Probabilistic Unit Life Status Estimation ( PULSE )

"The Warfighter Physiological Status Monitoring (WPSM) system must deal with artifact-laden, uncertain, and dynamic biomedical sensor data. We propose to address these critical issues directly with the development of the Probabilistic Unit Life StatusEstimation (PULSE) suite of advanced Bayesian Belief Network (BBN) algorithms. BBNs provide probabilistic reasoning for dealing with uncertainty, and flexible modeling of real world problems. However, static BBNs do not explicitly provide temporalreasoning, which is necessary due to the dynamic nature of the WPSM sensor array data. We plan to utilize a formal extension to BBNs known as Temporal Belief Networks (TBNs), which adds the required temporal reasoning. TBNs combine the key features offlexible modeling, probabilistic reasoning, and temporal reasoning needed for the PULSE algorithms. We will use TBNs primarily to model the subject's clinical state, but also to model the sensor array hardware in parallel, obtaining metrics of sensorreliability which will improve the clinical state estimates. Finally, we will use the powerful sensitivity analysis feature of BBNs to isolate faults in the modeled sensor array, enhancing operational robustness. We plan to develop a prototype PULSEalgorithm suite, and evaluate overall concept feasibility in a simulated physiological monitoring test, incorporating realistic sensor models. The Probabilistic Unit Life Status Estimation (PULSE) algorithm sui