Research On Space Registration Technology Of Operating Neurosurgery Robot

Neurosurgery craniotomy is a necessary method to treat patients with severe traumatic brain trauma,and it is also an effective method to treat brain tumors,cerebral hemorrhage,epilepsy and other diseases.According to statistics,the annual operation volume of Beijing Tiantan Hospital reaches about 12,000 units per year.However,the craniocerebral region involves a large number of important physiological tissues such as the central nervous system and venous sinuses.Traditional craniotomy has clinical problems such as high risk of important tissue damage,strong doctor operation,and heavy dependence on doctors’ experience and skills.Due to the limitation of medical conditions,craniotomy cannot be performed,and patients often have difficulty getting timely and effective treatment.Therefore,it is of great practical significance and clinical value to study surgical robots for neurosurgery.This paper takes the robotization of craniotomy as the starting point,and sets up a neurosurgical robot experimental platform with cranial drilling and milling operations around the craniotomy.The spatial registration technology,robot kinematics analysis and trajectory planning,and surgical mechanical object detection are mainly studied.The main contents and work of this thesis are as follows:Firstly,using traditional craniotomy as a reference and combining with the doctor’s experience,an experimental platform of operating neurosurgery robot was set up,and the layout of operating room and the corresponding surgical procedures for neurosurgery were introduced in detail.Secondly,according to the design requirements for high accuracy and stability of neurosurgical robot positioning,a spatial registration method combining SAC-IA and ICP is proposed,and the SAC-IA algorithm based on FPFH features is used to achieve initial registration and ICP algorithm is utilized to complete accurate registration.Simulation experiments were completed through VS and PCL database,and the results prove the effectiveness and feasibility of the proposed method.Thirdly,aiming at the surgical area planning functions required for surgical robots,taking JAKA ZU 7 robot as an example,the forward kinematics model of the surgical robot is established based on the D-H method,which is verified by simulation experiments using the MATLAB toolbox,and the results show that the forward kinematics model is correct.Based on the quintic polynomial interpolation algorithm,the trajectory planning in joint space is performed and verified by simulation experiment.Finally,aiming at the problem of single way of sensing the position information of the surgical instruments,in order to enhance doctors’ perception of real-time operation status of surgical instruments at the end of surgical robots,the object detection and location of surgical instruments are realized based on SSD algorithm,which lays a foundation for forming multi-source information feedback channel at the end of doctors.The feasibility of this method is verified by using dataset that doctors operate Da Vinci surgical robot.