Enhancing reliability and security in cloud-based telesurgery systems leveraging swarm-evoked distributed federated learning framework to mitigate multiple attacks.

Advances in robotic surgery are being driven by the convergence of technologies such as artificial intelligence (AI), 5G/6G wireless communication, the Internet of Things (IoT), and edge computing, enhancing clinical precision, speed, and real-time decision-making. However, the practical deployment of telesurgery and tele-mentoring remains constrained due to increasing cybersecurity threats, posing significant challenges to patient safety and system reliability. To address these issues, a distributed framework based on federated learning is proposed, integrating Optimized Gated Transformer Networks (OGTN) with layered chaotic encryption schemes to mitigate multiple unknown cyberattacks while preserving data privacy and integrity. The framework was implemented using TensorFlow Federated Learning Libraries (FLL) and evaluated on the UNSW-NB15 dataset. Performance was assessed using metrics including precision, accuracy, F1-score, recall, and security strength, and compared with existing approaches. In addition, structured and unstructured security assessments, including evaluations based on National Institute of Standards and Technology (NIST) recommendations, were performed to validate robustness. The proposed framework demonstrated superior performance in terms of diagnostic accuracy and cybersecurity resilience relative to conventional models. These results suggest that the framework is a viable candidate for integration into teleoperated healthcare systems, offering improved security and operational efficiency in robotic surgery applications.