In a groundbreaking development poised to reshape cardiac care, artificial intelligence (AI) has emerged as a promising tool for assessing the risk of sudden cardiac death (SCD). A team led by Dr. Ziad Obermeyer at the University of California, Berkeley, has developed an AI system that analyzes extensive electrocardiogram (EKG) data to identify patients at significant risk of SCD. This innovative approach has shown a potential to outperform current echocardiogram methods.
Sudden cardiac death accounts for 10-15% of global fatalities, with more than 300,000 deaths annually in the United States alone. One preventive measure is the use of implantable cardioverter-defibrillators (ICD), which can deliver a lifesaving shock to restore heart rhythm during a sudden cardiac event. However, determining which patients genuinely need an ICD is challenging, often leading to unnecessary surgeries and associated risks.
Traditionally, echocardiograms have been used to screen for potential ICD candidates. Specifically, a low left ventricle ejection fraction indicated by an echocardiogram suggests an increased risk of SCD, warranting an ICD implantation. Despite this, the echocardiogram’s predictive accuracy remains limited. Many patients without the echocardiogram’s abnormal reading still experience SCD, and others with abnormal readings never undergo cardiac events.
Dr. Obermeyer’s AI research is a leap forward in addressing this shortfall. By analyzing six years of Swedish health system data, the AI examined 110,000 EKG readings from 35,000 patients, training itself to recognize patterns associated with SCD. This AI identified a subset of patients with a 7.0% annual risk of SCD—significantly higher than the 4.6% risk detected via echocardiograms. Notably, 80% of high-risk patients, as per AI analysis, would have gone undetected with traditional methods and potentially missed the opportunity for ICD intervention.
The validation of these findings across independent datasets from the United States and Taiwan further strengthens the AI’s reliability. However, AI systems traditionally lack transparency, functioning as “black boxes.” To address this, a secondary AI system helped pinpoint the critical EKG alteration indicative of high SCD risk. It focused on an anomaly in the aVL lead of EKGs, which may correlate with heart tissue damage disrupting electrical signals.
While the precise connection between this EKG anomaly and increased SCD risk remains under investigation, the AI’s ability to visualize potential risks presents a significant advancement. This is the first time AI has visually detected an abnormality interpretable by human physicians, offering a practical screening tool that is both simple and cost-effective.
Dr. Obermeyer’s work is progressing toward broader clinical applicability, refining the tool for everyday use by cardiologists and primary care physicians. Although not yet ready for routine practice, this represents a transformative step in healthcare, leveraging AI to enhance clinical outcomes and deepen understanding of heart disease mechanisms.
For individuals interested in contributing to this pioneering research, further details are available on Dr. Obermeyer’s patient website. This technological leap not only sharpens the tools available for life-saving cardiac treatments but also paves the way for similar AI applications in various medical disciplines.
You can read the original article here: [Forbes](https://www.forbes.com/sites/paulhsieh/2026/07/31/how-artificial-intelligence-discovered-a-new-way-to-detect-patients-at-risk-of-cardiac-death-using-simple-ekgs/)
