Integrating modelling and knowledge into computer-aided orthopaedic surgery

拟议的研究计划的目标是开发新的计算机辅助骨科手术系统。这种手术系统的部件可以包括使用计算机辅助的术前规划、术中导航和术后评估。在实施术前规划阶段的系统中，处理医学图像以产生解剖结构的计算机模型。可以虚拟模拟外科手术以创建手术的优化计划，并且可以设计患者特定的仪器。实现术中导航的系统可以依赖于术中成像、定位解剖结构和手术器械的跟踪系统、执行手术的机器人设备或患者特定的仪器。实施术后评估的系统可以处理医学图像、术中跟踪数据和各种结果测量，以将患者结果与手术技术相关联。在拟议的研究计划的方法的一个主要主题是纳入知识和建模手术系统。解剖学知识可以用于实现医学图像中解剖结构的自动识别、定位和推断。特别令人感兴趣的是，这种表示的效用建模畸形的严重程度从轻度（骨关节炎关节侵蚀）到严重（多骨折）。拟议研究的新奇将是创建计算机辅助手术系统，通过知识集成实现高度自动化。当前一代的计算机辅助骨科手术系统依赖于训练有素的人员来分割医学图像、创建手术计划以及设计和制造患者特定的仪器;自动化这些任务将潜在地降低阻碍计算机辅助手术方法广泛采用的成本和技术障碍。一个系统，可以自动识别，本地化，并推断有关变形解剖的临床有意义的信息将是一个新的工程贡献的最先进的医学图像计算。

The goal of the proposed research program is the development of new computer-aided orthopaedic surgical systems. The components of such surgical systems can include pre-operative planning, intra-operative navigation, and post-operative evaluation using computer assistance. In systems implementing a pre-operative planning phase, medical images are processed to produce computer models of the anatomy. The surgical procedure can be virtually simulated to create an optimized plan of the procedure, and patient-specific instrumentation can be designed. Systems implementing intra-operative navigation may rely on intra-operative imaging, tracking systems to localize anatomy and surgical instruments, robotic devices to perform surgery, or patient-specific instrumentation. Systems implementing post-operative evaluation may process medical images, intra-operative tracking data, and various outcome measures to relate patient outcome to surgical technique. A major theme in the approach of the proposed research program is the incorporation of knowledge and modelling in surgical systems. Knowledge of anatomy can be used to achieve automated recognition, localization, and inference of anatomy in medical images. Of particular interest is the utility of such representations for modelling deformities ranging in severity from mild (joint erosion in osteoarthritis) to severe (multi-fragmentary fracture). The novelty of the proposed research will be the creation of computer-aided surgical systems with a high degree of automation enabled by knowledge integration. The current generation of computer-aided orthopaedic surgical systems rely on trained personnel to segment medical images, create surgical plans, and design and fabricate patient-specific instrumentation; automating these tasks will potentially reduce the costs and technological barriers discouraging the widespread adoption of computer-aided surgical methods. A system that can automatically recognize, localize, and infer clinically meaningful information about deformed anatomy would be a novel engineering contribution to the state of the art in medical image computing.