Low-field-intensity MR stomach segmentation method based on transfer learning image enhancement

Abstract

The invention discloses a low-field-intensity MR stomach image segmentation method based on transfer learning image enhancement. The method mainly solves the problems that a low-field-intensity 3DMR image is large in noise, more in artifacts and less in image data. According to the scheme, the method comprises the following steps: acquiring high-field-intensity and low-field-intensity 3DMR stomachimage data sets and corresponding label data sets, and preprocessing the image data; inputting the preprocessed data into a cyclic generative adversarial network to obtain an enhanced pseudo-low field intensity 3DMR stomach image set; constructing and training a 3D Res3Unet segmentation network; and inputting the pseudo low-field-intensity 3DMR stomach image set into a segmentation network, completing fine adjustment of segmentation network parameters, forming a 3D Res3Unet segmentation network after transfer learning, and inputting test data into the network to obtain a segmentation result.According to the method, the segmentation of the low-field-intensity 3DMR stomach image is realized, and the image segmentation precision is effectively improved.

Description

technical field [0001] The invention belongs to the technical field of medical image processing, and further relates to image segmentation technology, specifically a low-field-strength MR gastric image segmentation method based on transfer learning image enhancement, which can be used to accurately segment the gastric region in low-field-strength 3DMR images. segmentation. Background technique [0002] At present, the incidence rate of gastric cancer is second only to lung cancer, and the mortality rate ranks third. There are about 1.2 million new cases of gastric cancer every year in the world, and China accounts for about 40% of them. The detection rate of early gastric cancer in my country is low, only about 20%, and most of them have reached the advanced stage when they are discovered, and the overall 5-year survival rate is less than 50%. Therefore, research on the diagnosis of gastric diseases has received extensive attention, and more accurate and efficient medical ...

AI Technical Summary

Problems solved by technology

However, low-field-strength MR images have the problems of large noise, many artifacts, and less image data, and over-fitting will occur when using 3D deep learning networks for segmentation. Therefore, there is an urgent need for accurate, reliable, and generalization capabilities. The segmentation method to ensure the accuracy of image segmentation, so as to meet the requirements of practical applications

Images