Artificial Intelligence for Response Assessment in Pediatric Neuro-Oncology (AI-RAPNO), part 1: review of the current state of the art.

Abstract

Artificial intelligence (AI) has the potential to enable more precise, efficient, and reproducible interpretation of medical imaging data to improve patient care in paediatric neuro-oncology. Paediatric brain tumours present distinct histopathological, molecular, and clinical challenges that require tailored AI solutions. Recent advances have led to paediatric-specific AI tools for tumour segmentation, treatment response evaluation, recurrence prediction, toxicity assessment, and integrative multimodal analysis. These innovations have the potential to improve diagnostic accuracy, streamline workflows, and inform personalised treatment strategies. However, clinical implementation remains hindered by challenges related to data heterogeneity, model generalisability, and integration into clinical practice. In this Policy Review, we highlight key developments, challenges, and priority areas for imaging-based AI for paediatric neuro-oncology. Our goal is to provide oncology practitioners with a focused overview of current capabilities, unmet needs, and future directions at the intersection of AI and paediatric neuro-oncology.