Skip to content
Switch branches/tags

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time


JHU-MNI-ss (Eve) Template

Creator: Jean-Philippe Fortin,

Authors and Maintainers: Jean-Philippe Fortin, John Muschelli

Software status
Resource: Travis CI
Platform: Linux
R CMD check Build status

If using the EveTemplate package, please cite the following:

Description Citation Link
JHU-MNI-ss Atlas Kenichi Oishi, Andreia V Faria and Susumu Mori, JHU-MNI-ss Atlas, 2010, Johns Hopkins University School of Medicine, Department of Radiology, Center for Brain Imaging Science Link
WMPM White Matter Parcellation Map Oishi et al., Atlas-based whole brain white matter analysis using large deformation diffeomorphic metric mapping: Application to normal elderly and Alzheimer’s disease participants, Neuroimage, 2009 Link

Complete BibTeX citations can be found here.

Table of content

1. Introduction

The JHU-MNI-ss atlas, which is often called "Eve Atlas", is based on a single-subject data as described in Oishi et al, 2009. There are co-registered T1 (MPRAGE), T2, and DTI images as well as white matter parcellation map (WMPM). Once the image of interest is normalized to this atlas coordinate, the WMPM (which also includes gray matter structures) can be superimposed for anatomical definition (e.g. which structure is affected by a lesion or where exactly is the fMRI activation site) or automated segmentation.

The EveTemplate package contains the anatomical T1 and T1 images, together with or without the skull on, as well as a brain mask. It also contains the three types of White Matter Parcellaton Map (WMPM) with the dictionary of the labels. Finally, we have performed a 3-tissue class segmentaiton with the FSL FAST algorithm and have included the segmentation classes in the package. For the MNI152 template, see the similar package MNITemplate.

2. Reading the data into R

We first load the package into R:


Once the package is loaded into R, use the command readEve() to import the Eve template T1-w image as a nifti object into R:

eve_t1 <- readEve()

One can use the function orthographic from the oro.nifti package to visualize the template:


For the T2 image, T1 Brain only and Brain mask respectively, use the following:

eve_t2 <- readEve("T2")
eve_t1_brain <- readEve("Brain")
eve_brain_mask <- readEve("Brain_Mask")

To obtain the T2 image with the brain only, one can use the convolution of the T1 and brain mask as follows since the scans provided are already co-registered:

eve_t2_brain <- eve_t2*eve_brain_mask

In many preprocessing pipelines, the path of the template file in the system must be specified. For this, use the following:

eve_path <- getEvePath()

and similarly for the other images:

eve_t2_path <- getEvePath("T2")
eve_t1_brain_path <- getEvePath("Brain")
eve_brain_mask_path <- getEvePath("Brain_Mask")

3. Segmentation

We performed a 3-tissue class segmentation of the T1w Eve template using the FSL FAST segmentation algorithm via the fslr package. The script that was used to perform the segmentation can be found here. The segmentation labels are 0 for Background (outside of the brain), 1 for cerebrospinal fluid (CSF), 2 for grey matter (GM) and 3 for white matter (WM). Let's read the segmentation classes into R:

seg <- readEveSeg()

If one wishes to create a WM mask, could do the following:

wm_mask <- seg
wm_mask[wm_mask!=3] <- 0

and similarly for the other tissues.

4. White Matter Parcellation Map (WMPM)

The Eve template comes with an atlas of the different anatomical structures of the brain (not only the White Matter structures, despite the name White Matter Parcellation Map). There are 3 different types of WMPM available.

For instance, to obtain the WMPM type I, the following command should be used:

map <- readEveMap(type="I")

The label dictionary can be loaded as a data frame into R using

labels <- getEveMapLabels(type="I")

Curation of the labels

The script createLabels.R was used to clean the labels of the Eve atlas anatomical structures; there were some inconsistencies in the labels. The command getEveMapLabels() automatically loads the curated labels.

5. Summary of the files and functions

File Description Reader
JHU_MNI_SS_TI.nii.gz T1-w Eve Template readEve("T1")
JHU_MNI_SS_T2.nii.gz T2-w Eve Template readEve("T2")
JHU_MNI_SS_Brain.nii.gz T1-w Eve Template, skull stripped readEve("Brain")
JHU_MNI_SS_Brain_Mask.nii.gz T1-w Eve Template, brain mask readEve("Brain_Mask")
Tissue Segmentation:
JHU_MNI_SS_Brain_FAST_seg.nii.gz FSL FAST tissue classes (1=CSF, 2=GM, 3=WM) readEveSeg()
White Matter Parcellation Map:
JHU_MNI_SS_WMPM_Type-I.nii.gz White Matter Parcellation Map (Type I) readEveMap("I")
JHU_MNI_SS_WMPM_Type-II.nii.gz White Matter Parcellation Map (Type II) readEveMap("II")
JHU_MNI_SS_WMPM_Type-III.nii.gz White Matter Parcellation Map (Type III) readEveMap("III")
eve_map_labels.rda White Matter Parcellation Map labels (I) getEveMapLabels("I")
eve_map_labels.rda White Matter Parcellation Map labels (II) getEveMapLabels("II")
eve_map_labels.rda White Matter Parcellation Map labels (III) getEveMapLabels("III")

6. Miscellaneous

The GitHub repository from John Muschelli includes a useful discussion on how the Eve template compares to the MNI152 template and to the Rorden T1 image. The dimensions of the 3 templates do not agree, and different solutions are proposed to match the voxel locations, either by dropping slides, interpolating or performing a non-linear transformation. The files for the transformed templates are included in the repository.


JHU-MNI-ss (Eve) Template



No releases published


No packages published