Image segmentation method, device, electronic equipment and storage medium

Abstract

An image segmentation method applied in the technical field of medical image processing, comprising: encoding the image to be segmented step by step to obtain the highest-level feature image of the image to be segmented, and inputting the highest-level feature image to a compact feature learning model to obtain a compact feature image, the compact feature image is combined with the highest-level feature image to perform step-by-step decoding to restore the spatial information of the compact feature image to obtain the target image to be segmented, and the class to be segmented for each pixel in the target image to be segmented Predicting, obtaining a segmentation prediction map of the target image to be segmented, and extracting a prediction boundary of the target image to be segmented according to the segmentation prediction map. The application also discloses an image segmentation device, electronic equipment and a storage medium, which can improve the accuracy of image segmentation and at the same time reduce the influence of noise in image segmentation.

Description

technical field [0001] The present application relates to the technical field of medical image processing, and in particular to an image segmentation method, device, electronic equipment and storage medium. Background technique [0002] Limited by medical imaging methods and medical imaging equipment, medical images often have blurred edges, which makes it difficult for doctors to label, especially the labeling of boundaries is heavily dependent on the experience and supervisory standards of doctors. Between multiple annotations by the same doctor, there are differences in annotations between different doctors, and the resulting internal noise problem is often ignored in the automatic segmentation model training process. [0003] At present, there are existing research methods, and the way to optimize the boundary is often to assign higher weights to the boundary pixels during the training process, which enhances the sensitivity of the model to the boundary to a certain exte...

AI Technical Summary

Problems solved by technology

[0002] Limited by medical imaging methods and medical imaging equipment, medical images often have blurred edges, which makes it difficult for doctors to label, especially the labeling of boundaries is heavily dependent on doctors' experience and supervisory standards

Between multiple annotations by the same doctor, there are differences in annotations between different doctors, and the resulting internal noise problem is often ignored in the automatic segmentation model training process

[0003] At present, there are existing research methods, and the way to optimize the boundary is often to assign higher weights to the boundary pixels during the training process, which enhances the sensitivity of the model to the boundary to a certain extent, but also increases the noise brought by the training process. influences

Images