Photonic sensing of hidden vital signs

Non-contact monitoring of vital signs is critical for healthcare and emergency response but becomes especially challenging when the subject is hidden from direct view. Although electronic radar-based solutions have been validated, they encounter critical bottlenecks in detection range, resolution, and robustness against dynamic interference. Here, we report a photonic framework for hidden vital sign sensing to transcend these bottlenecks and explore its potential for intelligent respiration analysis via deep learning. Based on our developed picosecond-temporal-resolution, single-photon-sensitive LiDAR and specialized signal-processing algorithms, we achieve an order-of-magnitude improvement in detection range and resolution over previous radar-based methods and realize multi-target respiration monitoring, localization, and tracking under dynamic interference. Combined with deep learning, the system further identifies abnormal breathing patterns and issues timely alerts. These results demonstrate the strong potential of photonic sensing for continuous health monitoring and search-and-rescue, and motivate future clinical validation of the reported approach.