A Statistical Method for Predicting Clinical Outcomes Using Resting-State fMRI

• Activity Number: 583
• Type: Topic Contributed
• Date/Time: Wednesday, August 7, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Imaging
• Author(s): Ying Guo*+ and Tian Dai
• Companies: Emory University and Emory University
• Keywords: neuroimaging ; resting-state fMRI ; prediction methods

Abstract

Brain imaging data have shown great promise as a useful predictor for psychiatric conditions, cognitive functions and many other neural-related outcomes. Previous research on image-based prediction usually focused on developing predictive models based on task-related brain imaging data that were acquired when subjects were exposed to external experimental stimuli. Recently, the study of resting state brain activity patterns has become an important area of neuroimaging research. The functional magnetic resonance (fMRI) blood oxygenation level-dependent (BOLD) signal during resting state represents baseline brain activity in the absence of task-related neuronal action and external stimuli, hence reflecting intrinsic patterns of neural processing in the brain. In this talk, we present a statistical method for predicting clinical and psychiatric outcomes using resting-state fMRI images. We illustrate the proposed method using real fMRI data examples.