Pulse physician Ravi Hira, MD, and team develop AI tool to predict risk of PCI complications

This algorithm can help clinicians as they determine treatment for blocked arteries. Risk prediction is important in selecting the best treatment option for each patient, but because risks for patients undergoing PCI vary greatly, existing tools have only been modestly successful. The algorithm can help eliminate some over- and underestimating of those risks.

The research team used data from adult patients who underwent PCI, including age, cholesterol levels and blood pressure prior to the procedure, and analyzed it with a machine-learning AI software to create a risk prediction model. The AI-driven model showed significant accuracy at predicting death, major bleeding events and the need for blood transfusion for short-term outcomes.

“We continue to strive to provide the most up-to-date and individualized risk assessment for our patients so they can understand their own risk and make informed decisions for their heart health in collaboration with their cardiologist,” explains Dr. Hira, cardiologist at MultiCare’s Pulse Heart Institute.

The innovative technology has been made into a computer and phone application, which is publicly available for no cost to encourage widespread use. The complete study and results were published in European Heart Journal.

For more information about research opportunities, please contact the MultiCare Institute for Research & Innovation (MIRI). MIRI is dedicated to improving health care by providing access to clinical trials that connect our patients to cutting-edge research. This research enables our clinical providers to deliver high-quality care and support to patients throughout our service areas — and contribute valuable knowledge that improves well-being for patients everywhere. To learn how you can get involved, email [email protected].