Semantic segmentation method and device for tree structure in three-dimensional tomography image

Abstract

The invention provides a semantic segmentation method for a tree structure in a three-dimensional tomography image, and relates to the technical field of medical image processing, and the method comprises the steps: obtaining a to-be-tested three-dimensional tomography image; preprocessing the image, obtaining a preprocessed image, and preprocessing comprises the steps of unifying the image resolution, cutting the image to be in a unified size and normalizing the gray value of the image; and inputting the preprocessed image into the tree-structure semantic segmentation network to obtain a semantic segmentation prediction result corresponding to the image. According to the method, the multi-task full convolutional network and the graph convolutional network are used for completing feature extraction together, then a semantic segmentation result of a tree structure is obtained, spatial context information is modeled, structural priori knowledge is explicitly introduced, and good segmentation performance can be achieved.

Description

technical field [0001] The present application relates to the technical field of medical image processing, in particular to a method and device for semantically segmenting tree structures in three-dimensional tomographic images. Background technique [0002] Computed tomography technology uses the differences in the transmittance and absorption rate of X-rays by different tissues of the human body, and can provide doctors with a three-dimensional anatomical view of the human body without surgical operations. It has the advantages of fast scanning speed and high resolution. It is a mainstream medical imaging method. The anatomical tree structures in 3D tomographic images include trachea, arteries, veins, etc. Examination and analysis of these structures is an important auxiliary means for diagnosis and treatment of related diseases. Taking chest tomographic images as an example, it can reflect the anatomical structure and physiological conditions of the patient's lungs and t...

AI Technical Summary

Problems solved by technology

[0006] For this reason, the first purpose of this application is to propose a semantic segmentation method for the tree-like structure in the three-dimensional tomographic image, which solves the problem of complex distribution of the anatomical tree-like structure, various morphological changes between different individuals, and various semantic categories. In tomographic images, the volume ratio is often very small, there is no obvious boundary plane between classes, the local appearance of the image and the gray distribution may be affected by pathology, etc., and the use of multi-task full convolutional network and graph convolutional network is realized. To complete the feature extraction and obtain the semantic segmentation results of the tree structure, the spatial context information is modeled, and the prior knowledge of the structure is explicitly introduced to achieve good segmentation performance

Images