动态生理运动环境下脊柱手术辅助机器人术中感知方法研究

Spinal surgery is one of the high-risk surgical operations, according to the report the cumulative complication rate is as high as 50% above, so it is interesting line of research on using robot assisted surgery to achieve high stability, high safety and high precision of operations. In the previous experiments on in vivo animal we found that the slight deviation of position, rotation or expansion caused by the physiological movement will affect not only positioning and operation but also the recognition of intraoperative signal perception and operation state. Therefore it is highly required to carry out the kind of fundamental study of fine state perception mechanisms in dynamic conditions. In this study, based on the research work conducted before, focusing on the key problems of patients affected by coupling movement environment, The coupling motion modeling for operation area of the patient under the dynamic environment is proposed; the motion and perception of synchro-compensation and correction method of the surgical robot in the dynamic environment are set up, robot motion compensation and sensing signal correction on operation area under dynamic environment, to explore the relationship of more parameters with multi-info features in the operation and the anatomical structure, the operation state recognition method of robot based on multiple information fusion are put forward. These studies will expand the stable working ability of the doctor, and assist the doctor to implement clinical operations ensuring the operation safety and improving the operation quality. This research will expand the fine perception theory of the spinal surgery robot and has important theoretical and practical significance on promoting the clinical practical development of spinal operation robots.

脊柱手术是高风险外科手术之一,据报道累计并发症高达50%以上，采用手术机器人可辅助医生实现高稳定、高安全和高精准的手术。在前期活体动物实验研究发现，临床条件下患者受术区移位、旋转或扩张等耦合运动,不仅影响脊柱手术机器人的术中定位和操作，也影响手术状态的识别,因此亟需对术中动态条件下手术机器人精细状态感知理论和方法进行深入研究。本项目在前期研究基础上，重点针对患者生理呼吸运动对受术区的影响，提出患者脊柱受术区耦合运动模型;构建动态环境下手术机器人运动与感知的同步补偿方法;探索术中多信息特征参数与受术组织解剖结构之间相互关系,提出基于多信息融合的机器人手术状态识别方法。这些研究将拓展医生的稳定工作能力，辅助医生实施临床手术，确保手术安全，提高手术质量。本项目研究对发展脊柱手术机器人精细感知理论，推动脊柱手术机器人临床实用化具有重要意义。

本项目对机器人辅助脊柱椎弓根置钉、椎板减压手术技术进行了系统性研究，结合数学建模仿真、信息融合等技术，重点解决临床中医学图像规划、动态环境补偿、手术状态感知等问题。首先采集了生理呼吸信号，分析了深呼吸、平静呼吸模式对手术精度的影响，得出较小潮气量有利于提高手术安全性结论。并分别对脊柱在受负载过程中会发生扭转和弯曲进行分析、建模与耦合，得到了脊柱受力时的负载、角度及位移之间的数学关系模型。其次通过机器人只补偿呼吸运动和机器人同时补偿脊柱受力变形和呼吸运动的对比分析，验证出同时补偿脊柱受力变形和呼吸运动的情况下的力信号稳定性较高，有利于提高手术的安全性和精确性。面向机器人辅助手术中的手术状态感知问题，本项目提出了基于图像、声音、力信号的手术状态感知方法。建立了导航图像动态感受域模型，提高了对骨边缘感知的灵敏度，实现了主从操作模式下的低延迟风险预警；提出了基于声音和力信号特征、决策融合的更加精细的状态感知算法，提高了手术安全性。最终通过开发与集成机器人系统，实现基于动物实验、样本骨实验、系统仿真等方式的验证。通过动物实验验证了临床可操作性；通过脊柱模型有限元仿真验证了器械-脊柱交互作用模型与参数在线辨识方法的有效性，系统仿真分析了不同因素对系统跟踪精度的影响；通过样本骨实验统计了单一力特征、力特征融合、声音特征融合、决策融合感知方法的感知率，提高了手术安全性。