Research On TMS Therapy Image Registration And Dynamic Calibration Technology

In recent years,psychiatric diseases in my country have been showing a high incidence year by year.Transcranial magnetic stimulation technology has been greatly popularized as a non-direct contact non-invasive treatment of mental diseases.At present,cranial magnetic stimulation treatment is implemented under the manual guidance of a doctor,and planning treatment and selecting treatment targets require the doctor to have rich clinical experience.If the work experience of the psychiatrist is not rich,you will encounter in magnetic stimulation treatment More difficult;besides,because the treatment time is longer,and the stimulation coil is heavier,even if the doctor’s work experience is very rich,it will be affected by fatigue factors,which will cause jitter,making the coil and the preset Deviation of the trajectory.In order to solve the problems encountered in current cranial magnetic stimulation therapy,it is necessary to develop an image-guided cranial magnetic stimulation therapy robot system.This paper mainly studies the key technologies in the image navigation robot system for transcranial magnetic stimulation therapy,Including the real-time registration technology of the patient’s head point cloud image and the head point cloud collected by the patient through external equipment during operation,and the rapid self-calibration technology of the robot system,and rapid self-calibration technology of the robot system.For the various problems in the process of magnetic stimulation therapy,The treatment with the aid of optical positioning system and mechanical arm is proposed,and the whole treatment process is designed.Optical navigation can track the patient’s treatment and provide corresponding assistance for the patient’s treatment to achieve precise navigation and treatment.Comparison of various registration algorithms through experiments,and finally using fast point feature histogram(FPFH)for point cloud for initial registration of head point cloud,and then using ICP algorithm for precise registration,and further improvement of ICP algorithm,Reducing the time complexity of the algorithm.The registration of the point cloud of the patient’s head obtained by the optical locator during the treatment and the point cloud of the three-dimensional head model before the treatment are registered,so that the position information of the patient’s head can be fed back to the treatment navigation system in real time.This is verified in the paper The feasibility of the method.The paper constructs the hand-eye relationship between the navigation coordinates of the optical therapy and the base coordinates of the robot,proposes a fast method of the self-calibration system based on the singular value decomposition method,and verifies the calibration experiment.The method of calibrating the system only needs to install optical mark points that can be searched and obtained by the optical locator on the stimulation coil at the end of the robot and make the robot arm automatically move 12 points to accurately and quickly get between the optical navigation coordinates and the robot base coordinates Transformation relationship.The research results show that the calibration error is less than 0.0003 m,and the system calibration problem can be solved effectively and quickly.The thesis establishes a transcranial magnetic stimulation robot system based on the UR5 robotic arm.The D-H parameter method is used to analyze the kinematic characteristics of the robot,which lays a solid foundation for the subsequent theoretical research.