A method, apparatus, device and system for detect tissue nodules and model train thereof

Abstract

A method, apparatus, device and system for detect tissue nodules and model train thereof are disclosed. Source domain data and target domain data are obtained. That source domain data comprises a source domain image and an image annotation. A neural network model is used for feature extraction of the source domain image and the target image to obtain the source domain sampling feature and the target sampling feature, and the training result is determined according to the source domain sampling feature. Determining a distance parameter between the source domain data and the target domain data according to the source domain sampling characteristic and the target sampling characteristic; Determining a loss function value corresponding to the source domain image according to the training result and the image annotation; A tissue nodule detection model is determined based on the loss function value, the distance parameter, and the neural network model. Thus, the accuracy of detection can beimproved.

Description

technical field [0001] The present application relates to the technical field of medical image processing, in particular to a tissue nodule detection model training method, device and computer equipment, and a tissue nodule detection method, device, computer equipment and system. Background technique [0002] Tissue cancer is a leading cause of death, so early detection and treatment are crucial. The presence or absence of nodules in tissue is a strong indicator of cancer. At present, it is possible to judge whether there are tissue nodules with the help of medical images, such as chest thin-section (CT) images, which greatly increases the workload of doctors. In order to reduce the burden on doctors, automatic recognition of tissue nodules in tissue images has become a very critical technology. [0003] With the continuous development of artificial intelligence, in current tissue image processing methods, tissue nodules in tissue images can be identified based on a convol...

AI Technical Summary

Problems solved by technology

However, due to the inconsistency of medical imaging equipment in various hospitals, the sampling distance, noise level, and nodule diameter distribution of the captured tissue image data sets are different, which makes the image distribution in the actual detection and the image data set distribution in the training data set different.

[0004] Therefore, the traditional tissue nodule detection method has the problem of poor accuracy.

Images