Full-view pathological section image classification method and device, equipment and storage medium

Abstract

The invention discloses a full-view pathological section image classification method and device, equipment and a storage medium. The full-view pathological section image classification method comprises the steps of acquiring a to-be-classified full-view pathological section image; inputting the to-be-classified full-view pathological section image into a pre-constructed full-view pathological section image feature vector extraction model, and outputting a full-view pathological section image feature vector; and inputting the full-view pathological section image feature vector into a pre-constructed full-view pathological section image classification model, and outputting a full-view pathological section image classification result. According to the invention, the accuracy of full-view pathological section image classification can be improved.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a full-field pathological slice image classification method, device, equipment and storage medium. Background technique [0002] Cancer, as one of the diseases with the highest mortality rate in modern times, has shown a rapid growth trend in recent years. Pathology slides are the gold standard for doctors to diagnose cancerous tumors. Traditional cancer classification methods are screened by experienced pathologists through full-field pathological sections stained with hematoxylin and eosin (H&E). Massive visual screening work consumes a lot of time and energy for doctors. [0003] With the development of computer vision intelligent analysis, a deep learning-based method for classifying images of cancer full-field pathological slices has been proposed, which reduces the requirement for manual feature extraction, and uses deep neural networks to automaticall...

AI Technical Summary

Problems solved by technology

The full-view pathological classification method based on convolutional neural network lacks the surrounding spatial feature information due to the complete use of convolutional neural network; the breast cancer classification method based on residual network cannot capture the multi-scale feature information in the network, so the residual network cannot Good at exploring new features

In summary, the method based on deep learning full-field pathological slice image classification lacks surrounding spatial feature information and multi-scale feature information, which leads to model feature extraction using only current regional feature information, current scale information, surrounding spatial feature information, and multi-scale dimension information. Unable to provide auxiliary judgment, resulting in classification errors and affecting the accuracy rate

Images