Personalized Virtual Patient Simulation for Decision Support in Intensive Care Units

Challenges in today’s Intensive Care Units (ICUs) involve the management of patients’ life critical physiological systems. ICUs must constantly make fast and personalized treatment decisions (e.g. ventilation settings, fluid management, and vasopressors) that require constant monitoring, patient specific response predictions, and real time decision support. This project explores the use of Cellular Discrete Event System Specification (Cell-DEVS) as a formal modeling framework to develop personalized virtual patient simulations for ICU decision support. The project will identify important physiological processes in ICU care and define discrete representations that capture their spatial and temporal dynamics. Strategies for mapping continuous ICU measurements into discrete Cell-DEVS state transitions will be investigated, alongside parameter personalization approaches and traceability. In order to validate the final framework’s functionality, it will be compared to existing physiological, machine-learning, and agent-based virtual patient models. Overall, this project aims to establish Cell-DEVS as a complementary modeling approach for personalized ICU risk scenario simulation and AI-assisted clinical decision support.

Faculty Supervisor:

Gabriel Wainer

Student:

Partner:

Universidad de Granada

Discipline:

Computer science

Sector:

Health and Related Sciences and Technology; Artificial Intelligence; Biotechnology

University:

Carleton University

Program: