Research On Coordination Registration Of Seed Implantation Robot Based On Ultrasound And Binocular Vision Fusion

Brachytherapy technology implants radioactive particles into tumor cells to treat cancer,which has the characteristics of less harm to patients,accurate curative effect,fewer postoperative complications,and low side effects,and is widely used in the treatment of clinical tumor diseases.However,the current clinical seed implantation surgery is to manually implant radioactive particles into the target area through ultrasound image navigation,which requires high experience of doctors and assistance from physicists,and also has low accuracy,poor efficiency,long planning time,and labor costs issues of strength.Because of the above problems of traditional surgery,with the continuous development of surgical robots,combined with precise medical images,the robot system dedicated to assisting seed implantation surgery is gradually moving towards clinical practice,but is limited by the narrow layout of the surgical scene and the space for multi-image fusion The registration process is complicated,and it is difficult to popularize and promote it in actual clinical practice.Focusing on the above clinical application issues,this paper conducts the following research on the spatial registration method of the ultrasound-guided robot-assisted seed implantation surgical system.This article first analyzes the implementation methods of seed implantation surgery commonly used in clinical practice,determines the requirements of the robot-assisted seed implantation surgery system,and then builds the robot system according to the actual needs.Aiming at the matching between robot workspace and surgery space,analyze the needle insertion space,surgery implementation space,robot workspace,robot flexible workspace,etc.within the organization to ensure that robots have wider compatibility with surgical scenarios in different hospitals.On this basis,an ultrasound-guided robot spatial registration method using an ultrasound probe stepper as a registration medium is proposed.According to the wide application of binocular vision to spatial information recognition and pose measurement,the robot system in this paper uses binocular positioning to locate the stepper and obtains the distance between the robot and the stepper through eye-on-hand hand-eye calibration.Relative positional relationship.In this paper,a binocular vision system mounted on a robot is constructed,and the work of binocular positioning,3D reconstruction and pose measurement matching is completed by using the MATLAB Calibration toolbox.On this basis,the structure of the stepper and marking points are designed,and the optical marking points on the stepper are recognized by the built binocular vision positioning system to complete the spatial positioning of the stepper.Combined with the pose calibration of the robot puncture needle at the end,the relative pose of the needle tip relative to the robot base is obtained,and the calibration accuracy is analyzed and calibrated.The coordinates of the stepper,robot,and puncture needle are unified through binocular vision,and the registration of the robot’s physical space is completed.In this paper,based on completing the robot’s physical coordination registration,a calibration method of a double-sided transrectal ultrasound probe based on ultrasound and binocular vision fusion is proposed,and an improved ZN doublesided ultrasound probe calibration model is designed and produced for this method.And deduced the calibration algorithm.Through the calibration of the probe,the coordinates between the ultrasound image,the ultrasound probe,and the stepper are unified and combined with the registration results in the physical space of the robot,the coordination between the stepper,the robot,the puncture needle,the stepper,the ultrasound probe,and the ultrasound image is realized.The coordinates are unified,and the image coordination registration of the robot is completed,forming a complete closed-loop registration chain.This paper builds a software system based on the Treatment Planning System(TPS)under the Windows platform.Through the research on 3D visualization and 2D enhanced image display,it reflects the relative positional relationship after spatial registration and integrates it into the software system.,to realize the integration of diagnosis and treatment.To verify the registration accuracy of the system built in this paper and the feasibility of seed implantation surgery,this paper built a complete experimental platform for robot-assisted seed implantation surgery under the guidance of ultrasound,completed the experiments related to robot physical coordination registration and image coordination registration,and carried out error analysis.analyze.After completing the robot coordination registration experiment,a silicone phantom was made to simulate the prostate tissue for puncture and seed implantation experiments,and the implantation error was analyzed and compared with the actual implantation position.The preset function has been realized,the effective conclusion has been obtained,and the expected effect has been achieved.