“Sparsity Meets Dynamics: Signal Processing for High-dimensional Dynamic Neural Data”

Location: LTS Auditorium, 8080 Greenmead Drive

Speaker:
Behtash Babadi
UMD Department of Electrical and Computer Engineering

Abstract:
Thanks to advances in data acquisition technology, the process of collecting data from the nervous systems of animals and humans has been substantially facilitated, resulting in abundant pools of data under various modalities and conditions.

This data is typically high-dimensional, dynamic and complex, as it may span hours of brain activity. Therefore, modern-day neural data requires carefully-tailored signal processing techniques in order to be effective in deciphering the mysteries of the brain.

Converging lines of evidence in theoretical and experimental neuroscience suggests brain activity is a high-dimensional distributed spatiotemporal process emerging from underlying dynamic sparse structures.

In this talk, we will explore how in various scenarios of interest, the hypotheses of dynamicity and sparsity can be exploited in a principled way to address several outstanding challenges faced by existing neural signal processing techniques.

In particular, we will present ongoing research in MEG source localization, spectrotemporal analysis of EEG and spiking data, dynamic decoding of auditory attentional modulation from MEG, and adaptive neuronal system identification.

Bio:
Behtash Babadi is an assistant professor in the Department of Electrical and Computer Engineering at the University of Maryland.

His research interests include statistical and adaptive signal processing, neural signal processing, and systems neuroscience.

Babadi is a postdoctoral fellow at the Department of Brain and Cognitive Sciences at Massachusetts Institute of Technology, as well as the Department of Anesthesia, Critical Care and Pain Medicine at Massachusetts General Hospital.

He received his doctorate in engineering sciences from Harvard University in 2011.