Pathological image training set processing method and system

Abstract

The invention provides a pathological image training set processing method and system. A training set processed by the method is used for training an analysis learning network of pathological images, and the training set processing method comprises the following steps: utilizing a processor: acquiring the training set of the pathological images, the training set comprising a plurality of training images; grouping a plurality of training images in the training set to enable the training images in the same group to have the same image quality level; for each group of training images, determining a training priority based on the image quality level, so that the higher the image quality level is, the higher the training priority is; the groups of training images are sequentially used for training of the analytic learning network according to the training priority. The analysis accuracy of the analysis learning network can be improved, and a good analysis effect is ensured.

Description

[0001] This application is a divisional application of a Chinese invention patent application with the application number 202110985066.0, the application date is August 26, 2021, and the invention title is "A Method and System for Processing a Training Set of Pathological Images". technical field [0002] The present disclosure relates to the technical field of image analysis, in particular to a training set processing method and system for pathological images. Background technique [0003] Analysis and screening based on pathological images is an important means of cancer diagnosis in modern medicine. In the process of screening based on pathological images, professional pathological examiners are required to classify or segment pathological images through manual reading. The workload is heavy and the analysis takes a long time, which affects the efficiency of subsequent diagnosis and treatment. Therefore, In the prior art, an automatic analysis system for pathological imag...

AI Technical Summary

Problems solved by technology

However, due to the large number of cells contained in the pathological slices, the cells have different shapes, and there are also problems such as cell coverage, occlusion, or cell membrane folding, which affect the quality of the pathological images after digitization; In the process of digitizing slices, due to the influence of dye differences or scanning equipment, there will also be large quality fluctuations between pathological images. At this time, the above pathological images are input as a training set to the analysis and learning network for training. In the process, the training effect of the analysis and learning network will be poor, which will affect the analysis effect during subsequent use, and the accuracy of the analysis results such as cell classification and segmentation cannot meet the actual needs.

Images