Meeting Banner
Abstract #3362

Predicting Treatment in Patients with Major Depression Using Granger-Based Connectivity and Support Vector Machines

Gopikrishna Deshpande1, George Andrew James1, Richard Cameron Craddock2, Helen S. Mayberg3, Xiaoping P. Hu1

1Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Atlanta, GA, USA; 2School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA, USA; 3Department of Psychiatry, Emory University, Atlanta, GA, USA

Variations in effective connectivity within limbic-cortical network (as modeled with structural equation modeling from PET) have been identified in different major depressive disorder (MDD) subgroups associated with differential response to antidepressants. In this study, MDD patients were randomly assigned to one of two possible treatments cognitive behavioral therapy or a drug and treated for 12 weeks. Granger-based effective connectivity was calculated from fMRI data obtained before and during treatment. Recursive cluster elimination with support vector machines was used to predict which treatment the patients were receiving based on discriminative connectivity features. Our results show 100% accuracy in predicting treatment.