Multivariate Pattern Recognition for Neuroimaging Toolbox

Multivariate Pattern Recognition for Neuroimaging Toolbox (mPReNT) is a MATLAB toolbox 
based on the multivariate pattern recognition techniques for discriminative analysis of neuroimaging data. 
To identify informative patterns that distinguish subjects of different groups, 
we build a series of support vector machine (SVM) classification models in conjunction with a (simplified) forward component selection technique 
within a nested cross-validation procedure. 

In mPReNT, brain networks/patterns(independent components [ICs]) at an individual subject level are used as bases 
for a linear subspace on Grassmann manifold to facilitates a comprehensive characterization of neuroimaging data, 
and SVM classification models can be used to discriminate between groups of subjects with classification scores.

In mPReNT, crossvalidation_forward_svm_performance_with_component_main.m is the main function. 
Users should set the paths and parameters firstly. If the number of components as networks/patterns is too large, 
the simplified forward component selection technique should be used to reduce computational cost.

For computing ICs, a group information guided ICA method could be used (GIGICAR.m).

Written by Yong Fan (yong.fan@ieee.org), Rixing Jing (rixing.jing@nlpr.ia.ac.cn).

References:
Du, Y., & Fan, Y. (2013). Group information guided ICA for fMRI data analysis. 
Neuroimage, 69, 157-197, doi:10.1016/j.neuroimage.2012.11.008.

Fan, Y., Liu, Y., Jiang, T., Liu, Z., Hao, Y., & Liu, H. (2010). 
Discriminant analysis of resting-state functional connectivity patterns on the Grassmann manifold. 
7623, 76231J, doi:10.1117/12.844495.

Fan, Y., Liu, Y., Wu, H., Hao, Y., Liu, H., Liu, Z., et al. (2011). 
Discriminant analysis of functional connectivity patterns on Grassmann manifold. 
Neuroimage, 56(4), 2058-2067, doi:10.1016/j.neuroimage.2011.03.051.