Meeting Banner
Abstract #3813

Nonlinear kernel canonical correlation analysis (kCCA) in fMRI

Zhengshi Yang1, Xiaowei Zhuang1, Tim Curran2, and Dietmar Cordes1,3

1Cleveland Clinic Lou Ruvo Center for Brain Health, Las vegas, NV, United States, 2Department of Psychology and Neuroscience, University of Colorado Boulder, Boulder, CO, United States, 3Department of Radiology, University of Colorado-Denver, Denver, CO, United States

Kernel representation is an efficient method to extract nonlinear features without significantly increased computational complexity. Linear kernel CCA has been applied to fMRI data but the performance of nonlinear kernel CCA is still not clear. Here we investigate the accuracy of five types of kernel on simulated fMRI data and then apply Gaussian kernel CCA on real fMRI data. It provides a more sensitive and specific way to detect activation pattern.

This abstract and the presentation materials are available to members only; a login is required.

Join Here