Research Article
Diagnosis of Retinal Diseases Based on Bayesian Optimization Deep Learning Network Using Optical Coherence Tomography Images
Table 7
Performance of InceptionV3.
| Class labels | Precision (%) | Recall (%) | F1-score (%) | Feature extractor | Fine tuner | Feature extractor | Fine tuner | Feature extractor | Fine tuner |
| AMD | 90.12 | 95.62 | 91.81 | 92.31 | 90.60 | 93.27 | CNV | 88.15 | 96.71 | 90.81 | 95.91 | 87.96 | 90.06 | DME | 90.31 | 93.98 | 87.48 | 93.27 | 87.72 | 91.72 | CSR | 88.18 | 95.91 | 91.02 | 91.78 | 90.34 | 89.45 | DR | 87.99 | 93.11 | 90.18 | 93.91 | 90.06 | 91.62 | Drusen | 89.41 | 96.81 | 89.38 | 89.95 | 89.43 | 93.02 | MH | 90.01 | 96.21 | 91.64 | 96.81 | 89.23 | 91.62 | Normal | 88.62 | 95.47 | 89.74 | 95.99 | 90.82 | 93.49 | Macro average | 89.10 | 95.48 | 90.26 | 93.74 | 89.52 | 91.78 | Weighted average |
|
|