event rasters

What it is

A GUI for visualizing the detailed responses of individual neurons. Example views are shown at the bottom of this page.

How to run it

First add the spikes repository to your path, then execute "launchRasterViewer" from this repository. Using the prompts, choose the folder that contains the downloaded data, then choose a session and a probe, and the raster viewer will load.

How to use it

The viewer works best maximized on a screen with high resolution (e.g. 4k).

The controls are as follows:

• up/down arrows to switch between clusters
• left/right arrows to increase/decrease smoothing size
• h to hide/unhide behavioral icons on the rasters
• t/r to increase/decrease raster tick sizes
• c to go to a particular cluster by number
• click on anatomy at left to jump to cluster nearest there
• l to view the legend

What the plots mean

Probe layout: far left column

The vertical plot at far left is an image of the probe layout, where each dot is one site location and the size of the dot corresponds to the relative amplitude of this neuron's action potential on that site. Text labels on this plot are acronyms of brain regions.

Waveform and ACGs: second column from left

The next column, from top to bottom, contains:

• black lines: Plot of the neuron's waveform on the peak channel and surrounding channels. The snippets are about 2 ms long.
• red/blue pair of images: Another plot of the waveform, this one with x-axis time and y-axis probe depth, where color represents voltage. These are interpolated across space, and two separate interpolations are done at the left and right sides of the face of the probe. If this plot doesn't make sense, don't worry about it.
• Autocorrelogram on linear time axis up to 100 ms. The vertical dashed line is at 2 ms, and the horizontal dashed line is at the asymptotic ACG value.
• Same autocorrelogram but on a log time axis out to 1000 ms.

Rasters and PSTHs: main set of plots

In the main plots, black tick marks indicate times of spikes and colorful icons indicate behavioral events according to the legend below.

To clarify, triangles are the detected times of actual movement onsets, whereas x's are the times of the go cue, but colored by the eventual choice of the subject. But don't be confused: the go cue is the same auditory tone on every trial and does not instruct the subject in any way - coloring the go cue in this way is for convenience of visualization.

When a visual stimulus is shown on each side at the same time, the two corresponding visual icons are superimposed (the asterisk fits inside the circle).

Each raster plot is matched with the PSTH right below it - they share an x-axis. The top row of PSTHs average across all trials displayed, locked to different behavioral events. The rasters are sorted according to y-axis label (e.g. chronologically in the first and by reaction time in the second, from left).

The bottom row aligns to some of the same events but splitting the trials in different ways. In the first, with green traces, trials are split by left (ipsilateral) stimulus and the raster is sorted by this. In the second, with magenta traces, trials are split by right (contralateral) stimulus. The red/blue square plot in between these two shows the spike rate averaged across trials between 25 and 225 ms post-stimulus, for each of the 16 stimulus conditions (left contrast on the y-axis, right contrast on the x).

The third and fourth plots in the bottom row include only a subset of trials: those with equal contrast stimuli on left and right (including no stimulus on either side) and are split by choice.

The fifth and sixth are according to their y-axis labels.

The seventh, aligned to feedback, has solid lines for correct choices (positive feedback) and dashed for incorrect (negative). The trials in the raster are sorted by choice and then by feedback, so e.g. the top-most green trials are those with nogo choices and positive feedback, and match the black solid line, etc.

Example views