Chest x-ray analysis with deep learning-based software as a triage test for pulmonary tuberculosis: a prospective study of diagnostic accuracy for culture-confirmed disease.
Adult
Age Factors
Body Mass Index
Deep Learning
Female
HIV Infections
Humans
Image Processing, Computer-Assisted
/ methods
Lung
/ diagnostic imaging
Male
Middle Aged
Mycobacterium tuberculosis
/ growth & development
Pakistan
Prospective Studies
Radiology
/ methods
Sensitivity and Specificity
Sex Factors
Software
Sputum
/ microbiology
Triage
Tuberculosis, Pulmonary
/ diagnosis
X-Rays
Young Adult
Journal
The Lancet. Digital health
ISSN: 2589-7500
Titre abrégé: Lancet Digit Health
Pays: England
ID NLM: 101751302
Informations de publication
Date de publication:
11 2020
11 2020
Historique:
received:
16
06
2020
revised:
20
08
2020
accepted:
27
08
2020
entrez:
17
12
2020
pubmed:
18
12
2020
medline:
28
1
2021
Statut:
ppublish
Résumé
Deep learning-based radiological image analysis could facilitate use of chest x-rays as triage tests for pulmonary tuberculosis in resource-limited settings. We sought to determine whether commercially available chest x-ray analysis software meet WHO recommendations for minimal sensitivity and specificity as pulmonary tuberculosis triage tests. We recruited symptomatic adults at the Indus Hospital, Karachi, Pakistan. We compared two software, qXR version 2.0 (qXRv2) and CAD4TB version 6.0 (CAD4TBv6), with a reference of mycobacterial culture of two sputa. We assessed qXRv2 using its manufacturer prespecified threshold score for chest x-ray classification as tuberculosis present versus not present. For CAD4TBv6, we used a data-derived threshold, because it does not have a prespecified one. We tested for non-inferiority to preset WHO recommendations (0·90 for sensitivity, 0·70 for specificity) using a non-inferiority limit of 0·05. We identified factors associated with accuracy by stratification and logistic regression. We included 2198 (92·7%) of 2370 enrolled participants. 2187 (99·5%) of 2198 were HIV-negative, and 272 (12·4%) had culture-confirmed pulmonary tuberculosis. For both software, accuracy was non-inferior to WHO-recommended minimum values (qXRv2 sensitivity 0·93 [95% CI 0·89-0·95], non-inferiority p=0·0002; CAD4TBv6 sensitivity 0·93 [0·90-0·96], p<0·0001; qXRv2 specificity 0·75 [0·73-0·77], p<0·0001; CAD4TBv6 specificity 0·69 [0·67-0·71], p=0·0003). Sensitivity was lower in smear-negative pulmonary tuberculosis for both software, and in women for CAD4TBv6. Specificity was lower in men and in those with previous tuberculosis, and reduced with increasing age and decreasing body mass index. Smoking and diabetes did not affect accuracy. In an HIV-negative population, these software met WHO-recommended minimal accuracy for pulmonary tuberculosis triage tests. Sensitivity will be lower when smear-negative pulmonary tuberculosis is more prevalent. Canadian Institutes of Health Research.
Sections du résumé
BACKGROUND
Deep learning-based radiological image analysis could facilitate use of chest x-rays as triage tests for pulmonary tuberculosis in resource-limited settings. We sought to determine whether commercially available chest x-ray analysis software meet WHO recommendations for minimal sensitivity and specificity as pulmonary tuberculosis triage tests.
METHODS
We recruited symptomatic adults at the Indus Hospital, Karachi, Pakistan. We compared two software, qXR version 2.0 (qXRv2) and CAD4TB version 6.0 (CAD4TBv6), with a reference of mycobacterial culture of two sputa. We assessed qXRv2 using its manufacturer prespecified threshold score for chest x-ray classification as tuberculosis present versus not present. For CAD4TBv6, we used a data-derived threshold, because it does not have a prespecified one. We tested for non-inferiority to preset WHO recommendations (0·90 for sensitivity, 0·70 for specificity) using a non-inferiority limit of 0·05. We identified factors associated with accuracy by stratification and logistic regression.
FINDINGS
We included 2198 (92·7%) of 2370 enrolled participants. 2187 (99·5%) of 2198 were HIV-negative, and 272 (12·4%) had culture-confirmed pulmonary tuberculosis. For both software, accuracy was non-inferior to WHO-recommended minimum values (qXRv2 sensitivity 0·93 [95% CI 0·89-0·95], non-inferiority p=0·0002; CAD4TBv6 sensitivity 0·93 [0·90-0·96], p<0·0001; qXRv2 specificity 0·75 [0·73-0·77], p<0·0001; CAD4TBv6 specificity 0·69 [0·67-0·71], p=0·0003). Sensitivity was lower in smear-negative pulmonary tuberculosis for both software, and in women for CAD4TBv6. Specificity was lower in men and in those with previous tuberculosis, and reduced with increasing age and decreasing body mass index. Smoking and diabetes did not affect accuracy.
INTERPRETATION
In an HIV-negative population, these software met WHO-recommended minimal accuracy for pulmonary tuberculosis triage tests. Sensitivity will be lower when smear-negative pulmonary tuberculosis is more prevalent.
FUNDING
Canadian Institutes of Health Research.
Identifiants
pubmed: 33328086
pii: S2589-7500(20)30221-1
doi: 10.1016/S2589-7500(20)30221-1
pii:
doi:
Types de publication
Evaluation Study
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
e573-e581Commentaires et corrections
Type : ErratumIn
Informations de copyright
Copyright © 2020 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY-NC-ND 4.0 license. Published by Elsevier Ltd.. All rights reserved.