Table 6 Treatment outcome prediction performance.
From: Multimodal deep learning for cephalometric landmark detection and treatment prediction
Treatment type | Sample size | DeepFuse accuracy (%) | Precision (%) | Recall (%) | F1-score (%) | Prior SOTA [Ref] accuracy (%) |
|---|---|---|---|---|---|---|
Class I correction | 428 | 88.7 | 90.2 | 87.5 | 88.8 | 73.5103 |
Class II correction | 513 | 84.2 | 86.4 | 82.9 | 84.6 | 69.8104 |
Class III correction | 392 | 79.8 | 81.5 | 77.2 | 79.3 | 68.6105 |
Surgical intervention | 275 | 91.3 | 93.6 | 89.8 | 91.7 | 78.2106 |
Overall | 1608 | 85.6 | 87.4 | 83.8 | 85.6 | 69.2107 |
