A0649
Title: A fast and powerful spatial-extent inference for testing variance components in reliability and heritability studies
Authors: Jun Young Park - University of Toronto (Canada) [presenting]
Abstract: Clusterwise inference is a popular approach in neuroimaging to increase sensitivity, but most existing methods are currently restricted to the General Linear Model (GLM) for testing mean parameters. Statistical methods for testing variance components, which are critical in neuroimaging studies that involve estimation of narrow-sense heritability or test-retest reliability, are seriously underdeveloped due to methodological and computational challenges, which would potentially lead to low power. To fill this gap, a fast and powerful test for variance components called CLEAN-V ("CLEAN" for testing "V"ariance components) is proposed. CLEAN-V models imaging data's global spatial dependence structure and compute a locally powerful variance component test statistic by data-adaptively pooling neighbourhood information. Permutations achieve correction for multiple comparisons to accurately control family-wise error rate (FWER). Through analysis of task-fMRI data from the Human Connectome Project (HCP) across five tasks and comprehensive data-driven simulations, it is shown that CLEAN-V outperforms existing methods in detecting test-retest reliability and narrow-sense heritability with significantly improved power, with the detected areas aligning with activation maps. The computational efficiency of CLEAN-V also speaks of its practical utility, and it is available as an R package.
