Research On The Method Of Three-dimensional Reconstruction Of Coronary Arteries And Its Hemodynamic Analysis

Coronary atherosclerotic heart disease,referred to as coronary heart disease,is one of the most dangerous cardiovascular diseases in recent years.It is a myocardial ischemic heart disease caused by myocardial ischemia,hypoxia or necrosis due to stenosis or obstruction of the vascular lumen,which threatens people’s lives and health.In order to diagnose the degree of coronary artery stenosis accurately and quickly,assist the doctor to develop a reliable treatment plan for the patient.The calculation of the noninvasive fractional flow reserve derived from coronary computed tomography angiography(CTA)has received extensive attention.This thesis proposes a method to study the noninvasive fractional flow reserve derived from coronary CT angiography.Firstly,the region of interest of coronary is identified and extracted by deep convolutional neural network.Then an algorithm combines the improved Ostu threshold and the connected region is used to segment the coronary from CT angiography.A three-dimensional model of coronary artery was obtained by three-dimensional reconstruction of blood vessels,and hemodynamic analysis was performed based on this model.Finally,the method of calculating the noninvasive fractional flow reserve derived from coronary CT angiography is obtained by analyzing the results.In this thesis,combined with deep learning and traditional segmentation methods to accurately segment coronary vessels from CT angiography,hemodynamic analysis of blood vessels by computed fluid dynamics,is of great significance for the diagnosis and treatment of coronary artery stenosis.The main research work is as follows:First,the features of coronary artery was extracted through the deep convolutional neural network to train a model of coronary artery from the CT angiography image data.This model is used to detect regions of interest in the coronary artery.Second,According to the characteristic information of blood vessel shape and gray value in the region of coronary artery interest,the Otsu algorithm combined with the connected region algorithm is used for the segmentation of coronary artery region for blood vessel segmentation,which achieves high segmentation accuracy and good robustness.Three-dimensional reconstruction of coronary artery using volume rendering method based on ray casting algorithm.Thirdly,The hemodynamic model was used to analyze the coronary artery three-dimensional model to study the calculation method of the noninvasive fractional flow reserve derived from coronary CT angiography.In this thesis,an algorithm combines the improved Ostu threshold and the connected region is used to segment the coronary from CT angiography.It is proved by experiments that the accuracy of the segmentation algorithm meets the requirements of the actual index,and the reconstructed three-dimensional model of coronary artery is ideal.The hemodynamics of the geometric vessel model and the real coronary three-dimensional model were studied by computational fluid dynamics,and the noninvasive fractional flow reserve derived from coronary CT angiography was calculated.It provides assistance for the early noninvasive diagnosis of coronary artery is of great significance for doctors to develop treatment plans and save more lives.