Method for synchronously segmenting multiple organs of lung

Abstract

The invention relates to a method for synchronously segmenting multiple organs of a lung. The method comprises the following steps: a, performing interactive data annotation on the multiple organs of the lung; b, performing data preprocessing on the data annotation; c, constructing a data set for the data preprocessing result; d, enhancing and expanding the data set; e, carrying out DC-U-Net network model training on the data set to obtain a DC-U-Net training model; f, carrying out automatic segmentation on multiple organs of the lung based on the DC-U-Net training model. According to the invention, lung parenchyma, blood vessel, bronchus and pulmonary nodule areas can be automatically extracted based on the deep learning technology, and synchronous segmentation is realized.

Description

technical field [0001] The invention belongs to the field of medical image processing and analysis, and in particular relates to a synchronous segmentation method for multiple lung organs. Background technique [0002] The existing lung parenchyma, blood vessels, bronchi and pulmonary nodules segmentation methods are mainly extracted using various adaptive region growing methods and threshold segmentation methods. The disadvantages of this method are: 1) slow speed, poor generalization ability, It is necessary to repeatedly adjust multiple parameters. Due to the noise in the image, the method of judging each voxel can be accurately classified, but it is sensitive to noise, resulting in the need to repeatedly adjust parameters such as the growth threshold to achieve the best segmentation effect; 2) cannot Synchronous segmentation requires segmentation of the three organs in a certain order; 3) Frequent manual interaction and cumbersome segmentation process. At the same time,...

AI Technical Summary

Problems solved by technology

[0002] The existing lung parenchyma, blood vessels, bronchi and pulmonary nodules segmentation methods are mainly extracted using various adaptive region growing methods and threshold segmentation methods. The disadvantages of this method are: 1) slow speed, poor generalization ability, It is necessary to repeatedly adjust multiple parameters. Due to the noise in the image, the method of judging each voxel can be accurately classified, but it is sensitive to noise, resulting in the need to repeatedly adjust parameters such as the growth threshold to achieve the best segmentation effect; 2) cannot Synchronous segmentation requires segmentation of the three organs in a certain order; 3) Frequent manual interaction and cumbersome segmentation process

At the same time, some of the deep learning neural network methods are used to obtain segmentation results, which are divided into two types: three-dimensional network and two-dimensional network. Segmentation result Missegmentation is serious

No matter which method is used, the effect of synchronous segmentation of multiple organs cannot be achieved.

Images