Image-free system for non-contact cardiopulmonary monitoring in supine posture

Conventional contact-sensors for continuous vitals monitoring is uncomfortable for patients, especially in the post surgery day care units. We propose an image-free and completely non-contact hospital bed-based system to monitor cardiopulmonary activity while resting. The system utilizes the concept of respiratory sinus arrhythmia (RSA), which is the variation in heart rate in synchronism with breathing. This information is used to extract the respiratory information from the inter-beat intervals of the J peaks (J-J intervals) of the ballistocardiogram signals. This provides a viable option for continuous monitoring without the need for more modalities, complex processing, and affecting caregiving. Since the system does not have any imaging components, it is free from errors due to ambient light, insufficient illumination, and line-of-sight (LOS) mismatches. Also, since the system does not retain any patient-specific information, it does not raise any privacy concerns. The proposed system showed a strong agreement between the estimated and ground truth values with a pearsons's correlation coefficient (r) of 0.86 and 0.92 for respiratory rate and heart rate values respectively. The system also had a root mean squared error (RMSE) of 3.2 bpm for heart rate values and 2.8 brpm for respiration values, and a mean absolute error (MAE) of 2.8 bpm for heart rate values and 1.6 brpm for respiratory rate values.