The University of Vermont has received a federal research award worth up to $38 million to develop artificial intelligence models that could help doctors select treatments for critically ill patients.
The Advanced Research Projects Agency for Health's award record lists Burlington-based UVM as the lead recipient and Sept. 21 as the start date. Gary An, a trauma surgeon and researcher at UVM's Larner College of Medicine, leads the project.
The work will combine measurements of immune activity and physiological condition with electronic medical records to build a computational version of a patient's illness. Researchers would use those models to simulate possible treatments, including approved medications and experimental therapies.
In its Sept. 30 announcement, UVM described an initiative lasting up to five years, with funding and later stages tied to milestones. The first three years would focus on developing and validating the models. Further experimental work and clinical trials would follow if those milestones are met.
UVM says the researchers aim to shorten intensive care stays by at least 25 percent. That is a research goal, rather than a result already demonstrated by the project.
Larner will develop the computational platform. The Cambridge-based DNA Medicine Institute is among the partners and will supply a bedside molecular testing system, UVM said.