Automatic Extraction Of Coronary Vascular Intima And Segmentation Of Plaque In Optical Coherence Tomography

With the aging of society and the acceleration of urbanization,cardiovascular disease has become the disease with the highest morbidity and mortality in China,and it has become a major public health problem.At present,the auxiliary diagnostic imaging techniques for cardiovascular diseases include CT,intravascular ultrasound(IVUS)and optical coherence tomography(OCT).Due to the improved imaging speed and resolution of OCT systems,OCT technology has been widely used in the medical field.However,the use of OCT imaging to diagnose coronary artery disease still requires manual operation by doctors,which is time-consuming and laborious.Therefore,the computer automatic analysis algorithm based on coronary OCT image has important clinical significance for assisting doctors in the diagnosis and treatment of coronary artery disease.This paper has done the following research work on coronary OCT imaging:In the aspect of coronary endarterial extraction,an automatic extraction algorithm based on improved CV for coronary endocardial contour sequences was proposed.By improving the CV model evolution weight function,the curve local information is introduced to control the boundary evolution speed of the curve,and the gradient energy term and the shape constraint term are added to increase the information point capturing ability and the previous frame boundary information as a priori condition to further constrain the evolution contour shape.Finally,the sequence extraction of the coronary OCT intimal contour was achieved,and the intimal extraction efficiency was further improved by setting the initial evolution profile to the intimal extraction result of the front panel for the image with less difference in the intimal contour of the sequence.The results show that the intimal extraction algorithm not only improves the extraction quality of the endometrium,but also has a good extraction effect on special conditions such as fuzzy boundary,guide wire shadow,stent and plaque interference,and lays a foundation for subsequent plaque segmentation.In terms of coronary plaque segmentation,this paper proposes an improved kernel map cutting algorithm to achieve accurate segmentation of plaque regions.According to theobtained intimal contour and OCT light penetration characteristics,the intimal data points are expanded by the image rectangular coordinate and polar coordinate conversion method to obtain the region of interest covering various types of plaques,thereby reducing peripheral noise interference.,reducing the area of the image area to be processed later.The K-means clustering method is used to obtain the cluster center point of the region of interest.The cluster center point is used as a parameter input to improve the kernel map cutting algorithm,and the area ratio of the region of interest to the original image region is introduced into the energy function,and the smoothness of the segmentation boundary is controlled to finally achieve fibrosis,calcification and lipid plaque.Precise segmentation of the area.Experiments show that the proposed method can effectively improve the segmentation accuracy of various types of plaque regions,and accelerate the automation process of computer-aided analysis of coronary OCT images to some extent.