Robotic solution for orthopedic surgery

摘要

In orthopedic surgery, the accuracy of internal fixation is related to the success of the procedure. In the past, orthopedic surgeons usually placed screws freehand according to anatomical landmarks or fluoroscopy images. Although there are many surgical techniques for inserting screws freehand, the accuracy of freehand screw placement is not high enough, and the results can be unstable. The accuracy of freehand screw placement reported in the literature varies greatly. Particularly in the spine, with its unique anatomy and important adjacent tissues, inaccurate screw placement may lead to serious complications, such as nerve or vessel injury, cerebrospinal fluid leakage, thoracic or abdominal organ injury, and segment instability. Surgeons must undergo extensive professional training and accumulate in-depth surgical experience to reduce the incidence of inaccurate screw placement. Screw fixation usually requires percutaneous access to the bone, followed by insertion of catheters, guidewires, and screw cannulas. Thanks to developments in X-ray and computed tomography (CT) technology, surgeons can delicately plan the path of percutaneous access using personalized preoperative images and three-dimensional (3D) reconstruction images, avoid important tissue features, such as nerves and blood vessels, and select the most appropriate percutaneous entrance and access angle. However, it is difficult to translate well-designed percutaneous access on the image to the actual patient. Additionally, in the absence of a guidance device, percutaneous access established by manual operation will inevitably deviate from the planned path. The limitations of human hands (manipulation) and eyes (positioning) have become increasingly apparent. The limitations of orthopedic surgery are that surgeons cannot see deep structures and cannot directly transfer the preoperative plan to the operation. Additionally, surgeons' hand stability may be insufficient in some sophisticated procedures. Because the clinical use of orthopedic surgical robots requires precise and stable positioning, soft tissue surgical robots, such as the Da Vinci system (Intuitive Surgical, Inc., Sunnyvale, CA, USA), are unsuitable for use as orthopedic surgical robots. To improve the accuracy of orthopedic surgery, clinicians and researchers have jointly developed a variety of robotic systems. [1] Many robotic systems for orthopedic surgery, such as Renaissance, ROSA Spine, and MAKO, have been used clinically. Robot-assisted orthopedic surgery can improve the accuracy of implant placement and promote computer-assisted minimally invasive surgery; [1] however, these systems also have inherent limitations. Research on orthopedic surgical robots has focused on using robots for a single indication, but not for multiple indications. Additionally, most robots are registered by preoperative CT or intraoperative two-dimensional (2D) fluoroscopy images, but not by intraoperative 3D reconstruction images. During surgery, it is difficult to translate the percutaneous access designed on the image to the patient without intraoperative 3D reconstruction images. Therefore, our work was focused on developing a solution for robot-assisted orthopedic surgery and creating an orthopedic surgical robot system for multiple indications. Here, we describe how our key technologies are applied to our orthopedic surgical robot. Our goal was to build a robotic solution to provide real-time positioning and posture guidance to help surgeons establish percutaneous screw trajectories. The robot determines the planned percutaneous trajectory, places the guide device on the corresponding position on the patient, and holds the device steady. Key operations, such as puncturing, drilling, and screw placement, are manually completed by the surgeon under the guidance of the device. Similar studies have evaluated surgical positioning systems; however, the clinical practicability and versatility of the system are important f