Table 4 Performance metrics of state-of-the-art techniques for anemia screening in conjunctiva/palpebral images classification.
Author | Year | Main technique | Accuracy (%) |
|---|---|---|---|
Tamir et al.39 | 2017 | Color thresholding | 78.90 |
Sevani et al.40 | 2018 | K-means clustering | 90.00 |
Jain et al.42 | 2020 | ANN | 97.00 |
Asiyah et al.43 | 2022 | PCA and K-NN | 87.50 |
Appiahene et al.37 | 2023 | CNN + Logistic regression | 92.50 |
Dimauro et al.46 | 2023 | MobileNetV2 | 91.00 |
Purwanti et al.45 | 2023 | ResNet-50 | 93.70 |
Bhusham et al.47 | 2023 | DenseNet-201 | 93.70 |
Sehar et al.38 | 2024 | Multiple regression model | 83.30 |
Priyadarshini et al.41 | 2024 | XGBoost | 93.00 |
Mythily et al.44 | 2024 | LBP + SVM | 93.00 |
Muljono et al.48 | 2024 | SVM + MobileNetV2 | 93.00 |
ViT (with 1k TL) | 2024 | ViT | 95.42 |
ViT (with 21k TL) | 2024 | ViT | 98.47 |
