Segmentation method for esophagus cancer in chest CT images

Abstract

The invention discloses a segmentation method for esophagus cancer in chest CT images. According to the method, first, multiple groups of CT images containing the esophagus cancer are selected, and the CT images containing the esophagus cancer are used as training samples; the selected CT images are preprocessed, esophagus cancer features are acquired, and the images obtained after feature description is performed are used as training data; an esophagus cancer semantic segmentation model based on a full-convolution neural network is established, described esophagus cancer features are used asfeature input of the full-convolution neural network to serve as learning samples for training, and an esophagus cancer segmentation network model is obtained; three-dimensional reconstruction is performed on the esophagus cancer, wherein three-dimensional reconstruction and analysis are performed on an esophagus cancer segmentation result obtained through the obtained esophagus cancer segmentation network model, and esophagus cancer image omics parameters under a three-dimensional space are obtained; and the obtained esophagus cancer image omics parameters are visually displayed. The method is small in model scale, high in speed and high in accuracy.

Description

technical field [0001] The invention relates to the technical field of medical image processing, in particular to a method for segmenting esophageal cancer in chest CT images. Background technique [0002] With the continuous development of computer technology, medical imaging is also developing rapidly, so that more and more medical data can be processed and judged by computer, which improves the probability of successful disease diagnosis and treatment. In the diagnosis and detection of esophageal cancer, in addition to endoscopy, computerized tomography (CT) is commonly used. CT images can provide continuous slice grayscale images of a certain part of the human body, and can show the interior of the human body in detail. Therefore, the radiomics method based on CT image processing is often used in the diagnosis and prognosis of esophageal cancer. In the computer-aided diagnosis system, accurate segmentation of CT scan images of esophageal cancer is a key technology for s...

AI Technical Summary

Problems solved by technology

Since the esophagus only accounts for a small part of the body on CT slices and is closely related to other related organs, the proportion of the area occupied by other tissues and organs is small, and the contrast with other organs is also relatively low. Conventional images Segmentation methods are difficult to obtain esophageal cancer regions, which makes it difficult for radiomics analysis and prognosis prediction of esophageal cancer

Images