Research Article
Two-Stage Hybrid Approach of Deep Learning Networks for Interstitial Lung Disease Classification
Table 5
Comparison of the proposed classifier with the earlier CNN-based classifiers.
| Method | Image input type | -score (%) | Accuracy (%) |
| Li et al. [17] | ROI patch | 66.57 | 67.05 | LeNet [37] | ROI patch | 67.83 | 67.90 | AlexNet [38] | ROI patch | 70.31 | 71.04 | Pretrained AlexNet [38] | ROI patch | 75.82 | 76.09 | VGGNet [10] | ROI patch | 78.04 | 78.00 | Doddavarapu et al. [21] | ROI patch | 94.65 | 94.67 | Gao et al. [26] | Whole HRCT | 66.83 | 69.23 | Proposed two-stage hybrid classifier | Segmented HRCT | 89.45 | 89.39 |
|
|