Table 2 Classification performance of deep learning algorithms on LGG and GBM subtypes for validation set
Methods | Performance measures (Average of 10 fold cross-validation on test datset) | |||||||
|---|---|---|---|---|---|---|---|---|
Accuracy [95% CI] | Precision [95% CI] | Recall [95% CI] | F1-score [95% CI] | FPR [95% CI] | Gmean [95% CI] | MCC [95% CI] | ||
LGG | ANN | 90.18% | 82.40% | 84.23% | 82.42% | 0.07 | 90.06% | 0.8 |
[98.16–98.20] | [82.35–82.45] | [84.19–84.27] | [82.37–82.47] | [0.05–0.09] | [90.04–90.08] | [0.75–0.85] | ||
CNN | 95.23% | 92.08% | 92.63% | 91.84% | 0.03 | 95.30% | 0.9 | |
[95.22–95.24] | [92.05–92.11] | [92.60- 92.66] | [91.83–91.87] | [0.02–0.04] | [95.29–95.31] | [0.86–0.94] | ||
GBM | ANN | 93.85% | 93.85% | 93.85% | 93.85% | 0 | 1 | 1 |
[93.79–93.91] | [93.79–93.91] | [93.79–93.91] | [93.79–93.91] | [0–0] | [1–1] | [1–1] | ||
CNN | 90.26% | 85.38% | 87.69% | 86.15% | 0.02 | 95.26% | 0.92 | |
[90.20–90.32] | [85.29–85.47] | [87.62–87.76] | [86.06–86.24] | [0–0.04] | [95.21–95.31] | [0.84–1] | ||