Table 3. Comparison of diagnostic performance between using and not using the proposed pre-processing procedures.
Accuracy | Sensitivity | Specificity | Precision | F1-Score | |
---|---|---|---|---|---|
Proposed scheme | 0.853 ± 0.050 | 0.822 ± 0.087 | 0.912 ± 0.034 | 0.829 ± 0.089 | 0.822 ± 0.065 |
Basic scheme | 0.747 ± 0.027 | 0.738 ± 0.124 | 0.878 ± 0.057 | 0.735 ± 0.071 | 0.725 ± 0.070 |
The values are given as the mean and the standard deviation by 5-fold cross validation. The best performing CNN model (EfficientNet-b2) was used for comparison.