Research On Pose Of Guiding Robot For Lumbar Pedicle Screw Minimally Invasive Surgery

The accurate insertion of pedicle screw is an important part for Lumbar pedicle screw minimally invasive surgery.Nowadays doctors usually insert the pedicle screw by hands with the help of X-ray.Manytime shootings to correct screw poses make surgery time too long and doctors and patients hurt by X-ray radiation.In addition,the doctors' fatigue and tremble cause the deviation between the actual pose and the ideal pose of the screw inserted.Therefore,it is of great theoretical significance and clinical value to develop a guiding robot for lumbar pedicle screw minimally invasive surgery and study its key technologies.Firstly,this paper analyzed spinal surgery robots at home and abroad deeply,combined with doctors' actual demands and surgery characteristics,determined the functional requirements and performance indices of the guiding robot.Then this paper proposed a surgery robot in which a Stewart platform was in series with a translational degree of freedom before and after itself,and planned the surgery process with the robot.Secondly,this paper analyzed the pose and workspace of the surgery robot.The pose transformations inside the robot were computed,the inverse kinematics and forward kinematics were computed based on the Stewart platform.Then this paper computed the attitude workspace and position workspace using the Monte Carlo method.All of the computations laid the foundation for pose computation and motion control.They also provided the basis for the selection and design of drive systems.Thirdly,the motion performance,the bearing capacity and the rigidity indices were selected as optimization objective functions,the hinge distribution and the initial length were selected as optimization variables,the surgical posture range and the structure size were selected as constraints.This paper used the genetic algorithm to carry out the single objective optimization and after optimization three performances increased by 110%,111.62%,539.4%.This paper used the NSGA-? algorithm to carry out the multi-objective optimization,which determined the best structure parameters of Stewart platform.The mechanical structure and control system were designed in detail based on the optimization results.Forthly,two kinds of attitude error model were built based on inverse kinematics and forward kinematics.This paper used Monte Carlo method to analyse the attitude error probability distribution and found all attitude angle error obeyed normal distribution and the range of ? error was bigger.Sensitivity analysis was carried out and computation results showed that the sensitivity indices of overall attitude errors had similarity and symmetry.The methods and results can provide basis for accuracy design of Stewart platform.The calibration model was built based on error models,which could identify 36 parameters combined by 42 error variables in a particular linear relationship.To simplify error compensation steps,set 6 error variables as 0,which the zero diagonal elements of the R matrix of observability matrix corresponded.The simulation results indicated that errors after compensation were almost eliminated.The particle swarm optimization algorithm was used to choose calibration poses to solve the problem of inaccurate identification results under random measurement noise,which improved the identification and compensation results.Lastly,to verify calibration methods,a two-axis inclinometer was used to carry out the attitude error calibration experiments.The experiments results indicated most attitude errors were reduced greatly after compensation.The calibration method in this paper is simple to operate,low cost and effective.The research results in this paper can lay a foundation for the industrialization of surgery robots and provide a new thought for the development and clinical application of surgery robots.