Camera-based Continuous Heart and Respiration Rate Monitoring in the ICU

n this work, we provide new insights into the performance of camera-based heart and respiration rate extraction and evaluate their usability for replacing spot-checks on the general ward. A study was performed comprising of 36 ICU patients where we aimed to record RGB and NIR video for approximately 24 hours per patient, for a total recording time of 661 hours. The overall 5 beats/minute agreement between camera and ECG-based heart rate measurements was 81.4% and the coverage of the camera-based measurements was 82.8%, where the largest gap between two measurements was 239 minutes. The main remaining challenges that were encountered in this evaluation were situations where the visibility of the patient’s face was limited and patients with irregular heart rates, which in general led to poor agreement between camera and ECG-based heart rate measurements. To ensure that non-breathing motion did not disturb the extraction of the respiration rate, a new metric has been developed which can detect the presence of non-breathing motion. The overall 3 breaths/minute agreement between camera-based and contact sensor respiration rate measurements was 91.1% and a coverage of 59.1%, where the largest gap between two measurements was 114 minutes. There the main remaining challenges were the morphology of the extracted respiration signal and irregular breathing. While a few remaining challenges need to be overcome, the results in this work show promise for the usability of camera-based heart and respiration rate monitoring as a replacement for spot-checks of these vital parameters on the general ward.