In previous TRISH and NASA funded efforts, Nahlia has developed nested logic, Bayesian algorithms and software, to dynamically analyze and deliver evidence based clinical decision support to control evolving medical situations. This Autonomous Medical Response Agent (AMRA) was further advanced in phase I SBIR based on military Prolonged Field Care principles for autonomous field medical care. AMRA provides coordinated mission guidance for multiple caregivers, a clinical case simulator to systematically and verifiably improve AMRA’s decision structure, and an integrated clinical case-based training feature to maintain optimal human-machine performance for autonomous medical operations.
Phase II work focuses on expanding AMRA’s probabilistic structure to include parallel feedback control, Bayesian nested hidden state networks capable of providing resource constrained clinical decision support to optimize the health state of Astronauts. Learning and path optimization algorithms in AMRA will extend longevity and usability of AMRA. The efficacy and validity of AMRA will be demonstrated with an experienced NASA Flight Surgeon and non-clinician astronaut-like users at the Naval Post Graduate School.
Autonomous Medical Response Agent addresses multiple NASA HRP Gaps: Medical-101,201, 301, 401,601,701
Remote field applications to assist flight surgeons on HISEAS, or McMurdo Station
Autonomous medical response system has the potential to aid astronauts on long duration missions to the Moon and Mars.
DoD: Prolonged field care practitioners, Undersea Submarine Medical Caregivers, Air Force pararescue
Civilian: aeromedical evacuation, rural/indigenous peoples care, international disaster relief, prison health systems
Commercial: Assist nurse practitioners and physician assistants, health insurance to predict costs of care