The University of Minnesota Medical School has received $1.4 million in funding for a research project to develop methods and risk management processes for clinical artificial intelligence (AI) and machine learning to optimize continual patient safety and improve trust in these innovations.
“While AI is increasingly being utilized to improve patient care, critical methodologic and sociotechnical gaps in best practices remain,” said Genevieve Melton-Meaux, MD, PhD, a professor at the U of M Medical School in the Department of Surgery, core faculty in the Institute for Health Informatics and director of the Center for Learning Health System Sciences.
During this project, “ENsuring the TRUSTworthiness of AI/ML Models to Optimize Continual Patient Safety” (ENTRUST AI), this team will:
- Develop robust computational approaches to provide individualized reliability, harms and benefits;
- Extend current risk management best practices to the full clinical AI lifecycle and implement these optimized individualized risk management processes at M Health Fairview and Mayo Clinic, specifically for AI algorithms predicting patient decompensation and postoperative surgical complications.
“It’s important to understand the performance of clinical AI at the patient level. While a machine learning AI model may perform well for most patients across a larger population, what matters in patient care is the model’s accuracy as it applies to each individual patient,” said Gyorgy Simon, PhD, an associate professor at the U of M Medical School, core faculty in the Institute for Health Informatics and scientific co-director for the Program for Clinical AI in the Center for Learning Health System Sciences.
Drs. Melton-Meaux and Simon along with Hongfang Liu, PhD, a professor of biomedical informatics in the Mayo Clinic College of Medicine and Science, and Pedro Caraballo, MD, of Mayo Clinic serve as principal investigators of this study.
The funding for this project comes from Minnesota Partnership for Biotechnology and Medical Genomics (MNP) Program, which is funded by the State of Minnesota (Award number - MNP #22.29). The award period is March 1, 2023, through February of 2025.