Diagnostic accuracy of computer-aided detection of pulmonary tuberculosis in chest radiographs: a validation study from sub-Saharan Africa

Abstract

Background: Chest radiography to diagnose and screen for pulmonary tuberculosis has limitations, especially due to inter-reader variability. Automating the interpretation has the potential to overcome this drawback and to deliver objective and reproducible results. The CAD4TB software is a computer-aided detection system that has shown promising preliminary findings. Evaluation studies in different settings are needed to assess diagnostic accuracy and practicability of use.

Methods: CAD4TB was evaluated on chest radiographs of patients with symptoms suggestive of pulmonary tuberculosis enrolled in two cohort studies in Tanzania. All patients were characterized by sputum smear microscopy and culture including subsequent antigen or molecular confirmation of Mycobacterium tuberculosis (M.tb) to determine the reference standard. Chest radiographs were read by the software and two human readers, one expert reader and one clinical officer. The sensitivity and specificity of CAD4TB was depicted using receiver operating characteristic (ROC) curves, the area under the curve calculated and the performance of the software compared to the results of human readers.

Results: Of 861 study participants, 194 (23%) were culture-positive for M.tb. The area under the ROC curve of CAD4TB for the detection of culture-positive pulmonary tuberculosis was 0.84 (95% CI 0.80-0.88). CAD4TB was significantly more accurate for the discrimination of smear-positive cases against non TB patients than for smear-negative cases (p-value<0.01). It differentiated better between TB cases and non TB patients among HIV-negative compared to HIV-positive individuals (p<0.01). CAD4TB significantly outperformed the clinical officer, but did not reach the accuracy of the expert reader (p = 0.02), for a tuberculosis specific reading threshold.

Conclusion: CAD4TB accurately distinguished between the chest radiographs of culture-positive TB cases and controls. Further studies on cost-effectiveness, operational and ethical aspects should determine its place in diagnostic and screening algorithms.

Figure 1. Flow chart of individuals taking part in the study.

Figure 2. Distribution of CAD scores for patient groups A (s+/c+ M.tb), B (s−/c+ M.tb), C (s±/c+ NTM), D (s−/c− clin.TB) and F (Controls).

Figure 3. ROC analysis for the detection of M.tb culture-positive individuals.

Legend. —–A (s+/c+ M.tb), B (s−/c+ M.tb) vs. F (Controls): A_z = 0.84 (0.80–0.88), - - - A (s+/c+ M.tb), B (s−/c+ M.tb) vs. C (s±/c+ NTM), E (EPTB), F (Controls): A_z = 0.81 (0.77–0.85), p = 0.28.

Figure 4. ROC analysis for the detection of M.tb culture-positive individuals by smear status.

Legend. —– A (s+/c+ M.tb) vs. F (Controls): A_z = 0.90 (0.86–0.93), - - - B (s−/c+ M.tb) vs. F (Controls): A_z = 0.67 (0.58–0.75), p<0.01.

Figure 5. ROC analysis for the detection of M.tb culture-positive individuals by HIV Status.

Legend. —– HIV negative. A (s+/c+ M.tb), B (s−/c+ M.tb) vs. F (Controls): A_z = 0.89 (0.85–0.94), - - - HIV positive. A (s+/c+ M.tb), B (s−/c+ M.tb) vs. F (Controls): A_z = 0.79 (0.72–0.86), p<0.01.

Figure 6. Comparison of automated and human reading.

Legend. Sensitivity and specificity to distinguish group A (s+/c+ M.tb) and B (s−/c+ M.tb) vs. F (Controls). Line and shaded area: ROC curve and 95% CI for CAD4TB. The expert reader is represented by square symbols, the clinical officer by round symbols. The different fill of the symbols indicate different reading thresholds: empty symbols  = ‘any abnormality’, crossed symbols  = ‘TB consistent abnormalities’ and filled symbols  = ‘abnormalities highly suggestive for TB’.