Pulmonary tuberculosis identification method based on SE-ResNet

Abstract

The invention provides a pulmonary tuberculosis identification method based on SE-ResNet, and the method comprises the steps: obtaining a real pulmonary tuberculosis image, generating a pulmonary tuberculosis image through employing a DCGAN network, and constructing a matrix pulmonary tuberculosis image data set based on the real pulmonary tuberculosis image and the generated pulmonary tuberculosis image; preprocessing the pulmonary tuberculosis image data set, wherein preprocessing comprises data reading and data normalization processing; dividing the pulmonary tuberculosis image data set into a training set and a test set; constructing an SE-ResNet network based on a ResNet50 network and an SENet module, pre-training the network by using ImageNet data, initializing network parameters, and training the network by using a training set; and analyzing and identifying the pulmonary tuberculosis image to be identified by adopting the trained SE-ResNet network to obtain a corresponding analysis and identification result. According to the identification method, the pulmonary tuberculosis X-ray chest radiograph can be intelligently identified without manual intervention, the identification speed is high, and the precision is high.

Description

technical field [0001] The invention belongs to the field of image information detection, and in particular relates to a tuberculosis recognition method based on SE-ResNet. Background technique [0002] At present, the lack of professional doctors prevents tuberculosis patients from being diagnosed and treated in the first place, and pathology diagnosis from original medical image output is often a more complicated problem than the ideal problem faced by non-medical image classification, resulting in the existing The identification accuracy rate of tuberculosis intelligent diagnosis system is low. [0003] Problems or defects in the existing technology: tuberculosis patients often need to wait for a long time during the diagnosis process, and in some areas with insufficient medical conditions, patients will be mistakenly identified as negative, causing tuberculosis when they cannot receive treatment themselves. dissemination; and the existing tuberculosis detection technolo...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the above-mentioned technical problems of low efficiency and poor accuracy of the existing tuberculosis detection technology, the pre...

Images