Life Sign Decision Support Algorithms
GCAS is currently developing computational algorithms and the agent-based software for the Life-Sign Decision Support Algorithms (LSDSAs) for the Warfighter Physiological Status Monitoring system under U.S. Army Medical Research and Materiel Command contract. LSDSAs interpret data from a suite of wearable physiological sensors to infer a soldier’s current clinical state on the battlefield. LSDSAs use Bayesian Networks to model the human physiology. Other BNs are devoted to modeling the sensors, their probability and mode of failures, and the probability of their dislodgment. Further BN models take into account the probability and mode of failures of the data transmission system. These networks are integrated together and complemented with algorithms such as Sensitivity Analysis and detection of conflicting data via Conflict Measure. Thanks to this architecture, the LSDSAs system is capable of identifying a soldier’s physiological status with a high level of reliability by taking into account clinical uncertainty, data uncertainty and reliability of the integrated circuit of sensors and devices. Computational algorithms have also been devised so as to automatically evaluate the reliability of various sensor system configurations. These algorithms possess the flexibility of being readily applicable to different sensor and circuit configurations and clinical analysis.