Spatiotemporal fMRI data processing using generalized canonical correlation analysis

Common fMRI data processing techniques usually minimize a temporal cost function or fit a temporal model to extract an activity map. In our previous work we used generalized canonical correlation analysis to extract a highly, spatially reproducible statistical parametric map (SPM) from fMRI data using a cost function that does not depend on a model of the subjects' temporal response. Here we focus on using a cost function that simultaneously maximizes temporal and spatial reproducibility of MRI statistical parametric map. Based on a modified version of generalized canonical correlation analysis (gCCA) we propose a method to extract a highly reproducible map by maximizing the sum of the pair-wise correlations between pairs of maps while the associated temporal response to the extracted map follows the subjects' temporal response. The proposed method is applied to BOLD fMRI datasets without any spatial smoothing from 10 subjects performing a simple reaction time (RT) task. Using the NPAIRS split-half resampling framework with a reproducibility measure based on SPM correlations we compare the proposed approach with our previous work presented in. Our results show that the proposed modified gCCA is an efficient approach for extracting both default mode network and task mode network.