explore DCM

This is a single MATLAB function (download it here), which creates a GUI allowing to explore the results of a DCM analysis.

In addition to standard results display, this tool also extracts statistical quantities that serve as diagnostic tools for the DCM probabilistic inversion, such as the parameters posterior correlation matrix (indicative of identifiability issues) and the residuals autocorrelation (indicative of systematic underfit). Region-specific percentage of variance exlpained are also given.

The GUI is based on tabs, which makes it very easy to eye-ball, e.g., brain connectivity parameter estimates and significance scores (using Savage-Dickey ratios), observed versus fitted data in each region of interest, first-order Volterra kernels, etc ...