Federated Learning in Oncology: Privacy-Preserving Artificial Intelligence for Multi-Center Cancer Research

Cancer remains one of the leading causes of morbidity and mortality worldwide, creating an urgent need for advanced computational approaches that can improve diagnosis, prognosis, and personalized treatment strategies. Artificial intelligence (AI) has demonstrated significant potential in oncology by enabling automated medical image analysis, biomarker identification, treatment response prediction, and integration of complex biomedical datasets. However, conventional AI approaches generally rely on centralized data aggregation, where patient information from multiple institutions is transferred to a single location for model training. This approach presents major challenges related to patient privacy, data ownership, regulatory compliance, and institutional barriers, limiting large-scale multi-center collaboration. Federated learning (FL) has emerged as a privacy-preserving AI framework that enables multiple healthcare institutions to collaboratively develop machine learning models without sharing raw patient data. In oncology, FL provides opportunities for secure analysis of medical imaging, genomic information, electronic health records, and clinical trial data while maintaining data confidentiality. This approach can enhance cancer detection, tumor characterization, disease progression modeling, and personalized therapeutic decision-making. Despite its advantages, implementation of FL in oncology faces challenges including data heterogeneity, model bias, communication limitations, cybersecurity risks, and lack of standardized validation frameworks. Future integration of FL with deep learning, generative artificial intelligence, multi-omics analysis, and large-scale cancer research networks may accelerate the development of privacy-preserving precision oncology platforms. This review discusses the principles, applications, architectures, challenges, and future prospects of federated learning in multi-center cancer research.