Deep learning for pneumothorax diagnosis: a systematic review and meta-analysis
Background Deep learning (DL), a subset of artificial intelligence (AI), has been applied to pneumothorax diagnosis to aid physician diagnosis, but no meta-analysis has been performed. Methods A search of multiple electronic databases through September 2022 was performed to identify studies that app...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
European Respiratory Society
2023-06-01
|
Series: | European Respiratory Review |
Online Access: | http://err.ersjournals.com/content/32/168/220259.full |
_version_ | 1797793381954355200 |
---|---|
author | Takahiro Sugibayashi Shannon L. Walston Toshimasa Matsumoto Yasuhito Mitsuyama Yukio Miki Daiju Ueda |
author_facet | Takahiro Sugibayashi Shannon L. Walston Toshimasa Matsumoto Yasuhito Mitsuyama Yukio Miki Daiju Ueda |
author_sort | Takahiro Sugibayashi |
collection | DOAJ |
description | Background
Deep learning (DL), a subset of artificial intelligence (AI), has been applied to pneumothorax diagnosis to aid physician diagnosis, but no meta-analysis has been performed.
Methods
A search of multiple electronic databases through September 2022 was performed to identify studies that applied DL for pneumothorax diagnosis using imaging. Meta-analysis via a hierarchical model to calculate the summary area under the curve (AUC) and pooled sensitivity and specificity for both DL and physicians was performed. Risk of bias was assessed using a modified Prediction Model Study Risk of Bias Assessment Tool.
Results
In 56 of the 63 primary studies, pneumothorax was identified from chest radiography. The total AUC was 0.97 (95% CI 0.96–0.98) for both DL and physicians. The total pooled sensitivity was 84% (95% CI 79–89%) for DL and 85% (95% CI 73–92%) for physicians and the pooled specificity was 96% (95% CI 94–98%) for DL and 98% (95% CI 95–99%) for physicians. More than half of the original studies (57%) had a high risk of bias.
Conclusions
Our review found the diagnostic performance of DL models was similar to that of physicians, although the majority of studies had a high risk of bias. Further pneumothorax AI research is needed. |
first_indexed | 2024-03-13T02:46:28Z |
format | Article |
id | doaj.art-80170a3090b743569c0de4213e5c93e1 |
institution | Directory Open Access Journal |
issn | 0905-9180 1600-0617 |
language | English |
last_indexed | 2024-03-13T02:46:28Z |
publishDate | 2023-06-01 |
publisher | European Respiratory Society |
record_format | Article |
series | European Respiratory Review |
spelling | doaj.art-80170a3090b743569c0de4213e5c93e12023-06-28T15:28:52ZengEuropean Respiratory SocietyEuropean Respiratory Review0905-91801600-06172023-06-013216810.1183/16000617.0259-20220259-2022Deep learning for pneumothorax diagnosis: a systematic review and meta-analysisTakahiro Sugibayashi0Shannon L. Walston1Toshimasa Matsumoto2Yasuhito Mitsuyama3Yukio Miki4Daiju Ueda5 Department of Diagnostic and Interventional Radiology, Graduate School of Medicine, Osaka Metropolitan University, Osaka, Japan Department of Diagnostic and Interventional Radiology, Graduate School of Medicine, Osaka Metropolitan University, Osaka, Japan Department of Diagnostic and Interventional Radiology, Graduate School of Medicine, Osaka Metropolitan University, Osaka, Japan Department of Diagnostic and Interventional Radiology, Graduate School of Medicine, Osaka Metropolitan University, Osaka, Japan Department of Diagnostic and Interventional Radiology, Graduate School of Medicine, Osaka Metropolitan University, Osaka, Japan Department of Diagnostic and Interventional Radiology, Graduate School of Medicine, Osaka Metropolitan University, Osaka, Japan Background Deep learning (DL), a subset of artificial intelligence (AI), has been applied to pneumothorax diagnosis to aid physician diagnosis, but no meta-analysis has been performed. Methods A search of multiple electronic databases through September 2022 was performed to identify studies that applied DL for pneumothorax diagnosis using imaging. Meta-analysis via a hierarchical model to calculate the summary area under the curve (AUC) and pooled sensitivity and specificity for both DL and physicians was performed. Risk of bias was assessed using a modified Prediction Model Study Risk of Bias Assessment Tool. Results In 56 of the 63 primary studies, pneumothorax was identified from chest radiography. The total AUC was 0.97 (95% CI 0.96–0.98) for both DL and physicians. The total pooled sensitivity was 84% (95% CI 79–89%) for DL and 85% (95% CI 73–92%) for physicians and the pooled specificity was 96% (95% CI 94–98%) for DL and 98% (95% CI 95–99%) for physicians. More than half of the original studies (57%) had a high risk of bias. Conclusions Our review found the diagnostic performance of DL models was similar to that of physicians, although the majority of studies had a high risk of bias. Further pneumothorax AI research is needed.http://err.ersjournals.com/content/32/168/220259.full |
spellingShingle | Takahiro Sugibayashi Shannon L. Walston Toshimasa Matsumoto Yasuhito Mitsuyama Yukio Miki Daiju Ueda Deep learning for pneumothorax diagnosis: a systematic review and meta-analysis European Respiratory Review |
title | Deep learning for pneumothorax diagnosis: a systematic review and meta-analysis |
title_full | Deep learning for pneumothorax diagnosis: a systematic review and meta-analysis |
title_fullStr | Deep learning for pneumothorax diagnosis: a systematic review and meta-analysis |
title_full_unstemmed | Deep learning for pneumothorax diagnosis: a systematic review and meta-analysis |
title_short | Deep learning for pneumothorax diagnosis: a systematic review and meta-analysis |
title_sort | deep learning for pneumothorax diagnosis a systematic review and meta analysis |
url | http://err.ersjournals.com/content/32/168/220259.full |
work_keys_str_mv | AT takahirosugibayashi deeplearningforpneumothoraxdiagnosisasystematicreviewandmetaanalysis AT shannonlwalston deeplearningforpneumothoraxdiagnosisasystematicreviewandmetaanalysis AT toshimasamatsumoto deeplearningforpneumothoraxdiagnosisasystematicreviewandmetaanalysis AT yasuhitomitsuyama deeplearningforpneumothoraxdiagnosisasystematicreviewandmetaanalysis AT yukiomiki deeplearningforpneumothoraxdiagnosisasystematicreviewandmetaanalysis AT daijuueda deeplearningforpneumothoraxdiagnosisasystematicreviewandmetaanalysis |