• Laser & Optoelectronics Progress
  • Vol. 60, Issue 14, 1410001 (2023)
Shuai Zhang1, Junzhong Zhang2, Hui Cao1、*, Dawei Qiu1、**, and Xurui Ji1
Author Affiliations
  • 1College of Intelligence and Information Engineering, Shandong University of Traditional Chinese Medicine, Jinan 250355, Shandong, China
  • 2First Clinical Medical College, Shandong University of Traditional Chinese Medicine, Jinan 250355, Shandong, China
  • show less
    DOI: 10.3788/LOP222126 Cite this Article Set citation alerts
    Shuai Zhang, Junzhong Zhang, Hui Cao, Dawei Qiu, Xurui Ji. Coronavirus Disease X-Ray Image Diagnosis Method Based on ConvNeXt Network[J]. Laser & Optoelectronics Progress, 2023, 60(14): 1410001 Copy Citation Text show less
    GlstNet framework structure
    Fig. 1. GlstNet framework structure
    LSTM network structure
    Fig. 2. LSTM network structure
    CNN-LSTM network structure
    Fig. 3. CNN-LSTM network structure
    Confusion matrix for the test set on dataset I
    Fig. 4. Confusion matrix for the test set on dataset I
    GlstNet accuracy change curve with confusion matrix
    Fig. 5. GlstNet accuracy change curve with confusion matrix
    GlstNet network confusion matrix on Chest X-ray validation set
    Fig. 6. GlstNet network confusion matrix on Chest X-ray validation set
    Score-CAM visualization results
    Fig. 7. Score-CAM visualization results
    DatasetTypeTotalTraining imageValidation imageTest image
    COVID-19 Radiography Databasecovid36162314578724
    normal8851566414161771
    lung_opacity601238479621203
    Table 1. Division of dataset I
    DatasetTypeTotalTraining imageValidation image
    Chest X-raycovid576460116
    normal15831266317
    pneumonia42733418855
    Table 2. Division of dataset Ⅱ
    No.ModelRaccuracy /%Rprecision /%Rrecall /%ΔRaccuracy/%
    1Baseline94.0094.8094.00
    2Baseline+CBAM92.0092.8091.50↓2.0
    3Baseline+GCT+SA90.5091.2089.90↓4.0
    4Baseline+ST+MCSA92.8093.2092.63↓1.2
    5Baseline+CGMB95.0095.7095.00↑1.0
    6GlstNet95.6096.0395.76↑1.6
    Table 3. Test set metrics on dataset I
    ModelRaccuracy /%Rprecision /%Rrecall /%
    ResNet182793.4393.4393.43
    ResNet502793.0193.1293.02
    ResNet1012793.0193.0493.01
    ChexNet2793.2193.2893.21
    DenseNet2012792.7092.7892.70
    InceptionV32793.4693.4993.47
    ConvNeXt94.0094.8094.00
    Vision Transformer94.6495.2394.63
    Swim Transformer95.1895.6695.23
    GlstNet95.6096.0395.76
    Table 4. Comparison of GlstNet network with mainstream algorithms
    TypeRaccuracy /%Rprecision /%Rrecall /%
    covid98.5097.5099.60
    normal94.1093.8097.20
    lung_opacity95.5096.0095.80
    Table 5. Performance metrics for three categories on the validation set of dataset Ⅰ
    ModelRaccuracy /%Rprecision /%Rrecall /%
    ALEXNET CNN3094.7094.0094.00
    LENET CNN3093.2093.0093.00
    DenseNet-1693095.70
    ResNet-503193.30
    DenseNet3295.0394.9194.24
    EfficientNet3293.4093.6291.07
    VGG-163292.0892.0292.77
    GlstNet97.2097.0397.60
    Table 6. Comparison of GlstNet with mainstream algorithms
    TypeRaccuracy/%Rprecision /%Rrecall /%
    covid100.0099.10100.00
    normal92.4096.5097.40
    pneumonia98.7097.2097.50
    Table 7. Performance metrics for three categories on the validation set of dataset II
    Shuai Zhang, Junzhong Zhang, Hui Cao, Dawei Qiu, Xurui Ji. Coronavirus Disease X-Ray Image Diagnosis Method Based on ConvNeXt Network[J]. Laser & Optoelectronics Progress, 2023, 60(14): 1410001
    Download Citation