News

UVM Wins Up to $38 Million to Develop Digital Twins for ICU Treatment

• By John K. Waters
• 10/06/2026

Choosing a treatment for a critically ill patient could eventually include testing it on a virtual version of that patient's immune system. A University of Vermont research team has received a federal contract worth up to $38 million to develop that approach, with clinical trials dependent on earlier validation.

The university announced the award September 30. Led by trauma surgeon Gary An at UVM's Larner College of Medicine, the project seeks to help clinicians understand how an individual patient's condition is changing and evaluate possible interventions.

The Advanced Research Projects Agency for Health's award record identifies the project as ReSCUED, short for Reprogramming Severe Critical-illness Using Extensible Digital-Twins. The contract began September 21.

Researchers plan to combine measurements of immune activity and neurophysiological state with electronic medical records. Those inputs would inform computational models that simulate potential treatments for individual patients at the immune-cellular level. ARPA-H's description includes both FDA-approved medications and novel therapies.

A digital twin is a computational representation updated with information about the system it models. In this project, the intended use is to explore how a patient's immune system might respond to an intervention before a clinician administers it.

The work addresses a problem that conventional monitoring does not fully resolve. According to ARPA-H's CIRCLE program description, immune and inflammatory responses can contribute to organ damage when they become excessive. Clinicians lack tools to track those rapidly changing responses precisely enough to intervene before damage occurs.

CIRCLE, which funds ReSCUED, aims to identify points where treatment could influence those responses. Its proposed digital twins would help connect measurements of immune activity to care decisions.

“The complexity of critical illness exceeds what any individual can interpret in real time,” An said in the university's announcement.

UVM says its approach would collect blood every six hours and use molecular measurements to update the patient model. Researchers intend to use the twin to train an AI-based virtual consultant that could suggest intervention strategies for clinicians.

The project also sits within a broader effort to build the infrastructure needed to evaluate such tools. In its September 22 announcement of CIRCLE awards, ARPA-H outlined supporting work by Verily Health on a shared program database and by Sage Bionetworks on a development and testing environment.

Vanderbilt University Medical Center will support adaptive clinical trials and assess how acceptable and usable the participating teams' tools are in real-world ICUs. ARPA-H's total commitment to CIRCLE is up to $144.9 million over five years, with awards contingent on research milestones.

The program targets a 25% reduction in ICU length of stay. That figure is a goal, rather than a demonstrated result.

UVM describes ReSCUED as an initiative lasting up to five years. The first three years will focus on developing and validating the twin's computational predictions. Later experimental testing and clinical trials depend on meeting those milestones.

For healthcare technology teams, the project illustrates the work between building a patient model and making it useful at the bedside. Shared data and simulation environments support development, but the research must also establish whether the resulting guidance helps clinicians treat patients.