Basic principles and concept design of a real-time clinical decision support system for autonomous medical care on missions to Mars based on adaptive deep learning

Space agencies and private companies prepare the beginning of the human space exploration for the 2030s with missions to put the first human on the Mars surface. The absence of gravity and radiation, along with distance, isolation and hostile environment are expected to increase medical events with unidentified manifestations along the crewmembers. The current healthcare strategy based on telemedicine and the possibility to stabilise and transport the injured crewmember to a terrestrial definitive medical facility is not applicable in exploration class missions. Therefore, full autonomous capability to solve medical situations will guide design of future healthcare systems on-board. This study presents ten basic principles and the concept design of MEDEA, an on-board clinical decision support system to help crewmembers to deal with medical conditions, with special attention to emergency care situations and critical monitoring. Therefore, MEDEA is conceptually designed as a software suite of four interconnected modules. The main of them is responsible to give direct advice to the crew by means of a deep learning multitask neural network to predict the characters of the medical event, a classifier of the tertiary medical intervention and an optimiser of medical action plans. This module is continuously evaluate and re-trained with changing physiological data from the crew by an adaptive deep learning module, ensuring fairness, interpretability and traceability of decision making during the full operational time of MEDEA. Finally, MEDEA would be semantically interoperable with health information systems on-board by a FHIR module. The deployment of MEDEA on-board of future missions to Mars will facilitate the deployment of a comprehensive preventive medical strategy, future quantitative medicine on Earth and on the expansion of humans throughout the solar system.