Deployment of an On-the-Edge Clinical Decision Support System in Neonatal Intensive Care Units.

Preterm infants require continuous monitoring prior to discharge from neonatal intensive care units (NICU). Conditions such as icterus and sepsis are life-threatening and may cause long-term consequences for this population. The deployment of an AI-based clinical decision support system (CDSS) facilitating early prediction, diagnosis, and intervention in the NICU setting is challenging yet promising. The objective of this work is to design, implement, deploy, and technically evaluate a CDSS that integrates quasi-real-time signal processing chains and AI models on the edge in a clinical context. The proposed system consisting of data transmission, pseudonymization, data fusion, processing, and inference was deployed at the University Hospital of Rennes in Jan 2023. During the first six months of deployment, the service continuously received monitoring signal data from 138 neonates, processing live data and generating bilirubin level estimations at a temporal resolution of 15 minutes. Results on the stability and robustness of the service are presented. To our knowledge, this is the first description of a multi-source, on-the-edge CDSS deployed in a NICU scenario for patient-specific early detection of high-risk events. This proof-of-concept is particularly encouraging and further prospective evaluations on clinical performance are warranted.