MEDULLA: Monitoring Emergent Duress Under Long-Lasting Allostasis

Allostatic Load is a result of chronic stress that accumulates incrementally over an extended period of time, often without the awareness of the individual(s). MEDULLA (Monitoring Emergent Duress of Unified Long-Lasting Allostasis) is a collective allostatic load (CAL) estimator validated within a NASA study (HI-SEAS) which SIFT is currently participating in. MEDULLA has several advantages over current approaches: 1) A high-fidelity, isolated, confined environment to prepare for human missions to Mars is ideally suited to study continuous stress; 2) MEDULLA integrates multiple non-intrusive data streams: physiological, behavioral and cognitive/linguistic; and 3) Collaborative VR games provide an opportunity to study social behavior. We propose to leverage SIFT's cutting-edge research using non-obtrusive metrics, such as linguistic analysis and heart rate variability, to estimate and predict cognitive workload. We will extend this work and our expertise in using VR and games to study human behavior and integrate those methods and metrics to the estimation of allostatic load. MEDULLA will utilize machine learning methods such as recurrent neural networks to enable it to adjust its classification parameters to novel teams and situations. Providing a non-intrusive, scalable and distributable solution to study and diagnose team behaviors within a naturalistic setting, is a clear advantage.