Skip to main content
SpringerLink
Log in

A comprehensive segmentation of chest X-ray improves deep learning–based WHO radiologically confirmed pneumonia diagnosis in children

  • Chest
  • Published:
European Radiology Aims and scope Submit manuscript

Abstract

Objectives

To investigate a comprehensive segmentation of chest X-ray (CXR) in promoting deep learning–based World Health Organization’s (WHO) radiologically confirmed pneumonia diagnosis in children.

Methods

A total of 4400 participants between January 2016 and June 2021were identified for a cross-sectional study and divided into primary endpoint pneumonia (PEP), other infiltrates, and normal groups according to WHO’s diagnostic criteria. The CXR was divided into six segments of left lung, right lung, mediastinum, diaphragm, ext-left lung, and ext-right lung by adopting the RA-UNet. To demonstrate the benefits of lung field segmentation in pneumonia diagnosis, the segmented images and images that were not segmented, which constituted seven segmentation combinations, were fed into the CBAM-ResNet under a three-category classification comparison. The interpretability of the CBAM-ResNet for pneumonia diagnosis was also performed by adopting a Grad-CAM module.

Results

The RA-UNet achieved a high spatial overlap between manual and automatic segmentation (averaged DSC = 0.9639). The CBAM-ResNet when fed with the six segments achieved superior three-category diagnosis performance (accuracy = 0.8243) over other segmentation combinations and deep learning models under comparison, which was increased by around 6% in accuracy, precision, specificity, sensitivity, F1-score, and around 3% in AUC. The Grad-CAM could capture the pneumonia lesions more accurately, generating a more interpretable visualization and enhancing the superiority and reliability of our study in assisting pediatric pneumonia diagnosis.

Conclusions

The comprehensive segmentation of CXR could improve deep learning–based pneumonia diagnosis in childhood with a more reasonable WHO’s radiological standardized pneumonia classification instead of conventional dichotomous bacterial pneumonia and viral pneumonia.

Clinical relevance statement

The comprehensive segmentation of chest X-ray improves deep learning–based WHO confirmed pneumonia diagnosis in children, laying a strong foundation for the potential inclusion of computer-aided pediatric CXR readings in precise classification of pneumonia and PCV vaccine trials efficacy in children.

Key Points

• The chest X-ray was comprehensively segmented into six anatomical structures of left lung, right lung, mediastinum, diaphragm, ext-left lung, and ext-right lung.

• The comprehensive segmentation improved the three-category classification of primary endpoint pneumonia, other infiltrates, and normal with an increase by around 6% in accuracy, precision, specificity, sensitivity, F1-score, and around 3% in AUC.

• The comprehensive segmentation gave rise to a more accurate and interpretable visualization results in capturing the pneumonia lesions.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6

Similar content being viewed by others

Abbreviations

2DAN:

2D attention network

AUC:

Area under receiver operating characteristic curve

CBAM:

Convolutional block attention module

COM:

Combination

CXR:

Chest X-ray

DC-UNet:

Dual channel U-Net

DR:

Digital radiography

DSC:

Dice similarity coefficient

ext-left lung:

Extensive left lung area

ext-right lung:

Extensive right lung area

FN:

False negative

FOV:

Field-of-view

FP:

False positive

Grad-CAM:

Gradient-weighted Class Activation Mapping

PCV:

Pneumococcal conjugate vaccine

PEP:

Primary endpoint pneumonia

RAN:

Residual attention network

RA-UNet:

Residual attention-UNet

ResNet:

Residual neural network

ResNet-50:

Residual network with 50 layers

ResUNet:

Residual U-Net

TN:

True negative

TP:

True positive

TransUNet:

Transformer-based U-Net

WHO:

World Health Organization

References

  1. Black RE, Cousens S, Johnson HL et al (2010) Global, regional, and national causes of child mortality in 2008: a systematic analysis. Lancet 375:1969–1987

    Article  PubMed  Google Scholar 

  2. Zhao B, Guo Y, Zheng C et al (2019) Using deep-learning techniques for pulmonary-thoracic segmentations and improvement of pneumonia diagnosis in pediatric chest radiographs. Pediatr Pulmonol 54:1617–1626

    Article  PubMed  Google Scholar 

  3. Hwang S, Park S (2017) Accurate lung segmentation via network-wise training of convolutional networks. Deep Learning in Medical lmage Analysis and Multimodal learning for Clinical Decision Support. Springer, pp 92–99

  4. Mansoor A, Cerrolaza JJ, Perez G et al (2019) A generic approach to lung field segmentation from chest radiographs using deep space and shape learning. IEEE Trans Biomed Eng 67:1206–1220

    Article  PubMed  PubMed Central  Google Scholar 

  5. Garin M, Carballo DF, Montet R (2012) High discordance of chest x-ray and CT for detection of pulmonary opacities in ED patients: implications for diagnosing pneumonia. Am J Respir Crit Care Med 31:10.1164

  6. Ferreira JR, Cardenas DAC, Moreno RA, de Sá Rebelo MdF, Krieger JE, Gutierrez MA (2020) Multi-view ensemble convolutional neural network to improve classification of pneumonia in low contrast chest x-ray images2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC). IEEE, pp 1238–1241

  7. Longjiang E, Zhao B, Liu H et al (2021) Image-based deep learning in diagnosing the etiology of pneumonia on pediatric chest X-rays. Pediatric Pulmonol 56:1036–1044

    Article  Google Scholar 

  8. Mohammed I, Singh N, Venkatasubramanian M (2019) Computer-assisted detection and diagnosis of pediatric pneumonia in chest X-ray images. Available via https://www.patterncomputer.com/wpcontent/uploads/2022/02/Computer-Assisted-Detection-and-Diagnosis-of-Pediatric-Pneumonia-in-Chest-X-ray-Images.pdf

  9. Beadling C, Slifka MK (2004) How do viral infections predispose patients to bacterial infections? Curr Opin Infect Dis 17:185–191

    Article  PubMed  Google Scholar 

  10. Pavia AT (2013) What is the role of respiratory viruses in community-acquired pneumonia?: What is the best therapy for influenza and other viral causes of community-acquired pneumonia? Infect Dis Clin 27:157–175

    Article  Google Scholar 

  11. Bosch AA, Biesbroek G, Trzcinski K, Sanders EA, Bogaert D (2013) Viral and bacterial interactions in the upper respiratory tract. PLoS Pathog 9:e1003057

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Cherian T, Mulholland EK, Carlin JB et al (2005) Standardized interpretation of paediatric chest radiographs for the diagnosis of pneumonia in epidemiological studies. Bull World Health Organ 83:353–359

    PubMed  PubMed Central  Google Scholar 

  13. Fernandes V, Junior GB, de Paiva AC, Silva AC, Gattass M (2021) Bayesian convolutional neural network estimation for pediatric pneumonia detection and diagnosis. Comput Methods Programs Biomed 208:106259

    Article  PubMed  Google Scholar 

  14. Selvaraju RR, Cogswell M, Das A, Vedantam R, Parikh D, Batra D (2017) Grad-cam: Visual explanations from deep networks via gradient-based localizationProceedings of the IEEE international conference on computer vision, pp 618–626

  15. Jin Q, Meng Z, Sun C, Cui H, Su R (2020) RA-UNet: a hybrid deep attention-aware network to extract liver and tumor in CT scans. Front Bioeng Biotechnol 8:1471

  16. Woo S, Park J, Lee J-Y, Kweon IS (2018) Cbam: convolutional block attention module. Proceedings of the European conference on computer vision (ECCV), pp 3–19

  17. He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. Proceedings of the IEEE conference on computer vision and pattern recognition, pp 770–778

  18. Chen J, Lu Y, Yu Q et al (2021) Transunet: transformers make strong encoders for medical image segmentation. arXiv preprint arXiv:210204306

  19. Zhang Z, Liu Q, Wang Y (2018) Road extraction by deep residual u-net. IEEE Geosci Remote Sens Lett 15:749–753

    Article  CAS  Google Scholar 

  20. Lou A, Guan S, Loew M (2021) DC-UNet: rethinking the U-Net architecture with dual channel efficient CNN for medical image segmentation Medical Imaging 2021: Image Processing. SPIE, pp 758–768

  21. Wang F, Jiang M, Qian C et al (2017) Residual attention network for image classification. Proceedings of the IEEE conference on computer vision and pattern recognition, pp 3156–3164

  22. Jin D, Zhou B, Han Y et al (2020) Generalizable, reproducible, and neuroscientifically interpretable imaging biomarkers for Alzheimer’s disease. Adv Sci 7:2000675

    Article  CAS  Google Scholar 

  23. Piryonesi SM, El-Diraby TE (2020) Data analytics in asset management: cost-effective prediction of the pavement condition index. J Infrastruct Syst 26:04019036

    Article  Google Scholar 

  24. Ibrahim AU, Ozsoz M, Serte S, Al-Turjman F, Yakoi PS (2021) Pneumonia classification using deep learning from chest X-ray images during COVID-19. Cogn Comput 4:1–13

  25. Zhang X, Han L, Sobeih T et al (2021) CXR-Net: an encoder-decoder-encoder multitask deep neural network for explainable and accurate diagnosis of COVID-19 pneumonia with chest X-ray Images. arXiv preprint arXiv:211010813

  26. Tiwari A, Sharan TS, Sharma S, Sharma N (2022) Deep learning-based automated multiclass classification of chest X-rays into Covid-19, normal, bacterial pneumonia and viral pneumonia. Cogent Eng 9:2105559

    Article  Google Scholar 

  27. Mahomed N, van Ginneken B, Philipsen RH et al (2020) Computer-aided diagnosis for World Health Organization-defined chest radiograph primary-endpoint pneumonia in children. Pediatr Radiol 50:482–491

    Article  PubMed  Google Scholar 

  28. Chen Y, Roberts CS, Ou W et al (2021) Deep learning for classification of pediatric chest radiographs by WHO’s standardized methodology. PLoS one 16:e0253239

  29. Yang Z, Xu Q, Bao S, Cao X, Huang Q (2021) Learning with multiclass AUC: theory and algorithms. IEEE Trans Pattern Anal Mach Intell 44:7747–7763

    Article  Google Scholar 

  30. Van Calster B, Van Belle V, Condous G, Bourne T, Timmerman D, Van Huffel S (2008) Multi-class AUC metrics and weighted alternatives2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence). IEEE, pp 1390–1396

  31. Gimeno P, Mingote V, Ortega A, Miguel A, Lleida E (2021) Generalizing AUC optimization to multiclass classification for audio segmentation with limited training data. IEEE Signal Process Lett 28:1135–1139

    Article  Google Scholar 

  32. Nishino M, Ashiku SK, Kocher ON, Thurer RL, Boiselle PM, Hatabu H (2006) The thymus: a comprehensive review. Radiographics 26:335–348

    Article  PubMed  Google Scholar 

  33. Nikolić MZ, Sun D, Rawlins EL (2018) Human lung development: recent progress and new challenges. Development 145:dev163485

    Article  PubMed  PubMed Central  Google Scholar 

  34. Goldstein AJ, Oliva I, Honarpisheh H, Rubinowitz A (2015) A tour of the thymus: a review of thymic lesions with radiologic and pathologic correlation. Can Assoc Radiol J 66:5–15

    Article  PubMed  Google Scholar 

Download references

Funding

This study has received funding by the National Natural Science Foundation of China (grant number 61802330, 61802331) and Xiamen Science and Technology Plan Project (grant number 3502Z202009220).

Author information

Author notes

  1. Both Yuemei Li and Lin Zhang contributed equally. 

  2. Both corresponding authors Jungang Liu and Qiang Zheng contributed equally.

Authors and Affiliations

  1. School of Computer and Control Engineering, Yantai University, Yantai, 264005, China

    Yuemei Li, Hu Yu & Qiang Zheng

  2. Department of Radiology, Xiamen Children’s Hospital, Children’s Hospital of Fudan University at Xiamen, Xiamen, Fujian, China

    Lin Zhang, Jian Wang & Jungang Liu

  3. Yantai University Trier College of Sustainable Technology, Yantai, 264005, Shandong Province, China

    Shuo Wang

  4. Trier University of Applied Sciences, D-54208, Trier, Germany

    Shuo Wang

Authors
  1. Yuemei Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Lin Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Hu Yu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Jian Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Shuo Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Jungang Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Qiang Zheng
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding authors

Correspondence to Jungang Liu or Qiang Zheng.

Ethics declarations

Guarantor

The scientific guarantor of this publication is Qiang Zheng from Yantai University.

Conflict of interest

The authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.

Statistics and biometry

No complex statistical methods were necessary for this paper.

Informed consent

The scientific ethics committee of local hospital approved this retrospective study and waived the requirement for informed patient consent (Approval number: [2022] No.18).

Ethical approval

Institutional Review Board approval was obtained.

Study subjects or cohorts overlap

The study subjects were collected from the children aged 3–6 years between January 2016 and June 2021 at local children’s hospital.

Methodology

• retrospective

• diagnostic or prognostic study

• performed at one institution

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (PDF 1016 KB)

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Li, Y., Zhang, L., Yu, H. et al. A comprehensive segmentation of chest X-ray improves deep learning–based WHO radiologically confirmed pneumonia diagnosis in children. Eur Radiol 34, 3471–3482 (2024). https://doi.org/10.1007/s00330-023-10367-y

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00330-023-10367-y

Keywords

Search

Navigation