An Adaptable Framework to Extract Abnormal Brain Networks
assistant professor in the Department of Electrical and Computer Engineering at Johns Hopkins University
There is increasing evidence that complex neurological disorders reflect distributed impairments across multiple brain systems. These findings underscore the importance of network-based approaches for functional data. However, network analyses in clinical neuroimaging is largely limited to aggregate measures, which do not pinpoint a concrete etiological mechanism. In contrast, I will present a novel Bayesian framework that captures the underlying topology of the altered functional pathways.
In the first part of this talk, I will introduce our core framework to extract abnormal network foci from functional MRI data. This model relies on a latent structure, which captures hidden interactions within the brain; the latent variables are complemented by an intuitive likelihood model for the observed neuroimaging measures. The resulting variational EM algorithm produces clinically meaningful results by simultaneously localizing the centers of abnormal activity and the network of altered connectivity. Next, I will address three technical challenges: flexible network topology, multimodal integration and patient-specific analysis. I will demonstrate that our core framework can elegantly be adapted to each of these scenarios and yields novel insights into autism, schizophrenia and epilepsy, respectively. Finally, I will highlight some exciting future directions for our methodology that revolve around clinical understanding and interventions.
Clark Hall 110 (Homewood Campus)
Video-teleconferenced to Traylor 709 (Medicine Campus)
Lunch will be provided at 12:00 PM (Homewood location only)
Information on ICM seminars and links to previous talks can be found at http://www.icm.jhu.edu/calendar