Title: Leveraging Cross-Modal Contrastive Learning of ECG and Angiography Data to Classify Severe Stenosis

Original: arXiv:2606.02605v1 Announce Type: cross Abstract: Coronary artery stenosis is a common cardiovascular disease, with severe, untreated cases posing significant risks of heart attack. Although coronary (X-ray) angiograms remain the standard for stenosis diagnosis, they are invasive, time- and resource-intensive, and therefore only performed on patients with a high probability of disease based on symptoms and prior clinical tests. However, a subset of patients, especially those without symptoms, may remain undiagnosed. Detecting indications of stenosis from ECGs, which are fast, cheap, non-invasive, and thus routinely acquired even in asymptomatic patients, would support early diagnosis. However, as no reliable stenosis-specific signal has been identified in ECGs, they can not currently be used for stenosis risk stratification. To address this, we introduce StenCE, a pretraining framework, allowing stratification of patients based on features derived directly from ECGs. Evaluations across varying stenosis severity thresholds and additional ECG disease classification tasks demonstrate consistent performance improvements across different ECG encoders, outperforming previous work. The obtained models successfully detect signals for stenosis diagnosis in ECGs and are the first to achieve high performance in severe stenosis classification. The source code is available at https://github.com/NikolaCenic/ecg-stenosis-cls.

Rewrite: Coronary artery stenosis represents a prevalent cardiovascular condition, where untreated severe instances carry a substantial risk of myocardial infarction. While coronary angiography (X-ray) is currently the diagnostic gold standard, its invasive nature, coupled with high time and resource demands, restricts its use to individuals exhibiting strong clinical indicators based on symptoms or previous testing. Consequently, certain patient groups, particularly those who are asymptomatic, often go undetected. Since electrocardiograms (ECGs) are rapid, inexpensive, non-invasive, and frequently collected even in the absence of symptoms, identifying stenosis markers within ECG data could facilitate earlier diagnosis. Nevertheless, the absence of a dependable stenosis-specific signal in ECGs has prevented their current utility in risk stratification. To overcome this limitation, we present StenCE, a pretraining framework designed to stratify patients using features extracted directly from ECGs. Our evaluations, which span various stenosis severity thresholds and additional ECG disease classification tasks, reveal consistent performance enhancements across multiple ECG encoders, surpassing prior methodologies. These models successfully identify diagnostic signals for stenosis within ECGs and represent the first approach to achieve high performance in classifying severe stenosis. The source code is accessible at https://github.com/NikolaCenic/ecg-stenosis-cls.

Source: arXiv Generated at: 2026-06-03 00:00:00 UTC