Control Methods And Experiment Of Flexible Needle Puncture In The Soft Tissue

With the rapid development of economy and society,people's living standards gradually increased,thus modern medical technology has also made considerable progress.As a modern medical technology,paracentesis has also been taken seriously.Under the guidance of the imaging and other sensing methods,paracentesis is to put needle into the lesion to complete the living body sampling,local anesthesia,brachytherapy and other operations.At present,most of needles used in puncture surgeries are rigid.The rigid needle's controllability is poor and may cause trauma on the blood vessels and nerves.To overcome disadvantages of the rigid needle,the flexible needle is proposed which is made of nickel-titanium alloy.Due to the flexible needle body and small needle diameter,the surgery incision is reduced,the discomfort of the operation decreases and the recovery after operation is accelerated.What's more,the flexible needle can be driven by controlling the needle tail,to avoid the sensitive areas such as blood vessels and nerves.Compared with the rigid needle,flexible needle has above advantages,so the puncture control of the flexible needle is studied in this thesis.Firstly,we analyze the 6-DOF model of the flexible needle,and get the plane nonlinear model by dimension reduction,and then get the plane linear model based on the feedback linearization theory.Since the flexible needle's diameter is small,cameras used to locate the needle tip can acquire the tip position but cannot get the attitude.Therefore,Kalman filter is used to estimate the attitude of the needle tip.The linear quadratic optimal control method is used to design a closed-loop control system to keep the needle tip on the desired plane.In this method,the system noise and the measurement noise are regarded as Gaussian white noise with unknown characteristics,which lead to inaccurate results.To solve this problem,we estimate the measurement noise and the adaptive Kalman filter algorithm is employed to estimate the needle tip attitude.Simulation results show that,compared to the former method,the latter is indeed improved.Secondly,the previous process uses feedback linearization which ignores the model of nonlinear factors and reduces the control accuracy.Therefore,unscented Kalman filter is used to estimate the attitude of the needle tip for the plane nonlinear model.The same control strategy as above is used to keep the needle tip on the desired plane.The estimation of the measurement noise is also considered,and adaptive unscented Kalman filter is used toestimate the state of needle tip when the character of the measurement noise is unknown.Simulation and experimental results show that the two methods are feasible.Thirdly,different puncture curvature can be realized by controlling the rotation angle of the tail and the insert length of the needle.According to the characteristics of flexible needles,a three-dimensional control method of needle tip is proposed.The feasibility and effectiveness of the three-dimensional control method are verified by simulations and experiments.Finally,all parts of the experimental platform are introduced in detail.The localization method of the needle tip is described.The characteristics of the flexible needle puncture are studied.Filter methods and control methods are verified experimentally.To sum up,the estimation methods of the flexible needle tip state,and the algorithms to control the needle tip to puncture on the desired plane and three-dimensional space are studies in this thesis.