Coronary atherosclerotic heart disease is one of the main causes of death from cardiovascular diseases. Early detection of atherosclerotic lesions can help clinicians understand the condition of cardiovascular patients and provide reference for better treatment measures. Compared with other detection technologies, intravascular optical coherence tomography (IVOCT) has the advantages of no radiation, high resolution, and high imaging speed. Therefore, it plays an important role in the detection and evaluation of atherosclerotic plaque. Although IVOCT has been widely used in the detection of plaque in coronary vessels, the imaging system could not directly provide effective plaque feature identification information, and clinicians can only judge the characteristics of plaque according to their own experience. Based on a brief introduction of the application of IVOCT in detecting coronary atherosclerotic plaque, this paper introduces the method of eliminating the vascular motion artifact caused by cardiac pulsation. The automatic segmentation and classification of IVOCT images are studied by using machine learning method. And the plaque features of calcified plaque, lipid plaque and fibrous plaque in IVOCT images are extracted. The deep learning algorithm is used to analyze the characteristics of vulnerable plaque and put forward quantitative evaluation indicators. It is very important to develop the intelligent recognition system of IVOCT in plaque type, that provide objective, intuitive and accurate plaque classification marks, display and rupture risk assessment for the clinic. So that clinicians can get rid of the current situation of relying solely on experience for lesion recognition, and save patients' lives in time.

No abstract