Explainable ML-based Secure Communication Framework for Robotic Telesurgery in Healthcare

Telesurgery, a technology used in healthcare that provides access to remote surgical operations with precision and reliability, also faces potential cyber attacks because of the use of network-based communication. This paper proposes an explainable machine learning-based secure communication framework for a telesurgery system consisting of three layers: a doctor layer, a network layer that contains the AI layer, and a patient layer. The framework contains a random forest machine learning model (accuracy: 99%, features selected: 9) to classify whether the incoming data is an attack or a normal command/operation. Additionally, using SHAP, we also employ explainable AI, ensuring transparency in the process. The framework is designed to detect malicious commands from the doctor’s side and manipulated information from the patient’s side. The model shows satisfactory results in ensuring secure and reliable communication in the system.