Harnessing the potential of artificial neural networks for pediatric patient management

Abstract Both artificial intelligence (AI) and clinical data acquisition are advancing at a rapid pace, making this an exciting time as clinicians and computer scientists collaborate to develop automated tools for understanding disease pathophysiology and improving patient care. The development of machine learning algorithms specific to pediatric diseases may help compensate for the pediatric evidence gap as well as the lack of pediatric subspecialists outside of major tertiary care centers. In this chapter, we review early efforts and ongoing advances in applying AI to pediatric disease diagnosis, prognosis, and management. Specifically we discuss studies addressing prematurity, childhood brain tumors, epilepsy, autism spectrum disorder, mood and psychotic disorders, hydrocephalus, traumatic brain injury, and other entities. We highlight the ongoing transition of clinical pediatrics into an increasingly computer-assisted field and discuss challenges and opportunities for growth in this arena.