Analysis of Scientific Publications on the Use of Artificial Intelligence in Childhood Cancer: A Descriptive Study

Abstract

Aim: This study aimed to analyze the scientific literature on the use of artificial intelligence (AI) in childhood cancers through a bibliometric approach, combining descriptive analyses with analytical synthesis to identify publication trends, leading countries, journals, institutions, and major research themes.Methods: A bibliometric analysis was conducted using publications retrieved from the Web of Science Core Collection database between 2003 and 2025. A total of 1.091 publications were included in the analysis. Articles and review papers published in English were included. Bibliometrix (R), VOSviewer, and descriptive analyses were used to evaluate publication output, country collaboration, influential journals, institutions, and keyword trends.Results: The findings revealed a steady increase in scientific output, particularly after 2017. Research activity was predominantly concentrated in high-income countries, with the United States and China leading publication output and international collaboration. Keyword analysis indicated that machine learning, deep learning, diagnosis, risk stratification, radiomics, and survival were the most frequently studied topics.Conclusion: This bibliometric analysis provides a structured overview of how AI applications are widely used in the prediction, diagnosis, and treatment processes of childhood cancers, with scientific output increasing markedly in recent years and predominantly concentrated in high-income countries. Future research should prioritize the development, validation, and clinical implementation of AI-supported assessment tools and diagnostic models specifically designed for pediatric oncology.