariate approaches, but their output can also be more difficult to definitively interpret. This is the case for complex methods such as ICA, and especially for spatial ICA applied to fMRI data, where the basic BOLD signal properties that separate components remain under debate (Daubechies, PNAS 2009). Prior studies using ICA analyses have only identified a broad foveal-peripheral distinction in ‘rest’ data. Here, we demonstrate topographic correlation patterns across the central 15° of visual space that reflect the underlying eccentricity organization.</p><p>It is possible that ICA analyses may be less prone to revealing such topography. To investigate the ability of ICA analyses to reveal fine-scaled eccentricity organization, we applied ICA (FSL’s MELODIC) to data from an eccentricity mapping experiment (<xref ref-type="fig" rid="fig10">Author response image 1</xref>).<fig id="fig10" position="float"><label>Author response image 1.</label><caption><p>Spatially-specific ICA components from stimulus-driven (eccentricity mapping) data.</p></caption><graphic xlink:href="elife03952f010"/></fig></p><p>We used eccentricity-mapping data, since the ring stimulus evokes fine-scale eccentricity-specific topographic patterns of activity that can be differentiated in the BOLD signal. As seen in the phase maps, the underlying structure of activity patterns is spatially contiguous, not discrete, and can be clearly identified with conventional retinotopic mapping analyses (see <xref ref-type="bibr" rid="bib29">Engel et al. 1997</xref>; <xref ref-type="bibr" rid="bib82">Swisher et al. 2007</xref>; <xref ref-type="bibr" rid="bib6">Arcaro et al. 2009</xref>). In each subject, we identified about 45 ICA components. Most components were associated with distributed noise or relegated to non-visual networks (e.g., DMN and fronto-parietal). Across 5 subjects, we consistently identified between 2-3 ICA components in these data that were localized to contiguous parts of visual cortex and corresponded to spatially specific representations of the visual field. Each ICA component spanned several degrees of eccentricity. The ‘Peripheral’ ICA component was located largely outside of our eccentricity maps, and likely corresponds to regions of visual space >15° eccentricity. The ‘Mid’ ICA component spanned a region of visual space ranging between 4° – 8° eccentricities. In 3 of 5 subjects, we identified a ‘Foveal’ ICA component around the occipital pole and lateral surface that corresponded to foveal space less than ∼2° eccentricity.</p><p>In one subject, we identified another ICA component overlapping 13 - 15° eccentricity and part of ICA 1. ICA components did not reveal any finer-grained organization such as seen in the phase maps. In a few subjects, there were 1-2 additional ICAs that did not appear to be spatially-specific, and either spanned most of the visual field or covered multiple, non-contiguous eccentricity ranges (e.g., covered both ∼1° - 4° and ∼8° - 10°). Though these results demonstrate coarse eccentricity organization and could be consistent with the existence of a topographic organization, the components by themselves do not reflect a fine-scaled eccentricity organization as shown in the phase maps. These data suggest that such ICA analyses may not be sensitive enough to reveal fine-scale spatial gradients in data (also see Daubechies et al. 2009 PNAS), and may demonstrate why previous studies employing ICA analyses only report broad foveal-peripheral distinctions.</p><p>Upon close inspection of our ICA analyses with prior data, it appears that the previously reported ‘peripheral’ ICA in resting state studies does not correspond to the range of eccentricities probed in our current experiment. In our ICA analyses, our ‘Peripheral’ ICA component was located in anterior parts of the calcarine sulcus and retrosplenial cortex. This anatomically corresponds well with the ‘peripheral’ ICA previously reported by <xref ref-type="bibr" rid="bib79">Smith et al. 2012</xref> and <xref ref-type="bibr" rid="bib70">Raemaekers et al. 2014</xref> (see <xref ref-type="fig" rid="fig10">Author response image 1</xref>, right column). However, this component falls mainly outside of the retinotopically defined visual areas (> 15° eccentricity) in our current study (see <xref ref-type="fig" rid="fig10">Author response image 1</xref>, white line on flat maps). The ‘Mid’ and ‘Foveal’ ICA components were located around the occipital pole and overlap with each subject’s eccentricity map. These ICAs anatomically appear to correspond to the ‘foveal’ component previously reported. As such, the prior ‘foveal’ and ‘peripheral’ differentiation in resting state data may not be directly relatable to the eccentricity-based correlation patterns shown in our current study.</p><p>4) Though our ICA analysis distinguished foveal from peripheral cortex, it was not able to reveal a fine-scaled topography from eccentricity mapping data. Next, we tested whether other data-driven analyses based on correlation patterns between individual data points (i.e., nodes) could reveal such organization. We used a k-means clustering algorithm to segment our data. At a coarse segmentation (k=2), we observed a parcellation of visual cortex in individual subjects that reflects a foveal / peripheral distinction (note: this distinction is within our retinotopic maps, not outside like the ICA analyses). At larger k-segmentations, additional clusters appeared to be symmetric around the fovea, consistent with an eccentricity-based organization. At larger k-segmentations, there was also some differentiation between early visual (V1-V3) and higher-order cortical areas (V3A/B, LO / VO / MT / IPS), which is consistent with observations from our binning analyses and prior reports (<xref ref-type="bibr" rid="bib79">Smith et al. 2012</xref>). i.e., as seen in our binning analyses, peripheral regions of V1 were more strongly correlated with peripheral regions of V3A/B (vs. foveal V3A/B) even though peripheral regions of V1 were more strongly correlated with peripheral V3 (vs. peripheral of V3A/B). This differentiation between early visual and higher order areas is orthogonal to the current question of eccentricity-based connectivity patterns. To illustrate the relation of these clusters to the underlying eccentricity organization while minimizing the broad regional differentiation between early visual and higher-order cortex, we color coded each cluster based on the strength of their correlation with the foveal cluster in early visual cortex. In this visualization, a cluster organization similar to the eccentricity organization is apparent for segmentations based on intra- and inter- hemisphere correlations. See <xref ref-type="fig" rid="fig11">Author response image 2</xref>.<fig id="fig11" position="float"><label>Author response image 2.</label><caption><p>K-Means Cluster Analysis.</p></caption><graphic xlink:href="elife03952f011"/></fig></p><p>This analysis demonstrates that a gradient of eccentricity-based correlation patterns similar to that revealed in our seed-based analyses can be identified with data-driven segmentation methods. However, clustering was variable across subjects, and multiple organization patterns were evident in the data. As such, we feel that our current correlation analyses are better suited for examining functional connectivity patterns within visual cortex specifically related to eccentricity organization.</p><p>Reviewer #3:</p><p><italic>My main concern is in the interpretation of these striking findings. As the authors mention, functional connectivity reflects both direct and indirect anatomical connections, and the authors also mention that there have been no reports of anatomical connectivity that could support the observed correlations across eccentricities. The foveal-peripheral biases that are reviewed could be consistent with these observations. While I acknowledge that this is beyond the scope of the current report to identify the origin of this organizational feature, perhaps the manuscript could address further this issue which may be very challenging to reconcile</italic>.</p><p>We agree with the reviewer that the structure-function relationship of the eccentricity-specific correlation pattern needs to be explored and deserves further discussion. A foveal-peripheral distinction in connectivity has been noted in several prior anatomical studies (<xref ref-type="bibr" rid="bib20">Colby et al., 1988</xref>; <xref ref-type="bibr" rid="bib62">Nakamura et al., 1993</xref>; <xref ref-type="bibr" rid="bib36">Gattass et al., 2005</xref>; <xref ref-type="bibr" rid="bib84">Ungerleider et al., 2008</xref>; <xref ref-type="bibr" rid="bib85">Ungerleider et al., 2014</xref>) and does offer a plausible anatomical basis for the functional eccentricity-based correlation patterns observed in our data. We have expanded discussion of a possible structure-function relationship:</p><p>“Such a distinction has been characterized in the patterns of supra-areal anatomical connections between early visual and extrastriate cortex in non-human primates (<xref ref-type="bibr" rid="bib71">Rosa, 2002</xref>; <xref ref-type="bibr" rid="bib36">Gattass et al., 2005</xref>; <xref ref-type="bibr" rid="bib73">Rosa and Tweedale, 2005</xref>; <xref ref-type="bibr" rid="bib72">Rosa et al., 2009</xref>; <xref ref-type="bibr" rid="bib16">Buckner and Yeo, 2014</xref>). It is not known whether these anatomical connectivity patterns are ‘bi-modal’, and only distinguish central and peripheral space, or are part of a topography where connectivity patterns with intermediate eccentricity representations are distinguishable from central and peripheral connectivity profiles. Our results would predict that the anatomical connections reflect a gradient, though this remains to be explored. In particular, feedback projections from extrastriate areas with receptive fields covering wide swaths of the visual field to early and intermediate visual areas could facilitate such widespread, eccentricity-dependent correlation patterns. It is interesting to note that when comparing the profile of anatomical connectivity between V2 / V4 and higher order cortex (<xref ref-type="fig" rid="fig7">Fig 7</xref>; <xref ref-type="bibr" rid="bib36">Gattass et al. 2005</xref>) with the organization of eccentricity across visual cortex (<xref ref-type="bibr" rid="bib15">Brewer et al. 2002</xref>; <xref ref-type="bibr" rid="bib56">Kolster et al. 2009</xref>; <xref ref-type="bibr" rid="bib7">Arcaro et al. 2011</xref>, also see ; <xref ref-type="bibr" rid="bib6">Arcaro et al. 2009</xref> and ; Kolster et al. 2010), it is clear that higher order areas connected with peripheral parts of V2 and V4 (e.g., PO, PIP, LIP, DP, TF) have a peripheral visual field bias and higher order areas connected with foveal parts of V2 and V4 (e.g., TEO, TE) have a foveal visual field bias.”</p><p><italic>The issue of the effective resolution along the eccentricity coordinate versus the polar angle coordinate was not discussed, and I wonder how this might affect the findings. Because the cortical areas are elongated along the eccentricity direction there are more voxels to sample the progression of eccentricity than there are to sample the progressions of polar angles. Upon examining the plots in</italic> <xref ref-type="fig" rid="fig2"><italic>Figure 2B</italic></xref> <italic>(and to some extent the plots in</italic> <xref ref-type="fig" rid="fig3"><italic>Figure 3B</italic></xref><italic>) it appears that there is some consistent correlations across all areas averaged in the retinotopically corresponding region (especially well seen in seeds centered on 1 deg, 2.5 deg, and 5 deg), but also there seems to be some symmetry in the correlation patterns across quadrants in the 2.5 deg and 5 deg maps. One wonders if it is possible that there is a polar angle component to the organization as well, i.e., the correlations may not be only a function of eccentricity but could be a function of polar angle as well. This question gets at my main concern outlined above. Given the reduced resolution in the polar angle direction this question may be difficult to address using this data, and some attempt is made in</italic> <xref ref-type="fig" rid="fig8"><italic>Figure 8</italic></xref><italic>, but if the resolution is not sufficient perhaps the authors could discuss the potential of an unresolved polar angle organization as an alternate possibility</italic>.</p><p>We thank the reviewer for raising this point. We agree that it is very important to consider the effective resolution along eccentricity and polar angle dimensions. An imbalance of voxels between eccentricity and polar angle dimensions could affect interpretation of the results, in particular, whether any such polar angle connectivity is detectable at our current imaging resolution. This could manifest it two ways: 1) if the sampling space of individual voxels is greater along the eccentricity dimension than the polar angle dimension, and 2) if the distribution of voxels across bins in our topographic model analysis is imbalanced. When analyzing the data, we were well aware of these potential issues, and feel that aspects of our original analyses suggest that this was not actually a major issue. In addition, we performed a few new control analyses to further rule out this potential issue. Though our control analyses provide strong evidence for a lack of polar angle connectivity, we agree that this is an important point that deserves mention in the current study, and now include a description of the issue along with a new control analysis in paragraphs 16-18 of the Results section.</p><p>1) As the reviewer noted, cortical areas V1, V2, and V3 are elongated along the eccentricity dimension. As such, a given voxel will generally sample a greater extent along the polar angle dimension than eccentricity, though this becomes marginal as one moves further into the periphery due to the cortical magnification factor. This imbalance in sampling at the individual voxel level could have limited our ability to detect polar angle-based connectivity patterns. In the topographic model analysis, we did not find a consistent effect of polar angle connectivity after accounting for overlapping response fields (see <xref ref-type="fig" rid="fig9">Fig 9</xref>). However, we found that eccentricity and polar angle models equally contributed to the variance explained by the overlapping RF model paragraphs 18-19 of Results section. That is, we found some effect of polar angle connectivity, but it could be attributed to overlapping RFs. Given this dissociation in polar angle connectivity patterns with respect to within and outside the extent of an overlapping RF, we feel that our imaging resolution and analytical approach was sensitive enough for detecting polar angle-based connectivity.</p><p>2) To further evaluate whether an effect of polar angle-based connectivity was detectable in our data, we performed the binning analysis with respect to angular distance (collapsing across eccentricities). See <xref ref-type="fig" rid="fig12">Author response image 3</xref>. For within-quadrant + within-hemisphere and mirror-symmetric, within-quadrant + between-hemisphere comparisons, correlations steadily decreased at larger angular distances suggesting an effect of polar angle distance which likely reflects overlapping response field and homotopic connections, respectively. We found no effect of angular distance between dorsal and ventral V2 and V3 based on actual angular distance or when reflecting across the horizontal (i.e., mirror symmetry) with correlations either remaining about equal to that at iso-polar angle or actually increasing for larger angular distances. For all other bin pairs, there was either little difference in the correlation value at larger angular bin distances or correlations actually got stronger.<fig id="fig12" position="float"><label>Author response image 3.</label><caption><p>Polar Angle-based Intra-run Connectivity.</p></caption><graphic xlink:href="elife03952f012"/></fig></p><p>3) It is also possible that any imbalance in the number of voxels across data bins could bias our results and ability to detect polar angle-based connectivity. To test this, we conducted a new analysis where we equalized the number of nodes in each bin of the topographic model analysis. For bins that contained more than 20 nodes, we subsampled the data to only include 20 nodes. In some subjects, a few bins contained less than 20 nodes. We conducted the new analyses both including and excluding these bins with few nodes. Eccentricity and (lack of) polar angle residual correlations were observed in these new analyses and were consistent with the original results shown in <xref ref-type="fig" rid="fig9">Figure 9</xref>. See <xref ref-type="fig" rid="fig13">Author response image 4</xref>. In some cases, the effects of residual eccentricity correlations were slightly stronger relative to the original analyses (e.g., a few pairs with hV4). Similar results were observed for bins of 30 and 50 nodes as well as whether the bins with few nodes were included or excluded. We now include a description of this control analysis in the Results section.<fig id="fig13" position="float"><label>Author response image 4.</label><caption><p>Equal Bins.</p></caption><graphic xlink:href="elife03952f013"/></fig></p><p><bold>[Editors' note: further revisions were requested prior to acceptance, as described below.]</bold></p><p><italic>The main issue is that one reviewer is still concerned by the absence of any reflection of polar geometry in the data, and by the apparent new claim that this cannot be due to the quality of your data. During the course of the revision, a second paper with higher quality data has shown that it is possible to see reflections of polar geometry in resting connectivity (Gravel et al. <ext-link ext-link-type="uri" xlink:href="http://journal.frontiersin.org/Journal/10.3389/fnins.2014.00339/">http://journal.frontiersin.org/Journal/10.3389/fnins.2014.00339/</ext-link>).</italic></p><p>To the best of our understanding, our results are entirely consistent with those of <xref ref-type="bibr" rid="bib41">Gravel et al. (2014)</xref>. We became aware of this study while our manuscript was under re-review. This is an excellent paper demonstrating that retinotopic connectivity (based on overlapping RFs) can be identified in resting state data. We now cite the paper in our Introduction and Discussion sections. In their study, Gravel and colleagues reconstruct polar angle and eccentricity components in areas V2 and V3 for individual subjects based on modeling RF-based connectivity with V1 (referred to as connective field modeling). Consistent with their results, an effect of RF connectivity is qualitatively apparent in the correlation structure of our data (<xref ref-type="fig" rid="fig8">Figure 8</xref>), and we found a significant effect of overlapping RF connectivity in our topographic model analysis, which we report in paragraphs 15-17 of Results section. Further, we show in <xref ref-type="fig" rid="fig10">Author response image 1</xref> that correlations drop off with angular distance for areas that contain overlapping RFs (e.g., V2v – V3v) as well as for mirror symmetric connections across hemispheres (e.g., V2v RH – V2v LH). In contrast, we do not find clear effects of polar angle connectivity between areas that have minimal / non-overlapping RFs (e.g., V2v – V2d). Gravel and colleagues (2014) did not explore connectivity patterns based on non-overlapping RFs. The polar angle and eccentricity maps that are generated from resting state data in the Gravel study were derived from inter-areal overlapping RF models, and do not account for structure in connectivity between non-overlapping RF regions. To our knowledge, no prior study has systematically explored / shown effects of polar angle (or eccentricity) connectivity attributed to regions with non-overlapping RFs, which is the focus of our manuscript.</p><p>We now cite the Gravel paper in the Introduction:</p><p>Similarly, fMRI connectivity studies in humans have demonstrated topographically-local correlations between regions with overlapping visual field representations (<xref ref-type="bibr" rid="bib46">Heinzle et al., 2011</xref>; <xref ref-type="bibr" rid="bib42">Haak et al., 2012</xref>; <xref ref-type="bibr" rid="bib18">Butt et al., 2013</xref>; <xref ref-type="bibr" rid="bib27">Donner et al., 2013</xref>; <xref ref-type="bibr" rid="bib41">Gravel et al., 2014</xref>; <xref ref-type="bibr" rid="bib70">Raemaekers et al., 2014</xref>).</p><p>We also specifically mention the Gravel paper’s results in our Discussion:</p><p>By modeling V1 inter-areal connections based on overlapping RFs, Gravel and colleagues (2014) generated polar angle and eccentricity maps for V2 and V3.</p><p><italic>In concert with these changes, we would like you to be slightly clearer about the novelty of your claim with respect to other papers that have shown eccentricity organisation’s – albeit with less detail and precision than you show here</italic>.</p><p><italic>For example, we do not think that the following statement in the Introduction is a fair reflection of the literature</italic>:</p><p><italic>⋖In addition, widespread functional correlation patterns have been observed across visual cortex in both macaques (</italic><xref ref-type="bibr" rid="bib57"><italic>Leopold et al., 2003</italic></xref>; <xref ref-type="bibr" rid="bib87"><italic>Vincent et al., 2007</italic></xref><italic>) and humans (</italic><xref ref-type="bibr" rid="bib65"><italic>Nir et al., 2006</italic></xref>; <xref ref-type="bibr" rid="bib66"><italic>Nir et al., 2008</italic></xref>; <xref ref-type="bibr" rid="bib94"><italic>Yeo et al., 2011</italic></xref>; <xref ref-type="bibr" rid="bib27"><italic>Donner et al., 2013</italic></xref><italic>). The governing principles, if any, of such widespread connectivity patterns are still unknown.”</italic></p><p>This is a fair point. We agree that the cited statement could be improved to reflect previous literature. We have changed the text to read:</p><p>“In addition, widespread functional correlation patterns have been observed across visual cortex in both macaques (<xref ref-type="bibr" rid="bib57">Leopold et al., 2003</xref>; <xref ref-type="bibr" rid="bib87">Vincent et al., 2007</xref>) and humans (<xref ref-type="bibr" rid="bib65">Nir et al., 2006</xref>; <xref ref-type="bibr" rid="bib66">Nir et al., 2008</xref>; <xref ref-type="bibr" rid="bib94">Yeo et al., 2011</xref>; <xref ref-type="bibr" rid="bib27">Donner et al., 2013</xref>). These patterns contain broad differences between foveal and peripheral cortex (<xref ref-type="bibr" rid="bib70">Raemaekers et al., 2014</xref>), though may also be tied to the fine-scale organization of individual retinotopic maps.”</p><p>We have also expanded discussion of previous findings and the relation to our findings in the Discussion section.</p><p><italic>Reviewer 3 had related concerns that I reproduce here. We do not require you to perform the analysis in point 1 below, if you are clear (as above) that the polar connectivity can be seen in higher quality data, but if you choose to do the analysis, that would be welcome. However, we would like you to address point 2</italic>.</p><p>Reviewer #3:</p><p><italic>I thank the authors for responding my concerns and for adding the analyses shown in <xref ref-type="fig" rid="fig12 fig13">Author response images 3 and 4</xref>. The new results presented in <xref ref-type="fig" rid="fig12">Author response image 3</xref>, showing that in some cases within and across V2 and V3 (although V1 is not included?) the correlation falls with angular distance, are particularly helpful</italic>.</p><p><italic>I am not sure, however, that the concern I had raised about the potential asymmetry in the resolving power along the eccentricity coordinate and polar angle coordinate has been completely addressed. First, the analyses are restricted to only eccentricities from 1 degree to about 12 degrees, so it is not clear how much a given fMRI voxel will sample along the eccentricity direction relative to the polar angle direction given the cortical magnification and therefore how marginal the effect would be. Also, the new analysis accounting for the potential statistical biases arising from re-binning the data does not address the fact that there is higher effective spatial resolution of the eccentricity coordinate than the polar angle coordinate. The plots in <xref ref-type="fig" rid="fig12">Author response image 3</xref> are the most direct measure of this, since the observation that in some regions the correlation falls off with distance argues that there may be in some cases enough voxels along the polar angle direction to properly resolve a spatial gradient of correlation. Perhaps the authors could address this issue more directly. (Smoothing the data along the eccentricity direction to yield similar resolution for both directions may be one brute-force approach, but perhaps there is another, better way.)</italic></p><p>We appreciate the reviewer’s suggestions regarding the issue of the potential asymmetry in voxel sampling along eccentricity and polar angle dimensions. We believe that our current analyses do address this concern. Our analyses and imaging parameters were sensitive enough for detecting spatial patterns of correlated bold signal along both angular and radial dimensions. We observed significant effects of angular connectivity between regions with overlapping RFs and for mirror symmetrical regions between hemispheres (<xref ref-type="fig" rid="fig6s2">Figure 6–figure supplement 2</xref>, also see <xref ref-type="fig" rid="fig8">Figure 8</xref>). These effects are consistent with those previously reported (<xref ref-type="bibr" rid="bib46">Heinzle et al. 2011</xref>; <xref ref-type="bibr" rid="bib41">Gravel et al. 2014</xref>; <xref ref-type="bibr" rid="bib70">Raemaekers et al. 2014</xref>). The effects were consistent across binning sizes (i.e., the slopes for individual area pairs did not significantly differ across binning sizes), suggesting that the number of voxels in each bin did not greatly affect our sensitivity for revealing such patterns of connectivity. Thus, a subset of our analyses (<xref ref-type="fig" rid="fig6s2">Figure 6 – figure supplement 2</xref>) provides a positive control for our finding of eccentricity-based connectivity in the absence of angular connectivity for regions with non-overlapping RFs (<xref ref-type="fig" rid="fig9">Figure 9</xref>).</p><p>Any bias in voxel sampling along angular and radial dimensions would affect the (average) signal of individual bins. Since bin pairs, not individual bins, determined the labeling of correlations into overlapping and non-overlapping RF groups (i.e., all bins contributed to both overlapping and non-overlapping RF analyses), any bias would have affected both overlapping and non-overlapping RF analyses. We observed comparable contributions of polar angle and eccentricity predictors for the variance explained by the overlapping RF model in our topographic model analysis (i.e., there was no bias for these comparisons). Due to the topographic nature of cortex, overlapping RF bin pairs will be, on average, in much closer anatomical proximity than non-overlapping RF bin pairs, and thus would actually be more susceptible to any biases driven by lack of sensitivity along the polar angle dimension. Thus, it is unlikely that any such bias in voxel sampling prevented us from resolving effects of polar angle connectivity for non-overlapping RF analyses.</p><p>Further, our lab and others (Tootell et al. 1997; Press et al. 2001; Tootell & Hadjikhani 2001; Warnking et al. 2002; Saygin & Sereno 2008; Henriksson et al. 2012) have used comparable voxel resolutions to distinguish polar angle and eccentricity representations across V1-V4 at a finer detail than our binning approach. i.e. each bin contained a range of phase preferences along both polar angle and eccentricity dimensions that can be differentiated in mapping experiments using a voxel resolution (3 – 4 mm) comparable to our current study. As such, our voxel resolution should not have prevented our ability in identifying connectivity effects along eccentricity and polar angle dimensions in the binning approach.</p><p>We appreciate the reviewer’s suggestion of asymmetrical smoothing as a control analysis. Given the reasons outlined above, we do not feel this analysis is necessary. Also, such an approach will not directly address the issue for several reasons: (1) Voxel sampling is in 3 dimensions and does not directly translate to biases across the cortical surface along polar angle and eccentricity dimensions, particularly for V2-V4 where cortical folding becomes more complex. (2) Spatial smoothing affects SNR, which could introduce another source of an asymmetrical, eccentricity-based bias. (3) Implementation is a challenge since the asymmetry of the smoothing would have to vary as a function of the cortical magnification and would also vary across subjects. However, our results do hold with the application of a general asymmetrical smoothing similar to the reviewer’s suggestion. To achieve this, we reduced the effective resolution of eccentricity by blurring data across eccentricity bins. We applied a smoothing filter such that a weighted average was computed for the timeseries of each eccentricity bin and all bins at distance of 1 (50% weighting) and 2 (25% weighting) at the iso-polar angle representations. This has the effect of blurring the data along the eccentricity dimension, thereby reducing the effective resolution along that dimension specifically. Even when applying this blurring, we observed eccentricity effects comparable with the original analyses.</p><p>As discussed above, the effects of polar angle connectivity reported in previous studies (e.g., <xref ref-type="bibr" rid="bib46">Heinzle et al. 2011</xref>; <xref ref-type="bibr" rid="bib41">Gravel et al. 2014</xref>; <xref ref-type="bibr" rid="bib70">Raemaekers et al. 2014</xref>) are attributed to overlapping RFs. To our knowledge, no study has shown angular-based connectivity patterns between regions with non-overlapping RFs. However, we acknowledge that the lack of observing such connectivity is a negative finding. As with any negative finding, this should not be overstated. Thus, we added the following text to our manuscript in paragraph 11 of Results section:</p><p>“No effect of angular distance was observed between dorsal and ventral comparisons based on actual angular distance or when reflecting across the horizontal meridian (i.e., mirror symmetry). As with any negative finding, we cannot conclude with certainty that such angular connectivity does not exist, though, to our knowledge, no study has shown angular-based connectivity patterns between regions with non-overlapping RFs.”</p><p>And paragraph 20 of Results section:</p><p>“Our data suggests little effect of polar angle connectivity beyond overlapping RFs. As with any negative finding, we can only report that we failed to see an effect of angular connectivity between regions with non-overlapping RFs, but do not conclude with certainty that it does not exist.”</p><p><italic>Still, the new results provided in <xref ref-type="fig" rid="fig12">Author response image 3</xref> highlight that there is some level of “polar angle connectivity”, which was previously only indirectly demonstrated, e.g., in the second and third panels of <xref ref-type="fig" rid="fig2"><italic>Figure 2B</italic></xref> and in <xref ref-type="fig" rid="fig8"><italic>Figure 8</italic></xref>. As the authors point out, based on the topographic model regression analyses there is evidence for “local connectivity patterns” including overlapping RF effects and correlations between homotopic regions. Beneath these overlapping RF and homotopic correlations there are also the widespread correlations, and this pattern appears to be weaker in the polar angle direction. Perhaps the authors would consider including the results of <xref ref-type="fig" rid="fig12">Author response image 3</xref> in the main manuscript, and boiling down the results of <xref ref-type="fig" rid="fig9"><italic>Figure 9</italic></xref> by reporting the overall variance explained by local connectivity versus widespread connectivity and what proportion of the widespread connectivity is along the eccentricity direction and what proportion is along the polar angle direction. While the authors do point out that these four spatial components are not orthogonal, some clarification of how much of the observed correlations can be explained by widespread patterns would strengthen the manuscript.</italic></p><p>We have added <xref ref-type="fig" rid="fig12">Author response image 3</xref> into the main manuscript as <xref ref-type="fig" rid="fig6s2">Figure 6–figure supplement 2</xref> and now discuss polar angle connectivity effects in the Results. We have moved the former <xref ref-type="fig" rid="fig6s2">Figure 6–figure supplement 2</xref> to <xref ref-type="fig" rid="fig6s3">Figure 6–figure supplement 3</xref>.</p><p>We thank the reviewer for the suggestion, and agree that some comparison between the variance explained by local and widespread models is informative. However, a ratio is not ideal for comparing variance explained by the eccentricity and polar angle models since residual polar angle correlations (in contrast to the overlapping RF and eccentricity models) were often very close to zero or negative. We believe that the paired t-test between the residual eccentricity and polar angle correlations paragraphs 17-18 of Results section serves as a better comparison.</p><p>We now report the ratio between the variance explained by the local connectivity model and the variance explained by the residual widespread connectivity model in Results section:</p><p>“Comparison of local and widespread connectivity influences</p><p>The variance explained by the overlapping RF model (after accounting for patterns attributable to the “noise” model) was generally greater than the variance explained by the model of residual widespread connectivity (after removal of overlapping RF connectivity effects) for within hemisphere comparisons (mean ratio across areas, conditions = 1.69:1.00). For between hemisphere comparisons, variance explained by the overlapping RF model (after accounting for patterns attributable to the “noise” model) was generally weaker than the variance explained by the model of residual widespread connectivity without accounting for mirror symmetrical connections (mean ratio across areas, condition = 1.00:5.65), but was generally greater when accounting for these connections (mean ratio across areas, condition = 1.27:1.00).”</p></body></sub-article></article> |