Research On The Optimization Of Preoperative Placement Of Minimally Invasive Surgical Robot Based On Augmented Reality

Robotic-assisted minimally invasive surgery(RMIS)has developed quickly due to its advantages of small surgical trauma,low blood loss and rapid postoperative recovery.However,the surgical robot structure is complex.The robotic arm has redundant degrees of freedom.At the same time the robot's workspace is small.These problems cause trouble to preoperative placement,which has an influence on the feasibility of surgery and safety.Therefore,this thesis intends to solve the key problem in preoperative placement of RMIS,especially for minimally invasive surgery in narrow stenosis.Combined with multiple performance indexes,a method to optimize the preoperative placement is proposed in the thesis based on the optimal physical characteristics and physical parameters.Augmented Reality(AR)technique is also introduced to study the preoperative placement of surgical robots.This thesis analyzes the importance of clinical preoperative placement.First,a workspace model based on patient's body surface and lesion parameters is established.Then,the kinematic model of the robot is established according to the relationship between the location of the incision and the coincidence of the manipulator.The robot's joint velocity of the Jacobian Matrix is derived to prepare for the optimal preoperative placement of the robot.In this thesis,the optimal target model of surgical robots is established from the aspects of the performance of the single arm motion,multi arm coordination,surgical instruments and low collision rate of robotic arms.The multi-objective differential evolution algorithm based on Gaussian process prediction is used to optimize the preoperative placement of the robot.In addition,this theis presents an algorithm to solve the problem of arm collision during the operation.The Oriented Bounding Box(OBB)model of each arm component is established.Based on the OBB intersecting test,the maximum motion range of each joint is calculated,and the concept of collision index is proposed to limit the movement of each joint in order to avoid the collision between robotic arms.The real environment is built to assist in the preoperative placement of the robot and the simulated operation.The positional relationship between the virtual patient model and the actual patient is established.The final position of AR assisted puncture can be obtained.In addition,this thesis attempts to solve the interaction problem between virtual and real models in AR environment,including detecting collision between real surgical instruments and virtual soft tissue or virtual bone tissue.When the deformation occurs in the soft tissue,the main hand will produce the corresponding force feedback.The immersive sense of AR environment is also improved.This thesis validates the above research contents.First,the validity of multi-objective optimization algorithm based on Gaussian process is proved by data comparison.Secondly,the three different models of patients are built to complete the preoperative placement optimization.The performance parameters of the optimized results are verified by the real surgical robots.The effectiveness of the real-time collision avoidance algorithm is verified based on the virtual surgical robots.Then the relationship between the feedback force of the surgical instrument and the deformation of the soft tissues is verified by establishing AR environment.The location of the surgical incision can be obtained based on the augment reality,which validates the feasibility of AR technique in preoperative placement.