A method and system for identifying cancer focus regions based on full-slice pathological images

Abstract

The present invention proposes a method and system for identifying cancer focus areas based on pathological images, including: obtaining multiple pathological images with labels, each pathological image has a content label marking the cancer focus area and a contour label for the outline of the cancer focus; Construct a picture classification model including encoder, content decoder and contour decoder. The encoder is used to downsample the image to obtain the downsampling feature map. The contour decoder decodes the profile feature map according to the downsampling feature map. The contour decoder has to The skip connection of the content decoder, through the skip connection, the content decoder decodes the content feature map according to the downsampling feature map and the profile feature map; input the marked pathological image into the picture classification model, and train the picture at the same time in the way of joint supervision The classification model performs content recognition and contour recognition tasks, and calculates the loss function, and updates the image classification model through the loss as the cancer focus area recognition model; the cancer focus area is identified through the cancer focus area recognition model.

Description

technical field [0001] The present invention relates to the technical field of image processing, and in particular to a method and system for identifying cancer focus regions based on full-slice pathological images. Background technique [0002] In the field of computer-aided medical technology, the identification of cancer focus areas in whole-slice pathological images (WSI) has a broad application prospect. In addition to directly assisting doctors in diagnosis, cancer focus area recognition technology can also serve as the basis for many complex technologies, such as cancer cell segmentation, cancer classification and grading, and prediction of patient prognosis. [0003] A classic method is to treat this task as a classification problem, divide the full-section pathological image into small blocks and mark them at the block level, and train a CNN-based image classifier for cancer / non-cancer classification. For example, Nicolas et al. trained a tumor classifier based on ...

AI Technical Summary

Problems solved by technology

This feature makes the boundary problem of tumor region segmentation on pathological images more complicated than natural images and requires additional attention

Images