An interpretable pulmonary tuberculosis classification network recognition method based on CT images

Abstract

The invention discloses a CT image-based interpretable pulmonary tuberculosis classification network identification method, comprising the following steps: acquiring a CT image of a person to be tested for lung computed tomography, normalizing and scaling the lung parenchyma image; establishing and training Interpretable pulmonary tuberculosis classification network, interpretable pulmonary tuberculosis classification network is built on the basis of residual network and adding Dense idea and improved attention mechanism; normalized and scaled to 512x512 size lung parenchyma image is sent into To the trained interpretable tuberculosis classification network, the classification results of tuberculosis of the test person and the corresponding class activation heat map are obtained and output in the host computer. According to the invention, the classification result of pulmonary tuberculosis of the person to be tested can be obtained and the corresponding class activation heat map can be output according to the CT image of the person to be tested.

Description

technical field [0001] The invention relates to the fields of computer-aided diagnosis and treatment and image recognition, in particular to an interpretable pulmonary tuberculosis classification network recognition method based on CT images. Background technique [0002] At present, tuberculosis is one of the major infectious diseases that threaten health, and its typical manifestation is pulmonary tuberculosis. Tuberculosis is a chronic infectious disease with a long history of endangering human health. The most serious epidemic in the early 20th century even caused the death of millions of people around the world. At present, my country is one of the countries with a high burden of tuberculosis. Because pulmonary tuberculosis is highly infectious and lethal, early auxiliary diagnosis of pulmonary tuberculosis can help doctors detect early pulmonary tuberculosis patients, carry out early diagnosis and early treatment, and reduce the infection rate and mortality rate, so it...

AI Technical Summary

Problems solved by technology

In CT images, pulmonary tuberculosis has the characteristics of polymorphism, multiple sites, multiple nodules, and cavities, and the mixture of various forms makes it difficult to identify, and medical staff need to read a large number of CT image sequences of lungs of different patients. Under the pressure of work, medical staff are prone to fatigue, resulting in missed diagnosis or even misdiagnosis

Images