Research On Segmentation And Measurement Of Coronary Artery In Angiocardiography Image

In recent years,coronary heart disease has become the world's biggest killer of human health,and has become one of the most important diseases in the world,due to its high incidence and high mortality rate.Therefore,early prevention and diagnosis of coronary heart disease is also particularly important.For coronary heart disease,how to help doctors quickly and accurately process medical images with the help of computer and image processing technology has always been the goal of scholars,and it also faces enormous challenges.Combined with the clinical needs for the diagnosis of coronary heart disease,this paper deeply studies the enhancement and segmentation algorithm of angiocardiography image.In this paper,pre-processing and improved multi-scale vascular enhancement techniques are used to enhance the image.The functions of coronary artery segmentation,skeleton extraction,vessel diameter measurement and stenosis degree assessment are designed and implemented.Finally,an efficient computer aided diagnosis system was developed,which realizes the analysis and playback of DICOM files,and can quickly and accurately segment and measure the coronary arteries to assist doctors to evaluate the severity of coronary lesions.The main work is as follows:1.Enhancement and segmentation of angiocardiography image.Due to the low resolution,high noise and uneven background brightness of the angiocardiography image,this paper improves the imperfection of vessel enhancement arithmetic based on Hessian matrix,constructs a new vessel function and solved the problem of incomplete extraction of vessel structures,especially capillaries.In this paper,an improved region growing method is used to segment and extract the coronary structure based on the pretreatment of angiocardiography image,morphological methods is added to fill small holes.The experimental results show that this method can effectively enhance the image and extract the complete coronary artery results,and have strong robustness.2.Auxiliary assessment of coronary artery stenosis.In this paper,the characteristics of the coronary arteries are used to extract key information such as centerline,vessel bifurcation points,and endpoints,the maximum inscribed circle algorithm is used to calculate the vessel diameter point by point.Due to the problems of large amount of calculation and poor adaptability in automatic extraction,this paper proposes an interactive assessment mechanism of coronary stenosis based on skeleton to avoid adaptive parameters that are difficult to meet various complex situations.The experimental results show that the method proposed is basically consistent with the results of expert evaluation,and can effectively assist doctors in clinical diagnosis.3.Implementation of coronary assisted evaluation system.This paper develops a coronary artery assisted evaluation system,and completes the integration,optimization and acceleration of the system.The system includes four types of functional modules:import of image data,angiocardiography image processing,interactive functional design,and diagnostic data reading.It is designed to implement interactive auxiliary evaluation of coronary vascular lesions and generate auxiliary diagnostic reports.This provides software technical support for clinical coronary stenosis quantification and coronary heart disease diagnosis.It has important clinical value and social significance.