Research and Practice of X-ray Chest Film Disease Classification based on DenseNet

Abstract

X-ray examination, commonly used to examine chest diseases, is a common medical examination in patients’ physical examination. In this paper, the disease information is extracted based on the X-ray chest film report, and the disease types are classified by using DenseNet. The highest AUC can reach more than 0.935, the positive recall rate is about 0.82, and the negative recall rate is about 0.92, which has a good effect.