Digital Twin Interfaces for ICU Decision Making

October 24, 2025

The Intensive Care Unit (ICU) is a high-stakes environment, requiring a significant volume of data, which is necessary for guaranteeing the right decision contributes to the patient’s improvement in the long run. Minimizing potential human error and enhancing the ability to predict how the treatment is going to evolve are key aspects to improving this area of treatment. This dissertation proposes the creation and use of a Digital Twin (DT) interface for patients in the ICU with the goal of improving patient care and decision-making via predictive analytics and real-time data integration. This dissertation investigates how they might be able to help with issues in ICUs. With the purpose of identifying how DTs can be implemented in ICUs, a Systematic Literature Review (SLR) was conducted. Based on the results from the SLR, a proposed solution is described. The proposed information system uses Machine Learning (ML) and Retrieval-Augmented Generation (RAG) to simulate treatment alternatives and forecast results, and design and usability principles to enhance user experience. The system was implemented, tested and evaluated, using the Design Science Method. Three prototypes of the system were created and evaluated by medical specialists.

Authors in the community:

• 

  J

  João Pedro Proença Ramos

  ist199253

Supervisors of this institution:

• 

  M

  Miguel Mira da Silva

  ist13948

electrical-engineering-electronic-engineering-information-engineering - Electrical engineering, electronic engineering, information engineering

eng - English

Instituto Superior Técnico