Development and Validation of a Multimodal Performance Evaluation System (MPES) for Real-time Assessment of and Modifications to Collaborative Exercises

This project seeks to develop a method for delivering dynamically changing medical scenarios to enhance learner performance in stressful, complex, and changing battlefield environments. It will accomplish this by developing a model of human resiliency based on real-time biometric input. Biometric data (heart rate, galvanic skin response, socio-interaction) from learners will drive this model to derive an objective measure of current performance. In turn, this information will be used to alter the dynamic elements in the scenario (e.g., stressors, complexity of injuries, availability of resources). In this process, we will derive a method of training based not on static progression, but tailored toward individual traits and abilities. The model will be application and technology agnostic. The underlying methodology will work equally well on different curriculum (e.g., Tactical Combat Casualty Care) and gaming platforms. Development of the model will be done at the Val G. Hemming Simulation Center (VGHSC) of the Uniformed Services University of the Health Sciences.

The research question we aim to answer is what combination of data types, usable in virtual, head-mount, and live team events, provide the most reliable prediction of a trainee's state (cognitive and affective). Our hypotheses are (1) a trainee's personality traits can be a predictor of susceptibility to stress; (2) cognitive overload, triggered by scenario-related conditions, can increase stress; (3) non-invasive wearable biometrics as well as event, timing, and sequence data will enable sufficient triangulation of data to reliability predict stress; and (4) pre-test cognitive and affective data will enhance the predictive reliability of the model.

The model will be tested at VGHSC. It will be tested using a variety of training platforms: desktop, head-mounted displays, and immersive virtual environments such as the Wide Area Virtual Environment (WAVE). VGHSC conducts more than 40,000 hours of training annually for students at the University, as well as military clinicians from the 579 and 779th Medical Groups, the 166th Maryland Air National Guard, and first responders from the State Department and Department of Homeland Security.

We anticipate this work to take 18 months.

This project will leverage current understanding of how humans learn and react in high-stress environments. By applying these principles to virtual reality (VR)-based medical simulation applications, we can develop customized training scenarios that are specifically tailored to the military learner. As our model is application-agnostic, this work is applicable to both current and future military medical simulation applications build on VR gaming platforms.