Continuous movement quantification in preterm infants using fiber mat: Advancing long-term monitoring in hospital settings

Movement patterns in preterm infants can offer crucial insights into their physiological state including maturational development and sleep. These patterns can also serve as early indicators for potential deteriorations, such as cerebral palsy, sepsis, and epilepsy. In this study, we investigated a novel 2-D optical fiber mat system for the automated monitoring of infant movement, thereby enhancing the efficiency and safety of neonatal care in both the neonatal intensive care unit (NICU) and neonatal medium care unit (NMCU). The 20 preterm infants admitted to both NICU and NMCU were enrolled in the study. They underwent monitoring for a duration of 2–5 h using both an optical fiber mat and a camera which provided valuable movement annotations. The signals from the fiber mat were quantified, selected, and then integrated into a consolidated movement signal. This signal was subsequently transformed into binary states, distinguishing between “movement” and “still” based on the distribution of the movement signal. The proposed fiber mat system achieved a mean [standard deviation (SD)] area under the receiver operating curve (AUC) of 0.91 (0.05), and an F-score of 0.73 (0.09), when compared with manually annotated video recordings. This study demonstrates the feasibility of continuous movement monitoring for preterm infants within hospital settings. It illustrates the promising potential to evolve into a predictive tool for monitoring patient deterioration through the fusion of physiological information in both hospital environments and within the comfort of homes.