Method for intelligently segmenting intracranial plaques and carotid plaques of magnetic resonance image

Abstract

The invention discloses a method for intelligently segmenting the intracranial plaques and the carotid plaques of a magnetic resonance image, which comprises the following steps of S1, acquiring the blood vessel wall magnetic resonance image of a user; S2, preprocessing the blood vessel wall magnetic resonance image to obtain a preprocessed image; S3, segmenting a preset plaque region in the preprocessed image by a pre-trained convolutional neural network model; and S4, outputting a plaque tissue region segmentation image corresponding to the blood vessel wall magnetic resonance image. The method improves the accuracy of plaque segmentation of the blood vessel wall magnetic resonance image, thereby obtaining a high-precision segmentation image and contributing to improving diagnosis efficiency and improving the safety of an operation.

Description

Technical field [0001] The invention relates to a magnetic resonance image processing method, in particular to an intelligent segmentation method of intracranial plaque and carotid artery plaque of magnetic resonance images. Background technique [0002] Stroke is the disease with the highest mortality and disability rate in my country, and has become a serious burden on my country's medical expenditures and social families. Stroke includes hemorrhagic stroke and ischemic stroke. Stroke in my country is dominated by ischemic stroke, accounting for 79% of all stroke cases, and there is an increasing trend. Studies have shown that thrombosis caused by rupture of atherosclerotic plaque is the main pathogenesis of ischemic stroke. Among them, 46.6% of criminal plaque originates from intracranial arteries, 27.9% originates from carotid arteries, and the rest mainly originates from the heart and Thoracic aorta. Therefore, timely detection of stroke-related vascular beds, including vu...

AI Technical Summary

Problems solved by technology

However, the small diameter (1-2mm) and coiled anatomical characteristics of intracranial arteries have caused great difficulties in the graphic analysis and treatment of intracranial plaques

In addition, the deep learning network requires tens of thousands of data volumes for training, and its regions of interest are usually irregular and of different sizes, making it difficult to achieve accurate identification and segmentation

Images