ID: 4349377 IMPROVING WEARABLE CARDIAC ARREST DETECTION WITH CLINICAL DATA

Cardiac arrest affects over 4 million people globally each year. Early detection and timely intervention can double the chances of survival. Wearable devices offer a promising solution for continuous monitoring and automated emergency alerting. However, despite growing interest, the real-world performance of wearable-based cardiac arrest detection remains unknown. This is due to the limited availability of real-world cardiac arrest data, with prior work relying on simulated events for model development and evaluation.