From a physiological perspective, heartbeats and breathing do not operate independently in the human body. Cardiac rhythm varies with the respiratory cycle, and the close interaction between the two is known as cardiopulmonary coupling (CPC). CPC reflects the regulatory state of the autonomic nervous system and serves as an important physiological indicator for evaluating sleep quality, cardiovascular health, and stress levels.
For decades, however, cardiopulmonary coupling has been assessed almost exclusively using electrocardiogram electrodes, respiratory belts, or other contact-based sensors. While effective, these devices can compromise comfort and limit their use in home environments, long-term monitoring, and applications involving sensitive populations such as infants or post-operative patients.
In a study published in Intelligent Sports and Health , a research team led by Professor Rencheng Song at Hefei University of Technology presents a new solution: measuring cardiopulmonary coupling using only a standard video camera.
"Our bodies undergo extremely subtle surface changes with every breath and heartbeat, and cameras are capable of capturing this information," explains senior author of the study Rencheng. Song "This makes it possible to assess cardiopulmonary health under completely non-contact conditions."
The researchers developed an intelligent video-analysis framework that first automatically identifies facial and torso regions in video recordings. "From these regions, optical signals associated with blood flow variations—reflecting cardiac activity—and chest and abdominal motion—reflecting respiration—are synchronously extracted," says Song.
Through multi-region signal fusion and high-resolution time–frequency analysis, the method further improves signal stability and enhances the characterization of dynamic physiological changes. Experimental results showed that under different physiological conditions, including normal breathing and simulated apnea, the video-based approach produced cardiopulmonary coupling measurements that were highly consistent with those obtained using conventional contact-based devices.
Principal investigator Yi Qun Gao, noted that the work extends the application boundaries of non-contact physiological monitoring. "For a long time, cardiopulmonary coupling monitoring has depended mainly on medical equipment or wearable sensors. Our study demonstrates that remote video devices also have the potential to enable CPC assessment," he says.
The team highlighted that the approach is particularly well suited to scenarios requiring long-term, comfortable monitoring, such as in-home sleep health screening or remote post-operative rehabilitation. "In the future, such technology may even be integrated into smartphones or smart cameras, making health monitoring as natural as a video call," adds Gao.
