A NEW LESS INVASIVE APPROACH TO KNEE SURGERY USING A VISION-GUIDED MANIPULATOR

We describe new methods for on-line image processing and calibration in the context of a vision-guided robotics system for orthopaedic knee surgery: During knee surgery, e.g. insertion of knee implants, bone material of the femur and the tibia has to be removed. Conventionally, surgeons use templates to guide a hand-held saw. Therefore accuracy is limited by the surgeon's skills and dexterity. The proposed system uses a saw mounted on a guiding device xed on a manipulator's hand. During surgery the hand is guided to preoperatively planned cutting planes. Hence the surgeon freely moves the saw to remove bony material, only limited by the guiding device keeping the movement within the cutting plane. This approach combines the robot's high accuracy with the surgeon's expertise and is therefore expected to provide safety, high overall accuracy and lower complication rates. Preoperative planning is achieved interactively, based on 3D surface-oriented bone models derived from computer tomography (CT). Intraoperative usage of the system starts with calibration and registration of an X-ray system. For this matter a new calibration technique has been developed. It uses two permanently recognizable 3D calibration bodies and special image processing and recognition algorithms. Registration is based on error minimization between estimated and actual bone contours in the X-ray image. This new method's advantages-compared to well-known techniques-are the achievement of highly accurate and robust calibration results, while constraints concerning the placement of landmarks relative to the X-ray system are avoided. Rigid patient immobilisation is not necessary, because small knee movements during surgery are allowed. This is due to the use of a real-time vision system for intraoperative bone tracking.