A multivariate CCA method is introduced for fMRI 2nd level analysis to incorporate local neighboring information, and to improve the sensitivity in group activation and group difference detection in noisy fMRI data. Statistical thresholds for significance of the group-inferences in the multivariate method are computed non-parametrically. Results from both simulated data and real episodic memory data indicate that a higher detection sensitivity for a fixed specificity can be achieved in both 2nd level activation and difference detection with the proposed method, as compared to the widely used univariate techniques.
This abstract and the presentation materials are available to members only; a login is required.