Article
Author: M Itmamul Wafa Email M Itmamul Wafa.
The increasing integration of telesurgery systems in modern healthcare introduces critical challenges in maintaining secure and reliable communication channels between surgeons and remote robotic instruments. This study proposes a machine learning-based approach to predict encryption success in telesurgery communication, aiming to enhance cybersecurity monitoring and prevent potential encryption failures that may compromise patient safety. Two ensemble learning algorithms, Random Forest and XGBoost, were employed to model encryption performance using a dataset containing parameters such as network latency, data transfer rate, response time, threat severity, and robotic gesture information. The models were trained and evaluated through standard classification metrics, including accuracy, F1-score, and ROC-AUC. Experimental results revealed that the Random Forest model achieved superior performance, with an average accuracy and ROC-AUC of approximately 0.52 and 0.53, respectively, outperforming XGBoost across all key metrics. Feature importance analysis identified response time, data transfer rate, and network latency as the most influential predictors of encryption reliability, emphasizing the strong dependency between network efficiency and cryptographic stability. Although overall model performance was moderate, the findings demonstrate the potential of predictive analytics in identifying network conditions that contribute to encryption degradation. The proposed framework provides a foundation for developing intelligent cybersecurity systems capable of autonomously assessing encryption health in telesurgery, thereby improving the safety, integrity, and resilience of remote surgical operations.
Keywords: Telesurgery, Encryption, Machine Learning, Cybersecurity, Network Performance
How to Cite: Wafa, M. (2026) “Machine Learning-Based Prediction of Encryption Success for Cybersecure Telesurgery Systems ”, Journal of Cyber Law. 2(1). doi: https://doi.org//JCL.153