Camera-Based Cardiorespiratory Monitoring of Preterm Infants in NICU

Video health monitoring that combines video processing and physiological measurement has enabled contactless vital signs monitoring of infants. In the neonatal intensive care unit (NICU), intensive monitoring of vital signs like heart rate (HR) and respiratory rate (RR) is critical for early warning and proper treatment of infants. The monitoring of variability in HR (HRV) and RR respiratory rate variability (RRV) is equally important, as these indices provide valuable insights into the stability of the infant's autonomic nervous system (ANS) and cardiorespiratory system. This study collected video data from 50 preterm infants in NICU using remote cameras. We developed a novel framework to robustly measure HR, RR, HRV, RRV, and actigraphy from the videos. Various options in vital signs extraction are benchmarked in this framework and the best candidate for each measurement is concluded. This framework allows robust extraction of multiple cardiorespiratory parameters and actigraphy from infant's videos. In particular, HR and RR measurements are within the clinical acceptance range (reference ±5 bpm), which meets the ANSI/AAMI EC13:2002 standard and NIH recommendations.