Enabling “White-Box” Autonomy in Medical Cyber-Physical Systems

In pursuit of the long-term goal of enabling interpretable “white-box” autonomy in healthcare, the goal of this project is to investigate physiological modeling, coordinated and resilient multivariable closed-loop control, and regulatory science methodologies for white-box autonomy, as well as to illustrate its application to an important critical care scenario of circulatory resuscitation.  The idea of introducing autonomy to the healthcare domain is not new.  However, prior autonomy capabilities have not been suitably mature for real-world critical care, due to the limitations associated with the context awareness and interpretability of autonomy capabilities, coordination of multiple therapeutic control loops with resilience, and regulatory science methods for evaluation and approval of safety-critical physiological closed-loop control systems.  In this project, an integrated research, education, and outreach effort is proposed to develop and demonstrate generalizable methodologies for interpretable white-box autonomy that address these challenges.  On the research front, the project intends to develop novel generalizable methodologies for (1) patient physiology and pharmacology modeling and (2) coordinated and resilient multivariable closed-loop control, and demonstrate the value of these methodologies by implementing interpretable white-box physiological closed-loop control algorithms for circulatory resuscitation.  On the education and outreach front, this project intends to create next-generation M-CPS workforce as well as to advance regulatory science for evaluating critical care autonomy capabilities, by (1) attracting graduate, undergraduate, and K-12 students into STEM and M-CPS, and (2) collaborating closely with FDA to investigate testing methodologies and tools for physiological models and closed-loop M-CPS.  The advances in fundamental knowledge related to physiological modeling, closed-loop control, and regulatory science will close outstanding gaps and make scientific breakthroughs in M-CPS autonomy, which will ultimately improve the quality of life of human beings.  In addition, the autonomy testing methodologies and tools for M-CPS will facilitate the development, evaluation, and approval of emerging M-CPS autonomy capabilities as well as M-CPS research and education.