Skip to content
We updated gift to include the following:
- Algorithm name GIG-ICA is changed to MOO-ICAR to distinguish between the spatially constrained analysis and the GIG-ICA back-reconstruction approach.
- MOO-ICAR algorithm estimates the individual subject component maps and timecourses given spatial references or templates.
- GIG-ICA estimates individual subject component maps and timecoures given ICA aggregate maps as reference in the back-reconstruction.
Check it out here
- Two more cluster evaluation criteria like Daviesbouldin and Ray-Turi are added in the temporal dFNC toolbox.
- Subject ICA loadings (*ica_subject_loadings.nii) are written out to the disk when algorithms other than the constrained ica approaches are used.
- Nan version of mT function is fixed to include multiple contrasts when more than 2 levels is used in the categorical covariates.
- Default display template is changed to ch2bet_3x3x3.nii.
- SPM MEX binaries are updated.
- Kurtosis graphs are fixed to include y-limits separately for timecourses and spatial maps.