Skip to main content
. 2020 Nov 10;24(12):3595–3605. doi: 10.1109/JBHI.2020.3037127

TABLE VII. Results of the Baseline Classification Model With Segmentation, COVID-SDNet, Retrained COVIDNet-CXR-A and Retrained COVID-CAPS. Only Three Levels of Severity are Considered, Mild, Moderate and Severe.

Class N P Accuracy
Metric Specificity Precision F1 Sensitivity Precision F1
COVIDNet-CXR 83.42Inline graphic 15.39 69.73Inline graphic 10.34 74.45Inline graphic 8.86 61.82Inline graphic 22.44 79.50Inline graphic 11.47 65.64Inline graphic 15.90 72.62Inline graphic 7.6
COVID-CAPS 65.09Inline graphic 10.51 71.72Inline graphic 5.57 67.52Inline graphic5.29 73.31Inline graphic9.74 68.40Inline graphic5.13 70.20Inline graphic4.31 69.20Inline graphic3.61
With seg. 80.57Inline graphic8.72 78.68Inline graphic6.57 78.97Inline graphic3.20 76.80Inline graphic10.15 80.70Inline graphic5.56 78.01Inline graphic4.29 78.69Inline graphic3.00
FuCiTNet 82.63Inline graphic6.61 79.94Inline graphic4.28 81.05Inline graphic3.44 78.91Inline graphic5.88 82.43Inline graphic5.43 80.37Inline graphic3.16 80.77Inline graphic3.15
COVID-SDNet 85.20Inline graphic5.38 78.88Inline graphic3.89 81.75Inline graphic2.74 76.80Inline graphic6.30 84.23Inline graphic4.59 80.07Inline graphic0.04 81.00Inline graphic2.87