Oncology undergoes technological transformation through Artificial Intelligence (AI) along with Machine Learning (ML) and the Internet of Things (IoT) which provide both real-time patient monitoring and early diagnosis together with personal treatment strategies. A FL-based IoT oncology framework analyses patient data through distributed AI models which process wearable sensor data along with EHRs and imaging systems while protecting privacy integrity. The FL model serves three analytical purposes through its ability to assemble data-driven insights from different healthcare facilities while both preserving sensitive patient information and following privacy protection requirements. Medical data privacy improvements are accomplished through our method which eliminates security vulnerabilities and bias problems that occur in standalone AI systems. The FL predictive models deliver superior cancer prognosis outcomes than conventional centralized methods while preserving data privacy and achieving interoperability according to experimental outcomes. The new system provides secure and AI-enabled scalable cancer healthcare solutions which enhance patient health results as well as minimize hospital return frequencies.