Depth learning-based abnormal chest radiograph intelligent identification method and system

Abstract

The present invention discloses a depth learning-based abnormal chest radiograph intelligent identification method and system. The scheme is characterized in that: a deep neural network is trained byusing a large number of manually labeled samples, and the deep neural network identifies abnormal image features in the chest radiograph through self-learning of the abnormal image features in the chest radiograph. The resulting abnormal chest radiograph intelligent identification method can realize automatic identification of the abnormal image features in the chest radiograph, has high recognition efficiency and high recognition accuracy, can effectively avoid misdetection and unidentification, and can effectively solve the problem existing in the prior art.

Description

technical field

[0001] The invention relates to image recognition technology, in particular to the recognition technology of chest radiograph abnormality. Background technique

[0002] A chest X-ray is an X-ray of the chest, which is clinically called a chest radiography. Chest X-rays are widely used clinically, and people's understanding is getting deeper and deeper. Frontal chest radiography can show the size, shape, position and outline of the great vessels of the heart, and can observe the relationship between the heart and adjacent organs and the changes of blood vessels in the lungs, and can be used to measure the heart and its diameter. The left anterior oblique view shows the general picture of the aorta and the enlargement of the left and right ventricles and right atrium. Right anterior oblique view is helpful to observe the changes of left atrium enlargement, pulmonary artery segment protrusion and right ventricular infundibulum enlargement. The left side view ...

Images