2024 CR Forum Top Ten| Non-invasive detection of cardiac allograft rejection with AI

We developed an artificial intelligence (AI) tool to analyze recordings of the heart's electrical activity (electrocardiogram) among those who have had a heart transplant. We evaluated a total of 1,427 heart transplant patients with over 7,500 electrocardiograms (ECGs) between 1998 and 2021. The AI-ECG (artificial intelligence enhanced electrocardiogram) was able to determine whether the body is rejecting the donor heart without the need for a heart biopsy which is how this is currently determined. We also tested it in a sample of 56 heart transplant patients who were scheduled to get a heart biopsy as part of their heart transplant care. We recorded ECGs immediately before the biopsy and showed the AI-ECG was able to correctly identify all cases of organ rejection when compared to the heart biopsy results which came back a few days later.