An automatic liver CT segmentation method based on deep shape learning

Abstract

The invention discloses an automatic liver CT segmentation method based on deep shape learning. Firstly, a liver segmentation data set is established and preprocessed, and the liver CT is roughly segmented by liver segmentation; secondly, a liver shape set is established, and a variational autoencoder is used. The liver shape is learned, and a geometric shape regularization module is constructed, and then the geometric shape regularization module is added to liver segmentation to obtain a liver segmentation model constrained by geometric shape consistency, which is used for automatic liver CT segmentation. The present invention innovatively adds the represented shape features into the existing deep segmentation network through the regularization module, and introduces shape prior information in the training process of the convolutional neural network, which can improve the regularity and generalization of the segmentation model The ability makes the segmentation results more in line with the medical anatomical characteristics of the standard liver. The invention has the characteristics of automation, high precision, and transferability and expansion, and can realize automatic and precise segmentation of large abdominal organs represented by the liver.

Description

Technical field [0001] The present invention relates to the field of CT segmentation techniques, and more particularly to a method of automatic CT automatic segmentation based on deep shape learning. Background technique [0002] Liver cancer is one of the most common malignant tumors worldwide, seriously threatening human life health. my country is a high-risk area of ​​liver cancer, liver cancer ranks third in my country's malignant tumor death. Liver cancer is not only a high degree of malignancy, the disease is fast, and patients with early liver cancer lack specific clinical manifestations, and patients often in the middle and evening, treatment difficulties in the disease and prognosis. Computer tomogram (CT) is one of the clinical imaging examination methods of liver cancer diagnosis. The level of liver based on CT images can provide quantification information such as volume, size, shape, etc., and is the necessary steps of subsequent lesions detection, imaging analysis. B...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to address the deficiencies of the prior art, and propose a liver CT automatic segmentation method based on deep shape learning. The technical problem addressed by the present invention is that the existing deep learning liver segmentation method lacks geometric shape regularity, and cannot be well represented and Fusion Geometry Priors

Images