CAMEL AI Treats ECG as Language to Flag Looming Cardiac Crises

Penn researchers developed CAMEL, an AI system that analyzes hours of in-hospital ECG telemetry to forecast dangerous heart rhythms minutes before they occur, framing ECG signals like language to look for patterns that precede events. The goal is to provide earlier warnings (10–15 minutes) to improve care while avoiding false alarms, with plans to test real-time data in clinical settings and explore extending the approach to consumer wearables for broader monitoring.