Multivariate Neuroimaging Analysis: An Approach for Multiple Image Modalities

This study presents a multivariate technique for analyzing variable significance in neuroimages. Traditional univariate analysis presents the difficulty of controlling for the familywise error rate, and multivariate analysis across the entire brain of one subject presents computational difficulties. This study proposes a multivariate technique with different imaging modalities as the outcomes. This hybrid approach takes advantage of the possible correlations between imaging modalities. We demonstrate the analysis for a single voxel in the brain and compare image maps from the univariate and multivariate techniques for VBM and fMRI imaging modalities. Our results show that there is evidence that the multivariate technique is a viable method for the researchers toolkit. EUROIMAGING has become a very important technology in the medical field. It has helped doctors diagnose brain diseases, pinpoint structural damage, and identify disease related changes. Neuroimaging has also enhanced the abilities of medical researchers. Medical researchers are now able to examine the brain more closely in hopes of detecting brain diseases earlier. By conducting experiments and analyzing results, researchers hope to identify differences in brain structure and function associated with diseases earlier in peoples' lives. If they are able to do this, it may be possible to treat these diseases earlier with the ultimate goal of prevention.