Skip to main content
. 2023 Jun 29;13:10528. doi: 10.1038/s41598-023-37603-1

Table 4.

Median values for the evaluation metrics computed from the predictions of the additional test set during the 30 iteration rounds.

CNN Accuracy (%) Sensitivity (%) Specificity (%) F1 (%) AUC (%)
Non-augmented shallow 63.0 72.3 52.5 65.9 70.1
Augmented shallow 67.6 61.3 73.9 65.6 71.9
Non-augmented deep 70.4 73.5 67.6 70.7 78.1
Augmented deep 75.0 69.7 80.3 73.4 79.4