面向并联骨折手术机器人的复位轨迹自动式规划方法  

Automatic Planning Method of Reduction Trajectory for Parallel Fracture Surgery Robot

作　　者：李锦龙 刘传耙 孙涛[1,2] 张弢[3] 连宾宾 宋轶民[1,2] LI Jin-long;LIU Chuan-ba;SUN Tao;ZHANG Tao;LIAN Bin-bin;SONG Yi-min(School of Mechanical Engineering,Tianjin University,Tianjin 300350;Key Laboratory of Mechanism Theory and Equipment Design of Ministry of Education,Tianjin University,Tianjin 300350;Tianjin Hospital,Tianjin 300211)

机构地区：[1]天津大学机械工程学院,天津300350 [2]天津大学机构理论与装备设计教育部重点实验室,天津300350 [3]天津医院,天津300211

出　　处：《机械工程学报》2022年第5期26-33,共8页Journal of Mechanical Engineering

基　　金：国家重点研发计划(2018YFB1307800);国家自然科学基金(51775367);天津市科技计划(18PTLCSY00080,20201193,18YFSDZC00010)资助项目。

摘　　要：并联骨折手术机器人及其计算机辅助诊疗技术解决了骨折复位手术治疗创伤大、易二次感染的风险,但现有骨折复位轨迹规划方法存在碰撞检测效率低、复位路径长及肌肉牵拉力大等问题,难以实现精准安全的复位治疗。针对此问题,提出一种融合碰撞检测、肌肉力分析和最短路径搜索的骨折复位轨迹自动式规划方法。首先,基于患者的CT影像重建三维骨模型,以患者健侧骨模型为复位参考,获取骨折复位的目标。其次,定义碰撞检测阈值,基于八叉树搜索算法快速查找断骨间的最小距离点,实现断骨碰撞检测。随后,由OpenSim的标准模型通过缩放特征建立个性化骨折肌肉模型,实时获取复位过程的肌肉牵拉力。最后,以骨块无碰撞、肌肉牵拉力最小和路径最短为约束条件,设计A*算法的轨迹搜索节点与估价函数,实现轨迹规划。开展所述方法与交互式轨迹规划方法针对9组胫腓骨骨折的复位轨迹规划对比实验,本方法平均规划时间仅为3.2 s,比交互式方法耗时减小了两个数量级,平均复位路径长度为24.3 mm,减小了36.7%,平均肌肉力是96.4 N,减小了16.2%。结果表明,本方法轨迹规划耗时短、复位效率高且轨迹安全性高,为精准安全骨折复位治疗提供了新思路。Parallel fracture surgery robot and its computer-aided diagnosis and treatment technology solve the risk of trauma and secondary infection in fracture reduction surgery. However, the existing fracture reduction trajectory planning methods have some problems, such as low collision detection efficiency, long reduction path and excessive muscle tension, which make it difficult to achieve accurate and safe reduction treatment. To solve this problem, an automatic planning method is proposed, which combines collision detection, muscle force analysis and shortest path search. Firstly, three-dimensional bone model is reconstructed based on CT images of patients, and the target of fracture reduction is obtained by taking the healthy side bone model of patients as reduction reference. Secondly, the collision detection threshold is defined, and the closest pair of points between broken bones is quickly found based on the octree search algorithm, so as to realize the collision detection of broken bones. Then, a personalized fracture muscle model is established by the standard model of OpenSim through scaling features, and the muscle force in the reduction process is obtained in real time. Finally, the trajectory search node and evaluation function of A* algorithm are designed to realize trajectory planning with the constraints of no collision of bones, minimum muscle tension and shortest path. Compared with the interactive trajectory planning method, the reduction trajectory planning experiment of 9 groups of tibia and fibula fractures is carried out. The average planning time of this method is only 3.2s, which is two orders of magnitude less than that of the interactive method. The average reduction path length is 24.3 mm, which is 36.7%, and the average muscle force is 96.4 N, which is 16.2%. The results show that this method has short time-consuming trajectory planning, high reduction efficiency and high trajectory safety, which provides a new idea for accurate and safe fracture reduction treatment.

关 键 词：骨折复位 并联骨折手术机器人 复位轨迹规划 碰撞检测 肌肉力分析 

分 类 号：TH781[机械工程—仪器科学与技术]