Method for detection and recognition of cerebral microbleeds based on machine learning in swi images

Abstract

The invention discloses a method for detecting and identifying cerebral microbleeds in SWI images based on machine learning, which comprises the following steps: establishing a SWI image training library, reading any SWI image in the training library, and performing preprocessing; The image is processed to remove the skull image to obtain the SWI brain tissue image; obtain the microbleed candidate point area in the SWI brain tissue image; perform feature extraction on the extracted microbleed candidate point area to obtain the feature value of the microbleed point; use random forest Methods Training was carried out to obtain the classifier model of cerebral microbleeds; the SWI images to be identified were read and put into the classifier model of cerebral microbleeds for identification, the classification results of the SWI images to be recognized were obtained, and the information of cerebral microbleeds was output. This method solves the problems of time-consuming, labor-intensive, low reliability and repeatability for clinicians to manually identify CMBs lesions, and has important application value.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to a method for detecting and identifying cerebral microbleeds based on SWI images of machine learning. Background technique [0002] Cerebral microbleeds (Cerebral Micro-bleeds, CMBs) is a kind of tiny cerebrovascular lesions that lead to the deposition of hemosiderin in the blood of the brain, which widely exists in the patient's cortex, subcortical white matter and basal ganglia, etc. area, it will cause some damage to the corresponding brain tissue, which may cause cognitive dysfunction. At the same time, it is also one of the important risk factors for ischemic stroke. The existence of CMBs is highly correlated with the degree of loosening of white matter in the brain, and has become an important indicator for doctors to formulate anticoagulant and antiplatelet treatment plans as a reference. Therefore, the study and diagnosis of CMBs are considered to be helpful in...

AI Technical Summary

Problems solved by technology

Avoid the time-consuming and labor-intensive and low reliability problems of clinicians manually identifying CMBs lesions

Images