DEVELOPING A TRUST MODEL FOR AUTONOMOUS SURGICAL ROBOTS IN SPINE SURGERY

The global rise in spinal disorders, particularly in ageing populations, has led to increasing numbers of spinal procedures. Pedicle screw placement remains a technically demanding step, with reported misplacement rates between 17 and 40 percent. Such errors can lead to nerve injury, vascular complications, and revision surgery. Although artificial intelligence driven robotic systems have demonstrated improvements in accuracy, reduced radiation exposure, and enhanced clinical outcomes, surgeons' trust in robotic autonomy remains limited. Reasons include low transparency around robot decision-making and uncertainty regarding accountability and regulatory guidance. Addressing these issues requires a structured approach to building trust between surgeons and autonomous robots.

This project aims to develop a trust model that defines what surgeons need to confidently rely on autonomous robotic functions during spine surgery. To implement this trust model, the project is developing a co-robotic surgical platform that combines advanced AI control, a transparent and interpretable user interface, and a high-fidelity digital twin simulation. Surgeon input will shape the system's design and behaviour by means of qualitative methods in the form of a survey and interviews.

By enhancing transparency and maintaining meaningful human control, this work aims to support safer and more reliable adoption of autonomous robotic systems in spine surgery. The model integrates objective performance measures with clear and continuous communication, ensuring that the robot's actions are understandable, predictable, and open to surgeon oversight.