Application of Mass Multivariate Analysis on Neuroimaging Data Sets for Precision Diagnosis of Depression: Data and code

In the archive are the datasets and code used to create the results in article "Application of Mass Multivariate Analysis on Neuroimaging Data Sets fo Precision Diagnosis of Depression". The methods is based on the Multivariate Linear toolbox mad F. Kherif’s lab and available on GitHub. https://github.com/LREN-CHUV/MLM.   

A total of 44 patients with a current psychotic (n=19) or depressive (n=25) episode were analyzed by MLM (details can be found in the article itself) utilizing resting sate fMRI, tasks fMRI, and anatomical T1w images. 

The data results are provided in MATLAB format, in a matrix called MLM.mat. The matrix includes the canonical variables for each of the three imaging modalities, chi-square statistics, and p-values for each brain region.  Additionally, we provided a function MLM_plot_fun that allows mapping the statistics results to a canonical three-dimensional brain as per the article.