基于卷积辅助自注意力的胸部疾病分类网络

张自然,李锵,关欣

Classification network for chest disease based on convolution-assisted self-attention

Ziran ZHANG,Qiang LI,Xin GUAN

表 2 不同胸部疾病分类网络在CheXpert验证集上的结果比较

Tab.2 Comparison of result of different chest disease classification network on CheXpert validation set

疾病类别 AUC U-Ignore U-Zeros U-Ones PCAN DCNN MAE CAWSNet 肺不张 0.818 0.811 0.858 0.848 0.825 0.827 0.835 心脏肿大 0.828 0.840 0.832 0.865 0.855 0.835 0.856 肺实变 0.938 0.932 0.899 0.908 0.937 0.925 0.917 水肿 0.934 0.929 0.941 0.912 0.930 0.938 0.953 胸膜增厚 0.928 0.931 0.934 0.940 0.923 0.941 0.928 平均值 0.889 0.889 0.893 0.895 0.894 0.893 0.898