Mega Doctor News
By Children’s Hospital of Philadelphia
Newswise — Philadelphia – Researchers at Children’s Hospital of Philadelphia (CHOP) published new guidance to help responsibly implement artificial intelligence (AI) in pediatric neuro-oncology. The work includes a policy review from the AI in Response Assessment in Pediatric Neuro-Oncology (AI-RAPNO) subcommittee and a companion review in The Lancet Oncology, detailing how AI can enhance tumor assessment, support treatment planning and improve outcomes for children with brain tumors.
RAPNO is an international consortium that develops criteria for response assessment in pediatric brain tumors. AI-RAPNO is its subcommittee focused on translating AI into those criteria. The new policy blueprint addresses how to standardize imaging, validate algorithms in pediatric populations, and ensure equitable access so that AI tools are safe, effective, and trustworthy at the point of care.
Pediatric central nervous system (CNS) tumors are the most common solid tumors in children and adolescents, with approximately 48,000 new cases annually worldwide. Prognosis varies due to the differing treatability and lethality of tumors, and treatments often result in long-term side effects and reduced quality of life. MRI plays a crucial role in diagnosing and monitoring these tumors.
While AI advancements like 3D tumor auto-segmentation and radiation therapy planning have been FDA-approved for adult oncology, pediatric standards were lacking. The AI-RAPNO policy and its companion review address this gap by providing actionable clinical translation recommendations.
Key recommendations from the AI-RAPNO policy include:
Standardized Imaging Inputs and Labels: These are consistent and clear definitions used in MRI scans of children to improve how doctors evaluate and monitor brain tumors. This includes specific sequences, data descriptions, and outlined areas on images, all tailored to the RAPNO initiative for better understanding of children’s brain health.
Pediatric-Specific Validation: This means ensuring that the imaging techniques and data work well across different groups of children, considering their age, type of tumor, location, and the type of MRI scanner used.
Trustworthy Evaluation: This involves using clear and honest reporting of results, understanding the limits and accuracy of the techniques, and involving medical professionals in the process to ensure reliable outcomes.
Equity and Access: This focuses on making sure that all children, regardless of their background or location, have access to the best possible imaging care. It involves addressing data shortages, preventing biases, and supporting resource-limited areas.
Prospective Pilots: These are initial test runs where AI tools are used in real clinical settings and trials. They are designed to improve the methods used to assess children’s brain tumors, in line with RAPNO goals.
“These recommendations offer a practical roadmap to move from promising research to safe, equitable bedside use,” said Anahita Fathi Kazerooni, PhD, lead author of the policy article and senior author of the review, co-lead of AI-RAPNO, and Assistant Professor at CHOP’s Center for Data Driven Discovery in Biomedicine (D3b) within the Division of Neurosurgery, and the Department of Neurosurgery at the University of Pennsylvania. “As AI evolves, standardization, validation and transparency will be key to realizing personalized care for children with brain tumors.”
The companion review in The Lancet Oncology synthesizes evidence that AI can automate and improve precision of tumor measurements compared with manual methods; support non-invasive molecular subtyping to inform treatment; predict recurrence and progression risk to guide surveillance; and evaluate treatment-related side effects, informing survivorship and quality-of-life care.
“These guidelines help clinicians understand when and how to trust AI outputs, what evidence to look for, and how to integrate tools with RAPNO-aligned decision-making,” said Ali Nabavizadeh, MD, senior author of the policy article and co-senior author of the review, co-lead of AI-RAPNO, Director of the Translational Imaging Research Unit at CHOP’s D3b and Associate Professor at the University of Pennsylvania.
Policy review:
Fathi Kazerooni et al. “Artificial Intelligence for Response Assessment in Pediatric Neuro Oncology (AI-RAPNO) – Challenges, Opportunities, and Recommendations for Clinical Translation.”. The Lancet Oncology. Online October 28, 2025. DOI: 10.1016/S1470-2045(25)00484-X.
Companion review:
Kann et al. “Artificial Intelligence for Response Assessment in Pediatric Neuro Oncology (AI-RAPNO) – review of the current state-of-the-art.” The Lancet Oncology. Online October 28, 2025. DOI: 10.1016/S1470-2045(25)00489-9.
