Asclepius Phase IIB

Asclepius is an artificially intelligent (AI) perception system for automating TCCC scene interpretation and patient encounter documentation. Asclepius is designed to capture video from deployed cameras and data from inertial measurement units (IMUs) donned by the medic. A custom suite of machine perception algorithms analyzes the data to identify casualties and a set of procedures within the 68W health care specialist’s scope of practice, as identified in Soldier Training Publication STP 8-68W13-SM-TG. The system automatically provides output information on the detected procedure(s) in XML/JSON format that is human and machine readable. This approach supports integration and interoperability with other components of the medic’s kit. As a result of the work completed during the Phase II SBIR, Asclepius’s perception algorithms are being used as part of the Telemedicine and Advanced Technology Research Center (TATRC) AutoDoc effort. AutoDoc aims to support passive data collection and autonomous documentation of care delivery during tactical combat casualty care (TCCC) environments. By leveraging the training pipelines established in the Phase II SBIR, DI is able to extend its perception algorithms to continue support of the AutoDoc effort. The proposed effort aims to improve Asclepius and integrate its capabilities to support TATRC’s AutoDoc project through iterative software releases. This expansion includes;  1.) Increase the number of interventions and medical objects detected and tracked with Asclepius for documentation and resource tracking; 2.) Add injury detection, patient detection and note documentation modules to Asclepius 3.) Improve Asclepius’s performance through expansion of the training dataset from TATRC provided simulated TCCC scenario data; 4.) Continuously update and integrate Asclepius’s algorithms with TATRC’s AutoDoc software architecture.