Deep learning-based pneumothorax X-ray image recognition method and system

Abstract

The invention discloses a deep learning-based pneumothorax X-ray image recognition method and system. According to the system of the invention, a large number of manually-labeled samples are adopted to train a deep neural network; and the deep neural network learns the image features of pneumothorax to identify the X-ray image of the pneumothorax. The pneumothorax X-ray image recognition method and system can realize automatic recognition of chest X-ray images, have high recognition efficiency and high recognition precision. With the pneumothorax X-ray image recognition method and system adopted, missed detection and missed identification phenomena can be effectively avoided, and problems in the prior art can be solved.

Description

technical field [0001] The invention relates to image recognition technology, in particular to X-ray image recognition technology. Background technique [0002] A pneumothorax is when air enters the pleural space, causing a pneumothorax. The lung tissue and visceral pleura are usually ruptured due to lung disease or external force, or the tiny emphysema bubbles near the lung surface are ruptured, and the air in the lungs and bronchi escapes into the pleural cavity. It is more common in young males or those who suffer from chronic bronchitis, emphysema, and tuberculosis. This disease is one of pulmonary emergencies, which can be life-threatening in severe cases, and can be cured if treated in time. [0003] X-ray examination is an important method for diagnosing pneumothorax, and chest X-ray is the most commonly used examination method for diagnosing pneumothorax. see figure 1 Most of the pneumothorax chest films have a clear pneumothorax line, which is the junction line ...

AI Technical Summary

Problems solved by technology

[0004] However, the identification of existing chest films basically relies on purely manual film reading, which is not only costly, inefficient, and time-consuming; but also because pneumothorax lesions are low-brightness areas, they are sometimes overlooked when manually identifying and reading films. The recognition accuracy is not high; at the same time, there is still a phenomenon of missed diagnosis when reading images manually

[0005] It can be seen that there are generally problems of low efficiency and low accuracy in the existing manual-based identification of chest radiographs, so providing an efficient and high-precision pneumothorax chest radiograph identification technology is a technical problem that needs to be solved urgently in this field

Images