Physiological closed-loop control in critical care: opportunities for innovations

Abstract

Physiological closed-loop control (PCLC) systems are a key enabler for automation and clinician support in medicine, including, but not limited to, patient monitoring, diagnosis, clinical decision making, and therapy delivery. Existing body of work has demonstrated that PCLC systems hold the promise to advance critical care as well as a wide range of other domains in medicine bearing profound implications in quality of life, quality of care, and human wellbeing. However, the state-of-the-art PCLC technology in critical care is associated with long-standing limitations related to its development and assessment, including (a) isolated and loop-by-loop PCLC design without sufficient account for multi-faceted patient physiology, (b) suboptimal choice of therapeutic endpoints, (c) concerns related to collective safety originating from multi-PCLC interferences, and (d) premature PCLC assessment methodology. Such limitations naturally motivate research to generate new knowledge and create innovative methods. In this perspective, we propose several high-reward opportunities that can accelerate the advances in PCLC systems, which may be explored by deep fusion and collaboration among multiple disciplines including physiological systems and signals analysis, control and estimation, machine learning and artificial intelligence, and wearable sensing and embedded computing technologies.