Cargando…

Development and Validation of a Deep Learning–based Automatic Detection Algorithm for Active Pulmonary Tuberculosis on Chest Radiographs

BACKGROUND: Detection of active pulmonary tuberculosis on chest radiographs (CRs) is critical for the diagnosis and screening of tuberculosis. An automated system may help streamline the tuberculosis screening process and improve diagnostic performance. METHODS: We developed a deep learning–based au...

Descripción completa

Detalles Bibliográficos
Autores principales: Hwang, Eui Jin, Park, Sunggyun, Jin, Kwang-Nam, Kim, Jung Im, Choi, So Young, Lee, Jong Hyuk, Goo, Jin Mo, Aum, Jaehong, Yim, Jae-Joon, Park, Chang Min
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Oxford University Press 2019
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6695514/
https://www.ncbi.nlm.nih.gov/pubmed/30418527
http://dx.doi.org/10.1093/cid/ciy967
_version_ 1783444055815880704
author Hwang, Eui Jin
Park, Sunggyun
Jin, Kwang-Nam
Kim, Jung Im
Choi, So Young
Lee, Jong Hyuk
Goo, Jin Mo
Aum, Jaehong
Yim, Jae-Joon
Park, Chang Min
author_facet Hwang, Eui Jin
Park, Sunggyun
Jin, Kwang-Nam
Kim, Jung Im
Choi, So Young
Lee, Jong Hyuk
Goo, Jin Mo
Aum, Jaehong
Yim, Jae-Joon
Park, Chang Min
author_sort Hwang, Eui Jin
collection PubMed
description BACKGROUND: Detection of active pulmonary tuberculosis on chest radiographs (CRs) is critical for the diagnosis and screening of tuberculosis. An automated system may help streamline the tuberculosis screening process and improve diagnostic performance. METHODS: We developed a deep learning–based automatic detection (DLAD) algorithm using 54c221 normal CRs and 6768 CRs with active pulmonary tuberculosis that were labeled and annotated by 13 board-certified radiologists. The performance of DLAD was validated using 6 external multicenter, multinational datasets. To compare the performances of DLAD with physicians, an observer performance test was conducted by 15 physicians including nonradiology physicians, board-certified radiologists, and thoracic radiologists. Image-wise classification and lesion-wise localization performances were measured using area under the receiver operating characteristic (ROC) curves and area under the alternative free-response ROC curves, respectively. Sensitivities and specificities of DLAD were calculated using 2 cutoffs (high sensitivity [98%] and high specificity [98%]) obtained through in-house validation. RESULTS: DLAD demonstrated classification performance of 0.977–1.000 and localization performance of 0.973–1.000. Sensitivities and specificities for classification were 94.3%–100% and 91.1%–100% using the high-sensitivity cutoff and 84.1%–99.0% and 99.1%–100% using the high-specificity cutoff. DLAD showed significantly higher performance in both classification (0.993 vs 0.746–0.971) and localization (0.993 vs 0.664–0.925) compared to all groups of physicians. CONCLUSIONS: Our DLAD demonstrated excellent and consistent performance in the detection of active pulmonary tuberculosis on CR, outperforming physicians, including thoracic radiologists.
format Online
Article
Text
id pubmed-6695514
institution National Center for Biotechnology Information
language English
publishDate 2019
publisher Oxford University Press
record_format MEDLINE/PubMed
spelling pubmed-66955142019-08-21 Development and Validation of a Deep Learning–based Automatic Detection Algorithm for Active Pulmonary Tuberculosis on Chest Radiographs Hwang, Eui Jin Park, Sunggyun Jin, Kwang-Nam Kim, Jung Im Choi, So Young Lee, Jong Hyuk Goo, Jin Mo Aum, Jaehong Yim, Jae-Joon Park, Chang Min Clin Infect Dis Articles and Commentaries BACKGROUND: Detection of active pulmonary tuberculosis on chest radiographs (CRs) is critical for the diagnosis and screening of tuberculosis. An automated system may help streamline the tuberculosis screening process and improve diagnostic performance. METHODS: We developed a deep learning–based automatic detection (DLAD) algorithm using 54c221 normal CRs and 6768 CRs with active pulmonary tuberculosis that were labeled and annotated by 13 board-certified radiologists. The performance of DLAD was validated using 6 external multicenter, multinational datasets. To compare the performances of DLAD with physicians, an observer performance test was conducted by 15 physicians including nonradiology physicians, board-certified radiologists, and thoracic radiologists. Image-wise classification and lesion-wise localization performances were measured using area under the receiver operating characteristic (ROC) curves and area under the alternative free-response ROC curves, respectively. Sensitivities and specificities of DLAD were calculated using 2 cutoffs (high sensitivity [98%] and high specificity [98%]) obtained through in-house validation. RESULTS: DLAD demonstrated classification performance of 0.977–1.000 and localization performance of 0.973–1.000. Sensitivities and specificities for classification were 94.3%–100% and 91.1%–100% using the high-sensitivity cutoff and 84.1%–99.0% and 99.1%–100% using the high-specificity cutoff. DLAD showed significantly higher performance in both classification (0.993 vs 0.746–0.971) and localization (0.993 vs 0.664–0.925) compared to all groups of physicians. CONCLUSIONS: Our DLAD demonstrated excellent and consistent performance in the detection of active pulmonary tuberculosis on CR, outperforming physicians, including thoracic radiologists. Oxford University Press 2019-09-01 2018-11-08 /pmc/articles/PMC6695514/ /pubmed/30418527 http://dx.doi.org/10.1093/cid/ciy967 Text en © The Author(s) 2018. Published by Oxford University Press for the Infectious Diseases Society of America. http://creativecommons.org/licenses/by/4.0/ This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
spellingShingle Articles and Commentaries
Hwang, Eui Jin
Park, Sunggyun
Jin, Kwang-Nam
Kim, Jung Im
Choi, So Young
Lee, Jong Hyuk
Goo, Jin Mo
Aum, Jaehong
Yim, Jae-Joon
Park, Chang Min
Development and Validation of a Deep Learning–based Automatic Detection Algorithm for Active Pulmonary Tuberculosis on Chest Radiographs
title Development and Validation of a Deep Learning–based Automatic Detection Algorithm for Active Pulmonary Tuberculosis on Chest Radiographs
title_full Development and Validation of a Deep Learning–based Automatic Detection Algorithm for Active Pulmonary Tuberculosis on Chest Radiographs
title_fullStr Development and Validation of a Deep Learning–based Automatic Detection Algorithm for Active Pulmonary Tuberculosis on Chest Radiographs
title_full_unstemmed Development and Validation of a Deep Learning–based Automatic Detection Algorithm for Active Pulmonary Tuberculosis on Chest Radiographs
title_short Development and Validation of a Deep Learning–based Automatic Detection Algorithm for Active Pulmonary Tuberculosis on Chest Radiographs
title_sort development and validation of a deep learning–based automatic detection algorithm for active pulmonary tuberculosis on chest radiographs
topic Articles and Commentaries
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6695514/
https://www.ncbi.nlm.nih.gov/pubmed/30418527
http://dx.doi.org/10.1093/cid/ciy967
work_keys_str_mv AT hwangeuijin developmentandvalidationofadeeplearningbasedautomaticdetectionalgorithmforactivepulmonarytuberculosisonchestradiographs
AT parksunggyun developmentandvalidationofadeeplearningbasedautomaticdetectionalgorithmforactivepulmonarytuberculosisonchestradiographs
AT jinkwangnam developmentandvalidationofadeeplearningbasedautomaticdetectionalgorithmforactivepulmonarytuberculosisonchestradiographs
AT kimjungim developmentandvalidationofadeeplearningbasedautomaticdetectionalgorithmforactivepulmonarytuberculosisonchestradiographs
AT choisoyoung developmentandvalidationofadeeplearningbasedautomaticdetectionalgorithmforactivepulmonarytuberculosisonchestradiographs
AT leejonghyuk developmentandvalidationofadeeplearningbasedautomaticdetectionalgorithmforactivepulmonarytuberculosisonchestradiographs
AT goojinmo developmentandvalidationofadeeplearningbasedautomaticdetectionalgorithmforactivepulmonarytuberculosisonchestradiographs
AT aumjaehong developmentandvalidationofadeeplearningbasedautomaticdetectionalgorithmforactivepulmonarytuberculosisonchestradiographs
AT yimjaejoon developmentandvalidationofadeeplearningbasedautomaticdetectionalgorithmforactivepulmonarytuberculosisonchestradiographs
AT parkchangmin developmentandvalidationofadeeplearningbasedautomaticdetectionalgorithmforactivepulmonarytuberculosisonchestradiographs
AT developmentandvalidationofadeeplearningbasedautomaticdetectionalgorithmforactivepulmonarytuberculosisonchestradiographs