Low-differentiation gland segmentation method based on internal and external stress

Abstract

The invention discloses a low-differentiation gland segmentation method based on internal and external stress, and the method comprises the following steps: 1), carrying out the dyeing separation of apathological tissue dyeing image based on a ResUnet architecture, and obtaining a hematoxylin channel image and a background channel image; 2) segmenting a gland lumen region from the background channel image based on a variational level set image segmentation algorithm of an improved symbolic pressure function; 3) taking the hematoxylin channel diagram as an SC-CNN input feature to obtain an epithelial cell region boundary, namely an gland boundary formed by epithelial cell nucleuses; and 4) drawing the gland contour according to the lumen shape characteristics based on a graphic shape description method of the minimum inertia axis and the chain code. According to the method, information contained in the H&E dyed image is more independent and easy to identify, so that the conditions of non-uniform dyeing intensity and unobvious dyeing difference are processed, a group of new segmented gland contour features are developed and combined, and a method for expressing lumen and gland outerwheel width shape features is clearly provided.

Description

technical field [0001] The invention relates to the technical field of image information processing, in particular to a method for segmenting poorly differentiated glands based on internal and external stress. Background technique [0002] Adenocarcinoma is a malignant tumor formed by glandular structures in epithelial tissue. It affects the distribution of cells and also changes the structure of the gland. A biopsy is a tissue taken from a suspicious organ in a minimally invasive manner. Pathologists use a microscope to identify biopsy tissues. Pathologists need to be precise and have the ability to detect large amounts of data to detect small abnormalities in biopsies. Taking the digestive system as an example, the histopathological staining images of the colon are the main basis for detecting lesions. Typical histopathological images of colonic glands contain four tissue components: lumen, cytoplasm, epithelial cells, and stroma (connective tissue, blood vessels, neural...

AI Technical Summary

Problems solved by technology

All of the above methods require a large number of manually annotated images, but it is very difficult to annotate a large number of medical images

Images