Kamal Shah, Ph.D., a graduate of the UW Bioengineering and former researcher in the Yager research laboratory, is the lead author of a major new study published in Nature (“Automated loss of pulse detection on a consumer smartwatch”) that explores how smartwatches could help detect cardiac arrest. The research focuses on developing a machine learning algorithm that uses data from the watch’s photoplethysmography (PPG) sensor—a technology that measures blood flow—to automatically detect sudden loss of pulse, which is a key sign of cardiac arrest. Because unwitnessed cardiac arrests are often fatal, a wearable device that can recognize the event and call for help could be life-saving.

To test the system, researchers used a model that mimics pulselessness by temporarily cutting off blood flow in the arm, and found that the smartwatch PPG signal during this simulated condition closely resembles what happens during ventricular fibrillation (Vfib)—a life-threatening heart rhythm disorder closely linked with cardiac arrest. The algorithm was trained and validated using data from both these controlled tests and real-world conditions. In prospective studies, the system achieved a sensitivity of about 67% while generating very few false alarms—only one unintended emergency call per 21.67 user-years—making it potentially viable for widespread public use.

We hope that our research, and the Loss of Pulse Detection feature that came from it, can help improve outcomes from unwitnessed loss of pulse events. – Kamal Shah

Shah now works as a research scientist at Google. His work represents a significant step forward in using wearable devices for early detection of life-threatening emergencies, with the potential to improve survival rates by getting expedient aid to those who need it.