Table 3 Tuberculosis detection. The datasets, manifestations, assessment measures and results are shown in each column, respectively
From: Computer-aided detection in chest radiography based on artificial intelligence: a survey
Study | Datasets | Manifestations | Assessment measures | Results |
|---|---|---|---|---|
Rohmah et al. [80] | Custom (120) | Tuberculosis | Accuracy, false accept rate, false rejection rate | 95.7%, 3.33%, 6.67% |
Tan et al. [81] | Custom (95) | Tuberculosis | Accuracy, sensitivity, specificity, AUC, precision | 92.9%, 91%, 95.4%, 92.8%, 94.9% |
Noor et al. [82] | TPR (100) | Tuberculosis | Accuracy | 94% |
Song et al. [87] | Custom (200) | Focal opacities | Accuracy | Â |
Shen et al. [88] | Custom (243) | Cavities | True positive rate (or sensitivity) under FP/image | 0.237 FP/image: 82.35% |
Xu et al. [89] | Custom | Cavities | Densitivity, specificity, and accuracy | E-Group: 78.8%, 86.8%, 82.8%; D-Group: 69.4%, 81.6%, 75.5% |
Hwang et al. [90] | KIT, MC, Shenzhen | Tuberculosis | AUC, accuracy, positive precision, negative precision | 96.4%, 90.3%, 95.3%, 97.4% |
Lakhani et al. [91] | Shenzhen | Tuberculosis | AUC, sensitivity, and specificity | 99%, 97.3%, 100% |