NYU Study Combines ECG and Blood Biomarkers to Assess Heart Transplant Rejection
Published in Artificial Intelligence, Journals, Tools.
Publication date: October 7, 2026.
NYU Langone reported on October 6 an AI approach combining electrocardiograms with blood biomarkers to assess heart transplant rejection. The underlying study appeared online on September 25 in JHLT Open.
The team used 5,300 ECG recordings from 2,357 adult transplant recipients, with biopsy results as the reference. The combined model incorporated biomarkers reflecting immune-related gene activity and donor-derived DNA.
In an additional test group of 38 recipients, NYU reports that the combined model correctly identified 94% of patients without rejection. This percentage describes identification of non-rejection cases; it is not an overall accuracy estimate or a demonstrated sensitivity for detecting every rejection episode.
The findings suggest a possible way to reduce unnecessary biopsies, but do not establish that the model can replace biopsy-based care. The researchers plan further testing in larger groups across multiple transplant centers.
Source: NYU Langone, October 6, 2026. Study: 10.1016/j.jhlto.2026.100698.