Research

Research

Rapid advances in  smart surgical technologies, robotics, and data-driven medicine have transformed the modern operating room into a highly digitalized environment. Today's surgical suites continuously generate multimodal data, ranging from preoperative imaging and intraoperative video to sensor streams and robotic kinematics. However, the integration and clinical interpretation of these heterogeneous sources remain the primary bottlenecks toward achieving context-aware, patient-specific care. 

Within the interdisciplinary framework of TSO, we bridge AI, computer science, engineering, and clinical surgery to advance the emerging field of Surgical Data Science. Our mission is to transform surgical data into AI that supports decision-making, skill assessment, and autonomous or semi-autonomous execution in computer- and robot-assisted interventions.

Our research builds upon state-of-the-art machine learning and computer vision methods for video-based workflow understanding, context-aware scene interpretation, and prediction of surgical phases and instrument usage. We develop intraoperative guidance systems for soft-tissue navigation and visualization that leverage multimodal sensor fusion and deformable registration. In shared autonomy for robotic surgery, we explore human–machine collaboration, adaptive control, and real-time feedback mechanisms to enhance surgical safety and efficiency. Complementary to these research lines, we investigate synthetic data generation, simulation environments, and proof-of-concept (POC) translational studies to validate our algorithms under realistic clinical conditions. 

Together, these efforts advance the long-term vision of an AI-augmented, data-centric operating room, where human expertise and computational intelligence converge to improve surgical outcomes. Organized into two synergistic research groups, we span the full perioperative pathway—linking preoperative patient data with intraoperative guidance and postoperative analysis.