From: 2024-04-08 11:15 to 12:00
Place: Seminar Room M 3170-73 at Dept. of Automatic Control, LTH
Contact: kristian [dot] soltesz [at] control [dot] lth [dot] se

Date & Time: April 8th, 11:15-12:00
Location: Seminar Room M 3170-73 at Dept. of Automatic Control, LTH
Speaker: Taylor Baum
Title: A Closed-Loop System for Blood Pressure Control

Abstract: Blood pressure management of patients in the operating room and intensive care unit is currently done manually. Poorly managed perioperative hypotension (i.e., low blood pressure) and hypertension (i.e., high blood pressure) can lead to severe postoperative organ dysfunction. Improvements in patient care are possible and probable with the development of a physiological closed-loop control system for blood pressure management.In this talk, I will introduce a novel closed-loop system for blood pressure control. Many such systems have been developed, but have generally not incorporated mechanistic models (i.e., models which describe how a system behaves by representing its underlying physical processes) of the cardiovascular system. Our control approach bridges this gap through integration of the two-element Windkessel model, a mechanistic model which describes arterial blood pressure in relation to arterial resistance, compliance, and blood flow from the heart.

First, I will discuss why a closed-loop system for blood pressure control is warranted and provide background information about the cardiovascular system.

Second, I will present our model that relates drug infusions to the arterial blood pressure waveform through mechanistic cardiovascular components described by the two-element Windkessel model [ACC 2021, and ACC 2023]. This provides an underlying model from our actuators (i.e., drugs) to what we want to control (i.e., arterial blood pressure) through direct cardiovascular drug targets (i.e., arterial resistance and cardiac flow).

Third, I will present our novel approach for estimation of arterial resistance, compliance, and blood flow from the heart using a single arterial blood pressure waveform [in review at IEEE TBME]. Using a swine cardiovascular data set, we validated this estimation approach. These results emphasize the feasibility of reliable in vivo estimation of these mechanistic cardiovascular components.

Fourth, I will present an optimal control strategy to regulate arterial blood pressure using our system model and our mechanistic cardiovascular component estimates [ACC 2021, and ACC 2023]. Through explicit consideration of arterial blood pressure dynamics, our control approach selects drug infusion rates to regulate the blood pressure waveform.
The seminar will conclude with a forward-looking discussion on the path towards clinical translation of this work.