Future Program: Virtual-Integrated Twin for Autonomous Lifesaving (VITAL)

The Virtual-Integrated Twin for Autonomous Lifesaving (VITAL) program aims to develop continuously updating computational models of the cardiovascular system that integrate patient data with biological physics to predict outcomes in real time. The program has two phases: Phase 1 establishes technical credibility through the development of high-fidelity (HF) models, and Phase 2 evaluates the performance tradeoffs associated with HF-to-ROM transitions. The program will establish a foundation for causal, prediction-driven decision support using high-fidelity digital twins and will pursue a systematic reduction of HF models into fast, continuously updateable, mechanism-preserving surrogates using artificial intelligence techniques. • The program will develop HF models that integrate patient data with biological physics to predict outcomes in real time. • The program will establish an Image-to-Physics-to-Twin pipeline that automatically integrates multimodal clinical imaging with sparse physiological and biochemical measurements. • The program will pursue a systematic reduction of HF models into fast, continuously updateable, mechanism-preserving surrogates using artificial intelligence techniques. • The program will evaluate the performance tradeoffs associated with HF-to-ROM transitions, including scalability, real-time execution, and intervention-forecasting accuracy.