Table 5 Comparison with SOTA methods with proposed MSCAS model.
From: Transformer attention fusion for fine grained medical image classification
Dataset | Study | DR grading (classes) | Accuracy % |
|---|---|---|---|
APTOS 2019 | Mondal et al.7 | 5 | 86.08 |
Vijayan et al.15 | 86.20 | ||
Bodapati et al.35 | 84.17 | ||
Shaik and Cherukuri38 | 85.54 | ||
Proposed | 93.8 | ||
DDR | Zhao et al.9 | 5 | 83.10 |
Vijayan et al.15 | 84.80 | ||
Mubashra45 | 89.29 | ||
Oulhadj46 | 80.36 | ||
Proposed | 89.80 | ||
IDRID | Bodapati et al.35 | 5 | 63.24 |
Shaik and Cherukuri38 | 66.41 | ||
Jiwani50 | 77.60 | ||
Santos et al.51 | 77.50 | ||
Proposed | 86.70 |
