| | CXR | DTS |
---|
a |
Reader 1 | Sensitivity (%) | 24 (34/144) | 80 (116/144) |
Specificity (%) | 10 (33/321) | 95 (308/321) |
Accuracy (%) | 15 (69/465) | 91 (424/465) |
Diagnostic confidence: |
| Number | TP | TN | FP | FN | Number | TP | TN | FP | FN |
Score 1 | 1 | / | 1 | / | / | 295 | / | 295 | / | / |
Score 2 | 35 | / | 32 | / | 3 | 15 | / | 13 | / | 2 |
Score 3 | 379 | / | / | 303 | 73 | 17a | / | / | 10 | 7 |
Score 4 | 46 | 34 | / | 12 | / | 51 | 32 | / | 19 | / |
Score 5 | 4 | 0 | / | 4 | / | 87 | 84 | / | 3 | / |
(AUC) (95 % CI) | 0.571 (0.525–0.616) | 0.948 (0.924–0.967) |
b |
Reader 2 | Sensitivity (%) | 17 (25/144) | 85 (122/144) |
Specificity (%) | 13 (43/321) | 95 (308/321) |
Accuracy (%) | 17 (78/465) | 92 (430/465) |
Diagnostic confidence: |
| Number | TP | TN | FP | FN | Number | TP | TN | FP | |
FN |
Score 1 | 2 | / | 1 | / | 1 | 297 | / | 296 | / | |
Score 2 | 45 | / | 42 | / | 3 | 13 | / | 12 | / | |
Score 3 | 386 | / | / | 304 | 82 | 17a | / | / | 11 | |
Score 4 | 31 | 25 | / | 6 | / | 31 | 22 | / | 9 | |
Score 5 | 1 | 0 | / | 1 | / | 107 | 102 | / | 5 | |
(AUC) (95 % CI) | 0.612 (0.566–0.656) | 0.947 (0.923–0.966) |
- Visual prospective analysis in the pulmonary lesion diagnosis
- The confidence scoring system is reported in Table 1
- -CXR chest radiography; DTS digital tomosynthesis; TP true positive [lesion correctly assessed as a non-calcified pulmonary lesion (confidence score 4 or 5) or a lesion appearing as a parenchymal or ground-glass opacity, or a solid or subsolid ground-glass pulmonary nodule]; TN true negative (benign pulmonary lesion-centrally calcified lesion or lesion with gross calcifications or calcified fibrotic scars with pulmonary architectural distortion-or extra-pulmonary lesion or as a pulmonary pseudolesion; confidence levels 1, 2); FP false positive (benign pulmonary or extra-pulmonary lesion, or a pulmonary pseudolesion incorrectly assessed as a pulmonary lesion) (confidence levels 4 or 5) or assessed as indeterminate (confidence score 3); FN false negative (pulmonary lesion which should deserve further CT assessment assessed as an overt benign pulmonary lesion or pseudolesion or extrapulmonary lesion) (confidence levels 1 or 2) or assessed as indeterminate (confidence score 3); NPV negative predictive value; PPV positive predictive value. AUC area under the receiver-operating characteristic curve; CI confidence interval
- Sensitivity was defined as TP/(TP + FN); specificity, as TN/(TN + FP): accuracy, as. (TP + TN)/(TP + TN + FP + FN)
- All differences between chest radiography and digital tomosynthesis were statistically significant (P < 0.05)
- aIncluding 7 (reader 1) or 6 (reader 2) solid subpleural pulmonary lesions and 10 (reader 1) or 11 (reader 2) pseudolesions