Research On Rapid Enhancement Of Coronary Vessels And Plaque Detection Based On MSCT

Coronary heart disease is a common cardiovascular disease and one of the most lethal diseases.With the rapid increase of coronary heart disease death rate and the incidence of coronary heart disease in middle-aged and young people,it is an important and urgent task to actively promote the early diagnosis and prevention of coronary heart disease.In recent years,Multislice Computed Tomography(MSCT)technology has developed rapidly,providing new trends and research basis for the diagnosis and treatment of coronary heart disease with its non-invasive imaging and high temporal and spatial resolution.Computer-aided diagnostic techniques based on MSCT images have also evolved.In this paper,based on MSCT images,the research on coronary heart disease auxiliary diagnosis algorithm is carried out.The algorithm can enhance and extract the key information of MSCT images and automatically identify the plaques of coronary artery sections,and provide doctors with auxiliary diagnostic information to improve the efficiency and quality of doctors' reading.The main work of this paper is as follows:1.An algorithm for assisting diagnosis of coronary heart disease based on MSCT images was designed,including coronary artery preconditioning,coronary artery segmentation,and coronary artery plaque recognition.2.An image preprocessing algorithm based on Hessian matrix for vascular enhancement filtering is designed,which can improve the contrast of blood vessels in the image while suppressing image noise,and contribute to the quality of subsequent blood vessel segmentation.A coronary artery enhancement filtering algorithm based on the Routh–Hurwitz standard for improving the traditional preprocessing algorithm is designed,which can greatly reduce the computational time of the preprocessing algorithm while ensuring the accuracy of the algorithm.3.Automated extraction of coronary seed points was performed using an algorithm based on the Hough circle transformation and the dynamic Snake contour model.Using the obtained seed point as a source point,based on the gradation characteristics of the coronary vessels in the MSCT image,the vessel segmentation is realized by an adaptive growth algorithm based on the region of interest.4.The BP neural network based coronary plaque recognition method was designed.The image description feature vector including morphology,grayscale and texture was constructed.The BP neural network classifier was designed for the automatic detection of coronary plaque.5.The method of applying convolutional neural network to the identification of coronary artery plaque was proposed.The characteristics of the convolutional network were visually analyzed.The feature extraction of convolutional neural network from low to high level,from abstract to specific,was verified.The effects of two coronary plaque recognition methods proposed in this paper were compared.The experimental results show the effectiveness of convolutional neural network for coronary plaque recognition.