Acute ischemic stroke focus automatic segmentation method and system, terminal and medium

Abstract

The invention discloses an acute ischemic stroke focus automatic segmentation method and system. The method comprises the following steps of: registering a head CT (Computed Tomography) image; removing the skull to obtain a brain parenchyma three-dimensional mask; performing ellipse fitting on the brain parenchyma three-dimensional mask through PCA feature dimension reduction to obtain median sagittal plane data; calculating left and right symmetric parameter diagrams of the brain; setting different window widths and window levels for the median sagittal plane data to obtain multi-level window width and window level image data; marking training data; training a segmentation convolutional neural network model; inputting test data into the trained segmentation model for segmentation to obtain a segmentation probability graph of the brain parenchyma, the cerebrospinal fluid and the stroke focus of the corresponding voxel and a prediction probability of whether the stroke focus exists in the corresponding two-dimensional image; and outputting different segmentation results according to different prediction probabilities. According to the method, multi-task training is carried out in combination with brain anatomy priori knowledge, the performance of the convolutional neural network in the NCCT stroke focus automatic segmentation method is improved, and the acute ischemic stroke focus can be accurately segmented.

Description

technical field [0001] The invention relates to the technical field of data processing, in particular to a method, system, terminal and medium for automatic segmentation of acute ischemic stroke lesions. Background technique [0002] "Cerebral stroke" is an acute cerebrovascular disease, which is a group of diseases that cause brain tissue damage due to sudden rupture of blood vessels in the brain or blockage of blood vessels that prevent blood from flowing into the brain, including ischemic and hemorrhagic strokes . [0003] In 2018, the mortality rate of stroke in China was 149.49 / 100,000, accounting for 22.3% of the total mortality of Chinese residents, and it has become the number one cause of premature death and disease burden. The average age of hospitalized stroke patients was 66 years old, and the most important types of cerebral infarction, cerebral hemorrhage and subarachnoid hemorrhage accounted for 81.9%, 14.9% and 3.2% respectively. [0004] Acute ischemic str...

AI Technical Summary

Problems solved by technology

Most of the existing research on CNN segmentation of stroke lesions only considers the addition of three-dimensional spatial structure information, and lacks the use of prior knowledge related to NCCT images and stroke

Images