The current standard for monitoring transplant health involves a biopsy, which requires surgeons to remove a small piece of heart muscle for microscopic examination. To find a less invasive alternative, researchers trained AI models on 5,300 EKG readings from 2,357 heart transplant recipients. By integrating these electrical activity patterns with existing blood biomarker tests, the team created a tool that significantly improves diagnostic precision.

In a trial involving 38 patients, the combined AI model correctly identified 94 percent of those not experiencing rejection. While blood tests alone often yield false positives that trigger unnecessary biopsies, the new model effectively filtered these results. Dr. Lior Jankelson, senior author of the study published in the Journal of Heart and Lung Transplantation, noted that EKGs contain physiological information that remains largely underutilized in clinical settings. The research team intends to validate these results through multi-center trials, aiming to standardize a method that could enable earlier intervention and reduce the physical burden on transplant recipients.