Image Segmentation And Three-Dimensional Reconstruction Of The Left Atrium For Blocking Surgery Robot

Atrial fibrillation is one of the most common arrhythmic conditions,and according to statistics,the prevalence of atrial fibrillation has reached 2-4% in adults worldwide.Analysis and study of the patient’s left atrial structure after segmentation and threedimensional reconstruction from cardiac magnetic resonance images is an important basis for diagnosis and treatment of atrial fibrillation using robotic surgery for left atrial appendage closure.To address the need for computer-supported robotic surgical assisted procedures for left atrial appendage closure,this paper focuses on the following research work on cardiac MRI image segmentation methods and three-dimensional reconstructive visualization methods of the left atrium.1.An image segmentation method based on sequence relationship learning and multi-scale feature fusion is proposed for 3D to 2D sequence conversion in cardiac magnetic resonance images and the varying scales of left atrial structures within different slices.Firstly,a convolutional neural network layer with an attention module was designed to extract and fuse contextual information at different scales in the image,to strengthen the target features using the correlation between features in different regions within the image,and to improve the network’s ability to distinguish the left atrial structure.Secondly,a recurrent neural network layer oriented to two-dimensional images was designed to capture the correlation of left atrial structures in adjacent slices by simulating the continuous relationship between sequential image slices.Finally,a combined loss function was constructed to reduce the effect of positive and negative sample imbalance and improve model stability.The improved segmentation effect of the proposed method relative to other methods was verified in left atrial segmentation experiments based on the LASC2013 and ASC2018 datasets.2.A three-dimensional reconstruction visualization method based on an improved marching cubes(MC)algorithm is proposed to address the needs of medical personnel to observe the relative position of the left atrium and other tissue structures during surgery and to select a suitable blocker according to the morphology of the patient’s left atrial appendage.Firstly,a three-dimensional reconstruction method based on surface rendering is designed to reconstruct the left atrium structure with high efficiency and accuracy to meet the needs of the surgeon for real-time observation and selection of the left atrial appendage blocker during robotic surgery.Second,an improved MC algorithm is designed to meet the requirements for the morphology and accuracy of the left atrium in robotic surgery.The linear interpolation is used to calculate the coordinates of the equivalence points and normal vectors,and the principle of proximity is used to move the equivalence points to the nearest vertices to reduce the number of facets and topological complexity,which not only reduces the computation time and avoids repeated calculations,but also forms a smoother contour surface that can meet the accuracy of the left atrium morphology.Finally,based on the three-dimensional reconstruction visualization of the left atrium,a three-dimensional reconstruction visualization interactive system is designed,using a visualization library to realize interactive operations on the reconstructed object model,including rotation,scaling and panning functions.Based on the actual background of robotic surgery for left atrial appendage closure,a prototype system for robotic image visualization of left atrial appendage closure surgery is designed,and functional modules such as cardiac magnetic resonance image segmentation and three-dimensional reconstruction visualization of left atrium are developed to provide an effective tool and platform for improving the planning efficiency and safety of robotic surgery for left atrial appendage closure.