Automatic gallstone recognition and segmentation system based on deep learning, computer equipment and storage medium

Abstract

The invention discloses a gallstone automatic identification and image automatic segmentation system based on deep learning, and belongs to the technical field of image identification and image segmentation. The system comprises a target detection and target segmentation model, the target detection and target segmentation model comprises a feature extraction network and two branch networks comprising a target detection network and a target segmentation network; Firstly, convolution pooling and other operation feature extraction is carried out on an input abdominal CT image by using a feature extraction network, regression and classification are carried out on an extracted feature pattern by using the target detection network, and the position and probability of gallstone are outputted; andmeanwhile, the image segmentation network performs up-sampling on the convolution and pooling feature pattern by utilizing deconvolution, the image is restored to the size of the original image, andfinally, the segmented contour, namely a mask, of the gall-stone is obtained, so that the features such as the size and the shape of the gall-stone can be clearly and quantitatively analyzed.

Description

technical field [0001] The present invention relates to the field of image recognition technology and image segmentation technology, in particular to an automatic recognition and segmentation system for gallstones based on deep learning, computer equipment, and storage media. Background technique [0002] Gallstones are a common disease worldwide. In traditional diagnostic methods, physicians extract and mark gallstones based on past experience through collected medical images. This empirical diagnostic method brings great subjectivity. At the same time, the CT imaging characteristics of gallstones are very similar to those of acute cholecystitis and other diseases, which can easily lead to missed and misdiagnosed diagnoses, which greatly depend on the experience and professionalism of doctors, and the accuracy of diagnosis is limited. [0003] The medical field has gradually begun to use artificial intelligence technology and computer vision-related technologies for auxilia...

AI Technical Summary

Problems solved by technology

[0002] Gallstones are a common disease worldwide. In traditional diagnostic methods, physicians extract and mark gallstones based on past experience through collected medical images. This empirical diagnostic method brings great subjectivity. At the same time, the CT imaging features of gallstones are very similar to those of acute cholecystitis and other diseases, which can easily lead to missed and misdiagnosed diagnoses, which greatly depend on the experience and professionalism of doctors, and the accuracy of diagnosis is limited.

Images