BME PhD Thesis Proposal  

Date: 12/16/2020 

Time: 3:00 pm 

BlueJeans link: https://gatech.bluejeans.com/151676567 

Meeting ID: 151676567 

Advisor(s): 

Dr. Babak Mahmoudi 

Dr. Robert E. Gross 

Committee Members:  

Dr. Eva L. Dyer 

Dr. Svjetlana Miocinovic 

Dr. Helen S. Mayberg (Icahn School of Medicine at Mount Sinai) 

Dr. Jeffrey A. Herron (University of Washington) 

 Title: optimal design of experiments for developing closed-loop neuromodulation systems 

 Abstract: 

Open-loop deep brain stimulation (DBS) is a neurosurgical treatment that modulates the brain's neural functioning by delivering an electrical signal using predefined stimulation parameters to a specific deep anatomical structure of the central nervous system. The new generation of DBS therapy, called closed-loop DBS, would reduce the side effects and increase DBS therapy's efficacy by modulating the brain structure using optimized stimulation parameters. The current approach for finding the optimized stimulation parameters based on the grid-search is time-consuming, expensive, and even impossible in particular when the number of parameters scales up.  Active learning is a smart solution for designing an experiment in which human decision-making is less than optimal for the task. In more detail, active learning is a paradigm in which machine learning models can direct the learning process by providing dynamic suggestions/queries for the “next-best experiment.”  

This project is aimed at developing a framework that leverages interpretable machine learning techniques for characterizing the neurophysiological effects of DBS (Aim1) and active learning techniques for the optimal design of closed-loop DBS control systems (Aim2). We would implement and validate the proposed framework in a translational experimental setup (Aim3).