Title: Machine Learning in Neuroimaging 

Abstract: Neuroimaging, specifically magnetic resonance imaging (MRI), allows us to non-invasively investigate brain function both in health and disease. In principle, the application of machine learning to neuroimaging is attractive: it would potentially allow decoding of the mind from the data as well as identify features that are predictive of various psychiatric and neurologic disorders. However, the curse of dimensionality, i.e. larger number of features/descriptors as compared to number of examples/subjects, poses the biggest hurdle in realizing the potential of machine learning in neuroimaging. Here, I will discuss these challenges and show approaches that may provide us with realistic estimates of robust classifier performance in neuroimaging data.

Speaker: 

Gopikrishna Deshpande, Ph.D
Professor

AU MRI Research Center, Department of Electrical and Computer Engineering, Auburn University

Department of Psychology, Auburn University

Center for Health Ecology and Equity Research, Auburn University
Alabama Advanced Imaging Consortium, Auburn University and University of Alabama Birmingham