Deformable context coding network model and liver and liver tumor segmentation method

Abstract

The invention discloses a deformable context coding network model and a liver and liver tumor segmentation method, which can accurately determine the contour positions of a liver and a liver tumor, realize more accurate liver and liver tumor segmentation, have a wide application prospect, and are suitable for popularization and application. According to the network model of the invention, deformable convolution is utilized to enhance the feature representation capability of a traditional encoder, help the traditional encoder to learn a convolution kernel with adaptive spatial structure information, and eliminate interference of different sizes of liver tumor positions; according to the method, the global feature information in the image is coded by using the Ladder spatial pyramid poolingmodule for extracting the multi-scale context information, so that the contour positions of the liver and the liver tumor are determined more accurately, more accurate liver and liver tumor segmentation is realized, and the method has a wide application prospect.

Description

technical field [0001] The invention belongs to the field of image processing technology and pattern recognition, and in particular relates to a deformable context encoding network model and a segmentation method for liver and liver tumors. Background technique [0002] Currently, primary liver cancer has become one of the most common cancers with the highest lethality worldwide, threatening human life and health seriously. Accurate liver and liver tumor segmentation on abdominal CT images is of great value in assisting doctors in diagnosis, improving treatment success rate and reducing patient harm. However, CT images usually have the characteristics of large noise and low contrast, which makes the gray difference between the liver and liver tumors and other tissues smaller in the image, and the shape of liver tumors is highly variable and difficult to delineate intuitively. Segmentation of liver tumors and liver tumors is more difficult; in addition, manual labeling of ab...

AI Technical Summary

Problems solved by technology

Ronneberger et al. proposed that the fully symmetrical U-Net of the encoder and decoder obtains better segmentation results, but the convolution of fixed geometry usually cannot match the liver and irregular tumors, and cannot effectively extract semantic features.

Seo et al. added convolution and activation operations in the skip connection between the encoder and the decoder to enhance the network's ability to detect the liver and liver tumors, thereby obtaining better segmentation results, but this segmentation method cannot effectively utilize the image space context information, and the global feature information of the liver and liver tumor cannot be extracted, so the segmentation accuracy is poor

Christ et al. used the cascaded U-net model to segment the liver and liver tumors. This method can achieve better liver tumor segmentation results, but the segmentation edges are not fine enough and the smoothness is poor.

Images