Flexible Needle Trajectory Control Based On Needle Recognition And Needle-Tip Pose Prediction

In the modern minimally invasive interventional surgery,the oblique flexible needle can use the interaction force with the tissue to puncture the arc track,bypass the nerve,blood vessel and other obstacles,and achieve the human body depth accurate puncture.Only by real-time feedback of flexible needle track path and needle tip pose can high-precision puncture be achieved.According to the above problem,this paper combines ultrasonic imaging equipment with flexible needle puncture control.This paper presents a flexible needle trajectory control method based on needle body recognition and needle tip pose prediction.The identification of the flexible needle body,the prediction of the needle tip pose and the control of the flexible needle trajectory based on real-time feedback are deeply studied.First of all,the centroid of the section of flexible needle is extracted based on ultrasonic image.According to the working principle of ultrasonic imaging equipment,the influencing factors of ultrasonic imaging are analyzed.The ultrasound image is reduced by Median filtering and Gaussian filtering by the characteristics of the ultrasound image.The ultrasonic image of the needle body was processed by MATLAB.It is using the ROI Interest Zone algorithm to extract the center mass of a needle.By converting the pixel coordinates into actual coordinates,we can obtain the centroid coordinates of the cross sectio n of the needle body.Secondly,this paper presents an algorithm of "fitting + interpolating" needle trajectory recognition.The centroid points set of needle body cross section in XOY plane was fitted to obtain the two-dimensional trajectory function.According to the trajectory function and z-axis coordinates,spatial interpolation is carried out to obtain the three-dimensional needle trajectory.According to the characteristics of the length of the needle and the track of the flexible needle,the position of the needle-tip is predicted.The trajectory and posture of the whole flexible needle in human tissue were determined.Thirdly,based on the motion characteristics of flexible needle and the prediction of needle-tip,a real-time dynamic path planning algorithm based on the current state is proposed.Analyze the basic characteristics of RRT algorithm,embed reachable guidance and greedy heuristic strategy.According to the feedback of ultrasonic imaging equipment,the flexible needle ultrasonic image track and the real-time location of the target are obtained.The improved Greedy-Heuristic Reachability-Guided Rapidly-exploring Random Trees(GHRG-RRTs)algorithm based on reachable guidance and greedy inspiration was used to plan the needle entry path in real time and determine the optimal puncture path,and control puncture trajectory.It improves the path search efficiency and puncture target accuracy.Finally,a flexible needle trajectory control experiment based on ultrasonic recognition is carried out.The feasibility was verified by error analysis.The experiment selects nickel-titanium flexible needles and bionic tissue materials.And combined with ultrasonic imaging equipment,the experimental platform was built.The ultrasonic image extraction trajectory of flexible needle was compared with the real one.By error analysis,we verified the feasibility of flexible needle trajectory based on ultrasonic recognition.This paper mainly studies the track control of flexible needle based on ultrasonic image recognition.A method of needle recognition and tip prediction based on ultrasonic image is proposed.Based on real-time state feedback information of the needle body,this paper adopts GHRG-RRTs algorithm to plan and control the trajectory of flexible needle puncture.The accuracy of ultrasonic needle recognition was verified by experiments.This paper provides theoretical and experimental basis for flexible needle trajectory control based on ultrasonic image.