Study On Path Planning For Robot-Assisted Flexible Needle Insertion Sys-Tem

Needle insertion is a common minimally invasive surgery that is clinically used for biopsy,local anesthesia and brachytherapy.Compared to a short and rigid needle,flexible needle can better avoid sensitive areas(such as blood vessels,nerves,etc.)and cause less damage to the patient.Needle deformation and tissue deformation in the insertion process greatly increase the difficulty of needle insertion path planning.Robot-assisted flexible needle insertion has good stability,accuracy and safety,and has good application prospects.Its path planning has a very high research value and practical significance.In this paper,the deformation mechanism of the needle and the tissue during needle insertion process is studied in detail.The needle-tissue interaction model is established and the needle insertion path planning is taken into account the deformation of the needle and the tissue.First of all,in order to ensure the accuracy of path planning,the needle-tissue interaction model is established to simulate the deformation of needle and tissue during needle insertion pro-cess.The needle model is based on the vector form intrinsic finite element method,compared with the ANSYS calculation results,the deformation accuracy is within 0.5mm,while the tissue deformation model is base on finite element method.The needle-tissue interaction force model effectively combines the needle and the tissue.Through the analysis of the insertion force obtained by experiments,the insertion process is divided into four stages,and the parameters of the force model are calculated.The needle-tissue interaction model is programmed in Matlab.Genetic simulated annealing algorithm plans the initial path without considering the deforma-tion of the tissue in the insertion space.The needle-tissue interaction model and iterative learning control are used to continually adjust the deviation of the needle tip and target until the deviation is within acceptable limits.The needle insertion path planning method proposed in this paper is modular,and the algorithm of each module can be flexibly replaced,as long as the new module conforms to the format defined in this article.In order to verify the accuracy and feasibility of the needle-tissue interaction model and the path planning method,we set up some needle insertion experiments.The experimental results show that when the insertion depth is 80mm,the maximum error of non-rotating insertion exper-iment and model is 0.4736mm,the maximum error of rotating puncture is 0.6279mm.20 times path planning and insertion experiments were performed on the two typical insertion environments respectively.The average error between needle tip and target was 0.4586mm and 0.3517mm,the maximum error was 0.8555mm and 0.7947mm(less than 1mm),and the needle trajectory did not collide with obstacles.In this paper,many parameters are involved in the model.The PAWN method is used to an-alyze the sensitivity of parameters,including needle model parameters,path planning parameters and iterative learning parameters.The parameters with high sensitivity are optimized.Finally,the paper summarizes the contribution of this paper to clarify the contribution points and shortcomings,and prospects the future work of robot-assisted flexible needle path planning.