Accuracy of an automated system for tuberculosis detection on chest radiographs in high-risk screening
OBJECTIVE: To evaluate the performance of computerised detection software in triaging CXRs in a high-throughput digital mobile TB screening programme.
DESIGN: A retrospective evaluation of the software was performed on a database of 38 961 postero-anterior CXRs from unique individuals seen between 2005 and 2010, 87 of whom were diagnosed with TB. The software generated a TB likelihood score for each CXR. This score was compared with a reference standard for notified active pulmonary TB using receiver operating characteristic (ROC) curve and localisation ROC (LROC) curve analyses.
RESULTS: On ROC curve analysis, software specificity was 55.71% (95%CI 55.21–56.20) and negative predictive value was 99.98% (95%CI 99.95–99.99), at a sensitivity of 95%. The area under the ROC curve was 0.90 (95%CI 0.86–0.93). Results of the LROC curve analysis were similar.
CONCLUSION: The software could identify more than half of the normal images in a TB screening setting while maintaining high sensitivity, and may therefore be used for triage.
Keywords: TB; chest radiography; computer-aided detection; computerised image analysis
Document Type: Research Article
Affiliations: 1: Diagnostic Image Analysis Group, Radboud University Medical Center, Nijmegen, Thirona, Nijmegen, The Netherlands 2: Diagnostic Image Analysis Group, Radboud University Medical Center, Nijmegen 3: Department of Infectious Disease Informatics, Institute of Health Informatics, University College London, London 4: Department of Infectious Disease Informatics, Institute of Health Informatics, University College London, London, Institute of Epidemiology and Health Care, University College London 5: Institute for Global Health, University College London, UK
Publication date: 01 May 2018
The International Journal of Tuberculosis and Lung Disease (IJTLD) is for clinical research and epidemiological studies on lung health, including articles on TB, TB-HIV and respiratory diseases such as COVID-19, asthma, COPD, child lung health and the hazards of tobacco and air pollution. Individuals and institutes can subscribe to the IJTLD online or in print – simply email us at [email protected] for details.
The IJTLD is dedicated to understanding lung disease and to the dissemination of knowledge leading to better lung health. To allow us to share scientific research as rapidly as possible, the IJTLD is fast-tracking the publication of certain articles as preprints prior to their publication. Read fast-track articles.
- Editorial Board
- Information for Authors
- Subscribe to this Title
- International Journal of Tuberculosis and Lung Disease
- Public Health Action
- Ingenta Connect is not responsible for the content or availability of external websites
Receive new issue alert- Access Key
- Free content
- Partial Free content
- New content
- Open access content
- Partial Open access content
- Subscribed content
- Partial Subscribed content
- Free trial content