Method and device for automatically detecting vulnerable plaques based on cardiovascular OCT images

Abstract

The invention relates to a method and device for automatically detecting vulnerable plaques based on cardiovascular OCT images. The method includes acquiring a plurality of OCT images and establishingan OCT image data set; dividing the OCT image data set into an OCT training image set and an OCT test image set; using the OCT training image set to train a model to be trained, to form a post-training model; utilizing the post-training model to perform vulnerable plaque detection and vulnerable plaque classification on the OCT test image set; and prompting vulnerable plaques according to classification results. According to the method and device for automatically detecting vulnerable plaques based on the cardiovascular OCT images, the post-training model trained well is utilized to automatically extract features of the images, whether the vulnerable plaques are contained on the OCT images is automatically detected and recognized and the type of the vulnerable plaques is determined without manual intervention, so that a doctor can quickly and accurately analyze the OCT images. According to the method and device, detection can be performed in the complex OCT images, and the method anddevice are high in detection accuracy and have good robustness and detection speed.

Description

technical field [0001] The invention relates to the technical field of medical devices, in particular to an automatic detection method and device for vulnerable plaques based on cardiovascular OCT images. Background technique [0002] Vulnerable plaque refers to any plaque that is prone to thrombosis or can rapidly develop into a lesion. The capsule on the surface of the vulnerable plaque is very thin, and the plaque contains a lot of lipids, which is easy to rupture. For example, under the stimulation of emotional excitement, strenuous exercise, alcoholism, cold and other factors, the blood pressure of the human body will increase, the blood flow will violently impact the plaque, or the blood vessels will spasm. In these cases, the capsule of the vulnerable plaque will be damaged and ruptured. [0003] At present, there are many methods for vulnerable plaque detection, including coronary angiography, intravascular ultrasound, optical coherence tomography, etc., all of whic...

AI Technical Summary

Problems solved by technology

[0004] However, clinical evaluation and analysis of cardiovascular vulnerable plaque requires doctors to spend a lot of time and energy on analyzing and judging OCT images; on the other hand, it cannot meet the needs of real-time analysis in clinical practice, and cannot establish a unified clinical standard

And the current automatic identification of vulnerable plaque features is mainly based on two types of features: texture and elasticity. Using these two types of features to automatically identify vulnerable plaques has poor robustness and low detection speed.

Images