The paper "Federated Learning with Homomorphic Encryption for Ensuring Privacy in Medical Data" by Dr. R. Mohandas, Dr. S. Veena, G. Kirubasri, I. Thusnavis Bella Mary, and Dr. R. Udayakumar explores the integration of Federated Learning (FL) and Homomorphic Encryption (HE) to protect patient privacy in healthcare applications enabled by the Internet of Things (IoT). The authors highlight the benefits of IoT in healthcare, such as continuous tracking, remote support, and improved treatment efficiency, but also emphasize the need for robust privacy protection methods. The paper reviews existing FL and HE techniques, noting that while FL is effective in maintaining privacy by allowing devices to send only their locally trained model updates to a central server, it still faces challenges in protecting sensitive data. The proposed Secured Medical Homomorphic Encryption Algorithm (SMHEA) combines HE with secure computing to enhance privacy. Key contributions of SMHEA include a unique masking strategy that uses a weighted mean approach based on data quality, a solution resistant to collusion and dropouts using Diffie-Hellman key exchange and Shamir secret sharing, and a working model for handling medical information. The system's architecture involves a model aggregating server and dispersed clients, with a third-party Trust Agency (TA) to ensure privacy. The paper details the encryption and decryption processes, emphasizing the security of HE and the system's ability to resist collusion and external attacks. The effectiveness of SMHEA is evaluated using the HAM10000 dataset for skin lesion classification, showing higher accuracy (88.43%) and reduced computation time compared to other FL methods. The study concludes by discussing future directions, including optimizing the system for heterogeneous environments and enhancing security against malicious servers.The paper "Federated Learning with Homomorphic Encryption for Ensuring Privacy in Medical Data" by Dr. R. Mohandas, Dr. S. Veena, G. Kirubasri, I. Thusnavis Bella Mary, and Dr. R. Udayakumar explores the integration of Federated Learning (FL) and Homomorphic Encryption (HE) to protect patient privacy in healthcare applications enabled by the Internet of Things (IoT). The authors highlight the benefits of IoT in healthcare, such as continuous tracking, remote support, and improved treatment efficiency, but also emphasize the need for robust privacy protection methods. The paper reviews existing FL and HE techniques, noting that while FL is effective in maintaining privacy by allowing devices to send only their locally trained model updates to a central server, it still faces challenges in protecting sensitive data. The proposed Secured Medical Homomorphic Encryption Algorithm (SMHEA) combines HE with secure computing to enhance privacy. Key contributions of SMHEA include a unique masking strategy that uses a weighted mean approach based on data quality, a solution resistant to collusion and dropouts using Diffie-Hellman key exchange and Shamir secret sharing, and a working model for handling medical information. The system's architecture involves a model aggregating server and dispersed clients, with a third-party Trust Agency (TA) to ensure privacy. The paper details the encryption and decryption processes, emphasizing the security of HE and the system's ability to resist collusion and external attacks. The effectiveness of SMHEA is evaluated using the HAM10000 dataset for skin lesion classification, showing higher accuracy (88.43%) and reduced computation time compared to other FL methods. The study concludes by discussing future directions, including optimizing the system for heterogeneous environments and enhancing security against malicious servers.