Data-Centric Cyber-Defense Model for Connected Pediatric Devices

Abstract

The accelerated implementation of interlinked medical tools in pediatric medicine has increased the precision of diagnostic results and personalization of therapy at the same time as the size of the digital attack surface has grown. The current research offers a proposal of a Data-Centric Cyber-Defense Model (DCCDM) specifically adapted to the context of pediatric Internet of Medical Things (IoMT) system, which aims at safeguarding the continuous behavioral and physiological data against unauthorized subsequent access and manipulation. The model is a combination of anomaly-consciousness data pipelines, federated monitoring agents and adaptable encryption protocols, which are consistent with NIST AI Risk Management Framework (AI RMF). With simulated data of wearable and IoT-based pediatric devices, the DCCDM demonstrated 94% intrusion detection accuracy and a 41% decrease in false positive compared to the control intrusion detection systems. The results indicate that ethical and data-centric AI security systems can be used to guarantee privacy, resilience, and accountability in upcoming pediatric IoMT networks.