Table 2 Classification module mean performance on five folds for thin smears compared to the state-of-the-art
From: Malaria Screener: a smartphone application for automated malaria screening
| Â | Accuracy | AUC | Sensitivity | Specificity | F1-score |
|---|---|---|---|---|---|
Proposed Module (Patch-level) [14] | 98.6% | 99.9% | 98.1% | 99.2% | 98.7% |
Proposed Module (Patient-level) [14] | 95.9% | 99.1% | 94.7% | 97.2% | 95.9% |
Gopakumar et al. (2018) [5] | 97.7% | – | 97.1% | 98.5% | – |
Bibin, Nair & Punitha (2017) [22] | 96.3% | – | 97.6% | 95.9% | – |
Dong et al. (2017) [23] | 98.1% | – | – | – | – |
Liang et al. (2017) [24] | 97.3% | – | 96.9% | 97.7% | – |
Das et al. (2013) [25] | 84.0% | – | 98.1% | 68.9% | – |
Ross et al. (2006) [26] | 73.0% | – | 85.0% | – | – |