AI Research Catalyst Fund Awardees Virtual Seminar Series

“Combining Machine Learning and Large-Scale Brain Data to Predict Human Cognition and Behavior“

by Dr. Brian Odegaard
Assistant Professor in the Department of Psychology

Wednesday, December 1, 2021

ABSTRACT:

Over the last decade, machine learning algorithms have exhibited remarkable capacities to decode information from complex sources. These successes have been driven (at least in part) by access to large data sets: the more data one uses to train the model, the more accurate and robust the outputs. Specific machine learning applications, such as kernel ridge regression (KRR) and deep neural nets (DNN), have shown strong promise for enhancing predictions of cognitive and behavioral phenotypes in both mechanistic basic science and applied clinical work. Here, we aim to train KRR/DNN models on openly accessible, large-scale structural and functional magnetic resonance imaging (fMRI) data to improve prediction of cognitive and behavioral phenotypes in new smaller-scale samples collected in our labs at UF. Combining state-of-the-art brain decoding with a “big data” computational modeling approach, this project will (i) allow development and make publicly available a new research algorithm/infrastructure for machine learning on HiPerGator; (ii) provide basic scientific insights into how brain structure and function are linked to different phenotypes; and (iii) identify brain regions of interest for future targeted interventions, such as using neurofeedback training to enhance cognition and behavior.

Bio: 

Brian Odegaard is an Assistant Professor in the Department of Psychology at the University of Florida, and PI of the Perception, Attention, and Consciousness Laboratory at UF. His laboratory employs an interdisciplinary approach to study high-level sensory processes; specifically, the neural and computational basis of multisensory integration, perceptual metacognition, and attention, with an emphasis on how these domains inform current theories of visual awareness. Current research projects in the laboratory include work funded by the Templeton World Charity Foundation to study visual change blindness using high-resolution eyetracking, using online experimental platforms to study visual metacognition, and applications of deep learning to better understand the links between the brain, behavior, and consciousness.  His PhD and postdoctoral work were conducted at UCLA, and he recently held an appointment as a cooperating researcher at the Advanced Telecommunications Research Institute International near Osaka, Japan. This past year, he incorporated innovative paradigms from his work in Japan to conduct decoded neurofeedback at the McKnight Brain Institute AMRIS fMRI facility, as his research group implemented neuromodulatory techniques to try to reveal causal links between the brain and behavior. He is an active advocate for the Open Science movement, and looks forward to sharing the resources his research group have developed as part of this UF AI Catalyst Research Fund Award.

Wednesday, December 1, 2021
12:00PM – 1:00PM
Via Zoom

RSVP below. Instructions to attend talk via Zoom will be emailed to you.