Multilevel Independent Component Analysis to Identify Functional Networks During Multiple Cognitive Tasks

• Activity Number: 481
• Type: Topic Contributed
• Date/Time: Wednesday, August 12, 2015 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Imaging
• Author(s): Seonjoo Lee*
• Companies: Columbia University
• Keywords: independent component analysis ; multilevel principal component analysis ; fMRI ; NeuroImaging

Abstract

Independent component analysis (ICA) is an effective data-driven method for blind source separation. It has been successfully applied to identify function networks in neuroimaging. In cognitive neuroscience, cognitive abilities are optimally measured via multiple cognitive tasks. When we collect fMRI across multiple cognitive tasks, independent functional networks derived separately across different tasks and combined in the further steps. Although cognitive tasks are designed to demand certain cognition, their expression in functional networks may be subtle, and aggregating ICs across different tasks and applying dimension reduction for group analysis may not capture such subtle signals. Thus, we introduce multi-level independent component analysis to take advantage of repeated measures and account for within subject variations. A set of simulation studies and fMRI data analysis results will be presented.