MATLAB scripts for visualization of surface-mapped scalar values. Largely inspired by visualization functions in CAT12 (http://www.neuro.uni-jena.de/cat/) but more useful in serial/parallel batch processing (i.e., no user-interaction required).
- Various FreeSurfer (http://freesurfer.net/) surface models (white, pial, inflated, semi-inflated, sphere) can be used.
- OpenGL is required for transparent overlay. Otherwise binary curvature won't be rendered.
- Multiple figures can be created in the background (MATLAB won't bother you during batch processing).
- (NEW) Quick (0.07 sec per
fsaveragehemisphere) isocontour plotting using an isocontour algorithm fromft_plot_topo3d()from FieldTrip (http://www.fieldtriptoolbox.org/) and an edge-connecting algorithm.
Download/fetch the package and add all subdirectories to MATLAB path:
addpath(genpath(DIRECTORY_WHERE_YOU_SAVED_FILES))Replace "DIRECTORY_WHERE_YOU_SAVED_FILES" with an actual path (e.g., '/Users/me/Documents/MATLAB/surfviz/').
subdir = fullfile(getenv('FREESURFER_HOME'),'subjects'); % Find FREESURFER default subject directory
surfs = fsss_read_all_FS_surfs('fsaverage',subdir); % Reading template surfaces & lots of things in the directory
fsss_view(surfs, surfs.THNS) % for example, thickness, areas, white/pial curvature, sulcal depth, annotations, ...
You can also set the surface to visualize, layout, and a threshold via cfg structure:
cfg = struct('basesurf','pial', 'layout','1x2', 'thres',3); % just an arbitrary threshold for demo
fsss_view(surfs, surfs.thns, cfg)
... and some captions and colorschemes:
thns_z = {zscore(surfs.thns{1}), zscore(surfs.thns{2})};
cfg = struct('basesurf','inflated', 'layout','2x2', 'thres',1, ...
'colorbartitle','Rel. Ctx. Thns.', 'colorbarxlabel','Z-score', ...
'colorscheme','yellowblue');
fsss_view(surfs, thns_z, cfg)
... finally you can create as many figures as you want without Matalb taking away your attention (figure's visibility will be off):
fsss_view(surfs, surfs.THNS, struct('demo.png')) % Nothing pops up but it creates a PNG file _silently_
ls('demo.png') % check the file was created!(NEW) With an edge-connecting algorithm, isocontours can be plotted very quickly (0.07 sec per 160k-vert surface). Isocurvature can be very useful for unthresholded maps to mark the boundary between gyri and sulci. Any arbitrary contours (significant clusters, manual/atals-based ROIs) can be also overlaid:
subdir = fullfile(getenv('FREESURFER_HOME'),'subjects'); % Find FREESURFER default subject directory
surfs = fsss_read_all_FS_surfs('fsaverage',subdir, struct('isocurv','inflated')); % Computes isocurvature line groups when loading
fsss_view(surfs, surfs.thns)
Or perhaps the HCP-MMP atlas (projected on fsaverage)?
fsavg = fsss_read_all_FS_surfs('fsaverage');
labels = {}; cots = {};
[~, labels{1}, cots{1}] = read_annotation('lh.HCP-MMP1.annot', 0);
[~, labels{2}, cots{2}] = read_annotation('rh.HCP-MMP1.annot', 0);
[fsavg,cots] = fsss_isoclus(fsavg, labels, struct('cots',{cots})); % this takes a while for granular annotations
fsss_view(fsavg, fsavg.thns, struct('colormap',gray, 'isocluslinewidth',1,'isocluscolors',cots)
See documentation for more information:
doc fsss_viewThe current or previous version of this repo has been used in these publications:
- Kim et al., 2024, Cerebral Cortex
- Weidacker, Kim et al., 2020, Psychological Medicine
- Kim, Overath et al., 2022, NeuroImage
- Nord, Kim et al., 2019, Neuropsychopharmacology
- Zhang et al., 2019, Brain Stimulation
(cc) Seung-Goo Kim, 2019-2024.