Image segmentation method, computer equipment and storage medium

Abstract

The invention relates to an image segmentation method, computer equipment and a storage medium. According to the method, an input to-be-segmented image with a plurality of to-be-segmented structures is segmented through a segmentation network, wherein the segmentation network comprises multi-layer convolution operations, and the convolution operation of at least one layer comprises the network structures of a first convolution, a second convolution and a third convolution; and as the number of output channels of the first convolution, the number of input and output channels of the second convolution and the number of input channels of the third convolution are compressed according to a preset proportion, compared with the traditional convolution operation, the number of input channels andthe number of output channels are kept unchanged in the convolutional operation including three convolutions, and the number of the model parameters of the convolution operation is greatly reduced, sothat the number of the model parameters of the segmentation network comprising at least one layer of the convolution operation is reduced, and the consumption of a memory and a video memory when thesegmentation network is used for image segmentation processing is reduced.

Description

Technical field

[0001] This application relates to the technical field of medical image processing, and in particular to an image segmentation method, computer equipment and storage medium. Background technique

[0002] With the popularization and application of deep learning, more and more research institutions have entered the field of deep learning. In the field of medical detection technology, convolutional neural networks are usually used to classify, detect, and segment medical images. Among them, the segmentation of medical images is an important part of current medical detection. Moreover, due to the high complexity of human tissues and organs, it may include many sub-structures. For example, the brain atlas contains hundreds or even thousands of sub-structures. Based on the above application scenarios, there are correspondingly many large-scale sub-structures based on convolutional neural networks. Method of segmentation.

[0003] In recent years, the segmentation method...

Images