Study On Precise Puncture Method Of Liver Tumor Based On Interaction Mechanics

Percutaneous puncture is one of the most common minimally invasive surgical procedures.Due to the presence of the tip angle of the needle tip,the needle will naturally deflect when the needle is inserted into the patient’s soft tissue.Predicting the degree of tip curvature before surgery and adjusting the planned insertion trajectory accordingly is important for the surgeon to accurately insert the needle into the lesion.The traditional surgical method assumes th at the needle puncture trajectory is a straight line,but the accuracy is low.It is necessary to repeat the puncture operation to reach the target position,which inevitably causes greater damage to the patient.Predicting the bending trajectory of the puncture needle and planning the puncture path before surgery,this method can effectively compensate for the puncture error caused by the needle bending.In order to solve this problem,a mechanical-based puncture needle-soft tissue interaction model was established.U-net neural network was used to segment the CT image of the patient to obtain the target position,and the optimal trajectory was selected using optimization function,and finally the comparison experiment was carried out to verify the accuracy and effectiveness of the proposed algorithm.Considering the need to predict the bending trajector y of the needle before surgery,the subject proposed a method that does not rely on intraoperative feedback information,and an interactive mechanical model of the needle and soft tissue was established.The method was based on the material properties of the puncture needle,the geometric parameters and the material properties of the soft tissue,the error of the puncture needle at the multi-layer tissue interface was also modeled,and then these errors in the trajectory planning was compensated.Considering that the puncture needle should be safely inserted into the target while maintaining a safe distance from obstacles such as ribs and important blood vessel s.A multi-objective optimization method was proposed to determine the optimal needle insertion position and the puncture needle inclination angle.Considering the difficulty of accurate and rapid segmentation of liver tumors in CT images,a modified U-net neural network was used to segment the patient’s liver tumor image in two steps.The segmented tumor Dice value was about 0.86,which proved the improved UNet can effectively achieve automatic segmentation.The result of segmentation can show some information such as tumor size and position.The topic used gelatin with similar strength to human tissue as soft tissue substitute.A puncture experiment platform was built,and then multiple verification experiments were carried out,which proved the accuracy of the proposed algorithm.