Table 3.
Performance benchmarking of our approach against leading techniques on ADNI and BraTS datasets.
| Model | ADNI dataset | BraTS dataset | ||||||
|---|---|---|---|---|---|---|---|---|
| mAP | Precision | Recall | F1 score | mAP | Precision | Recall | F1 score | |
| Faster R-CNN (42) | 81.32 0.02 | 78.45 0.03 | 79.88 0.02 | 79.12 0.03 | 80.76 0.02 | 77.32 0.02 | 79.24 0.03 | 78.19 0.02 |
| YOLOv5 (44) | 85.78 0.03 | 82.34 0.02 | 83.29 0.02 | 82.76 0.03 | 84.62 0.02 | 81.08 0.03 | 82.85 0.02 | 81.96 0.02 |
| RetinaNet (43) | 83.12 0.02 | 80.67 0.03 | 81.42 0.02 | 80.91 0.02 | 82.55 0.02 | 79.33 0.02 | 80.87 0.03 | 80.22 0.02 |
| DETR (41) | 86.59 0.03 | 83.78 0.02 | 85.14 0.03 | 84.43 0.02 | 85.23 0.02 | 82.14 0.03 | 83.79 0.02 | 83.05 0.02 |
| CornerNet (45) | 84.02 0.02 | 81.23 0.03 | 82.75 0.02 | 81.94 0.02 | 83.47 0.02 | 80.02 0.02 | 81.95 0.03 | 81.14 0.02 |
| SSD (46) | 82.11 0.03 | 79.14 0.02 | 80.68 0.03 | 79.90 0.02 | 80.88 0.02 | 77.76 0.02 | 79.84 0.03 | 78.71 0.02 |
| Ours | 90.23 0.02 | 87.89 0.02 | 89.45 0.03 | 88.76 0.02 | 88.94 0.02 | 86.45 0.02 | 87.92 0.03 | 87.21 0.02 |
The values in bold are the best values.