| The single-port laparoscopic minimally invasive surgery robot is a teleoperation mode based on the human-computer interaction system.As the most important link in the system,the doctor’s operating skill level directly affects the operation effect of the entire robot system,and has a great impact on the quality and quality of the operation.Curative effect plays a decisive role.Therefore,how to effectively identify the skill level of robot operation and provide doctors with fair and objective theoretical guidance plays a very important role in researching and improving robot system control methods,helping medical staff improve operating skills,and ensuring standardization and safety of operations.The main work of this paper is as follows:Firstly,the single-hole laparoscopic minimally invasive surgical robot platform independently designed by the laboratory was introduced,the kinematics analysis of the end surgical instrument of the robot was carried out,the forward and reverse kinematics model of the end 6-DOF surgical instrument was deduced,and MATLAB software was used A simulation experiment is designed to verify the position and attitude of the proposed forward and inverse kinematics model,and it is obtained that the average maximum error of the manipulator movement is not less than 0.5% of the total length of the manipulator.Finally,the cloud image of the working space of the end manipulator is calculated by Monte Carlo method.The kinematics analysis of the end-effector arm of the single-port laparoscopic minimally invasive surgery robot is carried out to test the feasibility of the design and the accuracy and flexibility of the control.Secondly,based on the methods of literature analysis and theoretical research,the research on the surgical skills of single-port laparoscopic minimally invasive surgery robots was carried out.The kinematics evaluation index system constructs the robot-assisted surgery skill evaluation index system,which proves that the kinematics evaluation index system can accurately reflect the skill level of robot-assisted minimally invasive surgery.Thirdly,based on endoscopic vision technology,a more accurate and efficient automatic assessment method for robot-assisted surgical skills was studied,and an automated surgical skill assessment framework based on visual motion tracking technology that can be applied to real-time stage online assessment and feedback was proposed.Based on the Kernel Correlation Filter(KCF)tracking algorithm,this method extracts key motion signal features in video images,converts high-dimensional image information into low-dimensional motion features to reduce computational overhead,and maximizes the real-time performance of the result feedback is guaranteed.Finally,four deep neural network models,CNN,LSTM,CNN-LSTM and Res Net,were established to classify operational skill levels,and the performance of the models was compared and analyzed.The effectiveness of the proposed method was verified using the public and mature JIGSAWS dataset.The accuracy of the model was 83.36%,78.04%,79.88% and 84.80%.Efficient online automated assessment.This paper provides an effective scheme for automatic online assessment of target skills in single-port laparoscopic minimally invasive surgical robots. |
