Computational Learning Methods for Neuroimaging Data Analysis

• Activity Number: 150 - Recent Advances for Modeling Neuroimaging Data
• Type: Topic Contributed
• Date/Time: Monday, July 31, 2017 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Imaging
• Author(s): Don Hong* and Xin Yang and Jingsai Liang
• Companies: Middle Tennessee State Univ and Southern Arkansas University and Middle Tennessee State University
• Keywords: neuroimaging ; machine learning ; computational statistics ; fMRI data analysis ; brain science ; spatial data

Abstract

The studies on brain science including many neuropsychiatric disorders using fMRI technology usually incorporate data spatial information and investigate functional connectivity of the brains at the network level. However, the identification of region of interests for disease and differential expressions in connectivity networks remain challenging on theory, computation, and statistical inferences. In this talk, we'd like to report some recent progress on fMRI data analysis using computational learning schemes including multi-task learning, spatial regularization neural network, and non-Gaussian penalized PARAFAC analysis for fMRI data processing with applications in autism and Alzheimer disease (AD) studies.