Phase II Amount
$1,096,126
BioMojo proposes to develop a language-based intelligent patient simulation platform for the purpose of providing students with a realistic training environment and instructors better control over the training environment involving simulated patients. Building upon the development and lessons learned of Phase I, Phase II will deliver details on the design and performance of the product in an intelligent patient simulation platform in the following areas: functional requirements, architecture design, component design, coding, testing of the software, delivery of the software (on DoD funded mannequins and screen based simulators). The intent of this phase is development of AI software that can learn from available sources or trainings to understand questions asked in different ways; develop simulated patients personalities through learning and coaching (AI/machine learning); build a library of trained patient simulators for different training scenarios; and provide structure for future use with multiple languages. Tactical Combat Casualty Care specific medical training scenarios (e.g. hemorrhage control, airway compromise, tension pneumothorax, fasciotomy, or Advanced Cardiac Life Support scenarios) will be identified for testing the capability of the platform developed in either the mannequin or screen based patient platform. With the selected mannequin simulator, screen based virtual patient simulator, and training scenarios, the platform will be tested to prove the concept of configuring and adjusting parameters on a medical mannequin simulator or screen based virtual patient simulator using voice commands. An Intelligent Patient Simulator built with the platform developed for this topic should: Accurately understand learners questions in various ways in natural conversations Respond to commands by the instructor to change the physiology (e.g. slow down breathing, increase heart rate, become unresponsive, cough, etc.
) Answer and ask questions in a sensible way in the context of medical training Learn from available sources or trainings to understand questions asked in different ways Develop simulated patients personalities through learning and coaching (AI/machine learning) Build a library of trained patient simulators for different training scenarios Provide structure for future use with multiple languages (other than English) Be compatible with DOD funded simulators Integrate with DOD Virtual Patient Simulator Leverage tools provided with the DOD Biogears physiology engine