Advancing Paediatric Care Through Integrated Machine Learning, Deep Learning, Internet Of Things, Data Science And Mathematical Modelling: A Comprehensive Review
TL;DR
The objective of this research is to review the recent evidence on pediatric applications of ML/DL/IoT/DS and highlight promising but heterogeneous findings across sepsis prediction, remote physiological monitoring, autism spectrum disorder (ASD) screening, precision dosing, child maltreatment risk modelling, and geospatial resource allocation.
Abstract
The intersection of machine learning (ML), deep learning (DL), Internet of Things (IoT), and data science (DS) is transforming healthcare for children, but the evidence base is fragmented across disease areas and technologies, and pediatricpopulations remain underrepresented in both algorithmic training sets and regulatory approval processes. The objective of this research is to review the recent evidence on pediatricapplications of ML/DL/IoT/DS. We focused on four domains: predictive analytics and precision dosing, remote monitoring and developmental screening, child protection and mental health risk prediction, and population-level resource allocation. We performed a narrative review searching PubMed, Google Scholar, and specialist repositories for systematic reviews, scoping reviews, and primary studies published 2019-2026. We included articles explicitly specifying a pediatric population (typically under 18 years). We highlight promising but heterogeneous findings across sepsis prediction, remote physiological monitoring, autism spectrum disorder (ASD) screening, precision dosing, child maltreatment risk modelling, and geospatial resource allocation. In all cases, pediatric-specific studies were underrepresented, often reporting smaller, single-center cohorts. Fewer than 1 in 10 FDA-approved medical devices incorporating AI were explicitly approved for pediatric use (as of 2024). While early ethical frameworks such as ACCEPT-AI and PEARL-AI provide guidance on pediatric AI/ML development, pediatric-specific governance, privacy regulation, and algorithmic auditing lag behind those for adult populations. ML/DL/IoT/DS offer unprecedented opportunities to improve the personalization and timeliness of pediatric care. To realize this potential, multi-centerpediatric-specific datasets, rigorous subgroup and developmental-stage analyses, external validation, and pediatric-specific data-governance guidelines will be essential.