Thi repo contains the matlab script to generate some of the results reported in Ni et al.
Initially the Allen Mouse Brain Atlas https://www.nature.com/articles/nature05453 is used to register lables to the brain volumes and then plaques are counted according to specific regions of the atlas. The process is run separately per hemisphere. Initially the reference volume of the atlas is registered to the acquired volume in a 2-step process. Then the transformation used to obtain the registration is applied to the lables reference volume.
The registration is performed by using the tool Elastix http://elastix.isi.uu.nl/ The configuration for the registration is given into the 2 files: affine.txt and bspline.txt, which represent respectively the first and second step.
Being the regions of the atlas grouped hierarchically, we follow the hierarchy summarized into the file region_columns.xls
main_script.m should be used to call the function and pass parameters.
plaque_region_count.m is the script practically performing the threshold and counting the connected elements in specific rois. It uses adaptive threshold in 3D which is implemented in Matlab since version 2018b
saveastiff.m is a utility to save large TIF files for visual inspection (e.g. the plaque segentation). The script was originally developed by YoonOh Tak https://ch.mathworks.com/matlabcentral/fileexchange/35684-multipage-tiff-stack
left_results.csv and right_results.csv summarize the plaque count.
The used images (including the intermediate stesp) are available downsampled on Figshares at the URL: https://figshare.com/articles/arcAB_Mouse/11473707