Taylor Elise Baum
Massachusetts Institute of Technology
tbaum@mit.edu
Bio
Taylor Elise Baum is an Electrical Engineering and Computer Science PhD Candidate at the Massachusetts Institute of Technology (MIT), advised by Profs. Emery Brown, Munther Dahleh and Thomas Heldt. She is interested in improving how we understand and control physiological systems (e.g., the cardiovascular system and nervous system) in critical care. Her work has advanced systems that automate cardiovascular management, including arterial blood pressure regulation and treatment of serious cardiac arrhythmias. Taylor’s research has recently been recognized with a featured article in IEEE Transactions on Biomedical Engineering (2024), an invited article at the American Control Conference (2025), the Hugh Hampton Young Memorial Fund Fellowship (2024) and selection as an MIT Graduate Woman of Excellence (2023). She also founded Sprouting Education, which develops bilingual curricula and has hosted 30+ events across South America and the U.S., earning her MIT’s Seth J. Teller Award for Excellence, Inclusion and Diversity (2023).
Areas of Research
- Signal Processing
An Adaptive Closed-Loop System for Arterial Blood Pressure Control with Phenylephrine and Clevidipine
Poor management of arterial blood pressure (ABP) has been linked with increased morbidity and mortality. Although closed-loop systems for ABP control have been shown to improve management over standard manual titration of intravenous medications in the operating room (OR), their adoption has thus far been prevented by safety concerns and their inconsistent performance across patients. We present a novel closed-loop ABP control system consisting of a set of two adaptive model predictive controllers, one which brings ABP up with phenylephrine (PE) and one which brings ABP down with clevidipine (CL). Each controller is patient specific without requiring prior patient data and explicitly incorporates safety constraints such as limits on the rate of change of ABP. We find that, in non-emergent, post-transient periods, our PE-actuated controller maintains mean ABP within 5 mmHg of the target 99.86% of the time (95% CI: 99.69, 100.00) and has 1.09 mmHg (95% CI: 0.87, 1.32) root mean square error (RMSE) and our CL-controller, 97.13% (95% CI: 94.31, 99.04) with 1.64 mmHg (95% CI: 1.29, 2.03) RMSE. We additionally find that both controllers rapidly reject severe disturbances and that we can switch between the controllers mid-session. These results establish the attainable precision of ABP control and provide further support for clinical adoption of such closed-loop systems. As no FDA-approved closed-loop ABP control system is in use in the United States, our system has immense potential to revolutionize cardiovascular care in hundreds of millions of OR procedures and intensive care unit stays every year.