Skip to main content
SpringerLink
Log in

Automated quality assessment of chest radiographs based on deep learning and linear regression cascade algorithms

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

Abstract

Objectives

Develop and evaluate the performance of deep learning and linear regression cascade algorithms for automated assessment of the image layout and position of chest radiographs.

Methods

This retrospective study used 10 quantitative indices to capture subjective perceptions of radiologists regarding image layout and position of chest radiographs, including the chest edges, field of view (FOV), clavicles, rotation, scapulae, and symmetry. An automated assessment system was developed using a training dataset consisting of 1025 adult posterior-anterior chest radiographs. The evaluation steps included: (i) use of a CNN framework based on ResNet - 34 to obtain measurement parameters for quantitative indices and (ii) analysis of quantitative indices using a multiple linear regression model to obtain predicted scores for the layout and position of chest radiograph. In the testing dataset (n = 100), the performance of the automated system was evaluated using the intraclass correlation coefficient (ICC), Pearson correlation coefficient (r), mean absolute difference (MAD), and mean absolute percentage error (MAPE).

Results

The stepwise regression showed a statistically significant relationship between the 10 quantitative indices and subjective scores (p < 0.05). The deep learning model showed high accuracy in predicting the quantitative indices (ICC = 0.82 to 0.99, r = 0.69 to 0.99, MAD = 0.01 to 2.75). The automatic system provided assessments similar to the mean opinion scores of radiologists regarding image layout (MAPE = 3.05%) and position (MAPE = 5.72%).

Conclusions

Ten quantitative indices correlated well with the subjective perceptions of radiologists regarding the image layout and position of chest radiographs. The automated system provided high performance in measuring quantitative indices and assessing image quality.

Key Points

• Objective and reliable assessment for image quality of chest radiographs is important for improving image quality and diagnostic accuracy.

• Deep learning can be used for automated measurements of quantitative indices from chest radiographs.

• Linear regression can be used for interpretation-based quality assessment of chest radiographs.

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

CNN:

Convolutional neural network

DL:

Deep learning

ICC:

Intraclass correlation coefficient

MAD:

Mean absolute difference

MAPE:

Mean absolute percentage error

MOS:

Mean opinion score

PCK:

Percentage of correct key points

References

  1. Mettler FA Jr, Mahesh M, Bhargavan-Chatfield M et al (2020) Patient exposure from radiologic and nuclear medicine procedures in the United States: Procedure volume and effective dose for the period 2006-2016. Radiology 295:418–427

    Article  Google Scholar 

  2. Tesselaar E, Dahlström N, Sandborg M (2016) Clinical audit of image quality in radiology using visual grading characteristics analysis. Radiat Prot Dosimetry 169:340–346

    Article  Google Scholar 

  3. Whaley JS, Pressman BD, Wilson JR, Bravo L, Sehnert WJ, Foos DH (2013) Investigation of the variability in the assessment of digital chest X-ray image quality. J Digit Imaging 26:217–226

    Article  Google Scholar 

  4. Andersen ER, Jorde J, Taoussi N, Yaqoob SH, Konst B, Seierstad T (2012) Reject analysis in direct digital radiography. Acta Radiol 53:174–178

    Article  Google Scholar 

  5. Miyata T, Yanagawa M, Hata A, Honda O, Tomiyama N (2020) Influence of field of view size on image quality: ultra-high-resolution CT vs. conventional high-resolution CT. European Radiology 30:3324–3333

    Article  Google Scholar 

  6. Hardy M, Scotland B, Herron L (2015) Assessing sagittal rotation on posteroanterior chest radiographs: the effect of body morphology on radiographic appearances. J Med Imaging Radiat Sci 46:365–371

    Article  Google Scholar 

  7. Vañó E, Guibelalde E, Morillo A, Alvarez-Pedrosa CS, Fernández JM (1995) Evaluation of the European image quality criteria for chest examinations. Br J Radiol 68:1349–1355

    Article  Google Scholar 

  8. Cui S, Ming S, Lin Y et al (2020) Development and clinical application of deep learning model for lung nodules screening on CT images. Sci Rep 10:13657

    Article  CAS  Google Scholar 

  9. Ye Q, Shen Q, Yang W et al (2020) Development of automatic measurement for patellar height based on deep learning and knee radiographs. Eur Radiol 30:4974–4984

    Article  Google Scholar 

  10. Xue J, Wang B, Ming Y et al (2020) Deep learning-based detection and segmentation-assisted management of brain metastases. Neuro Oncol 22:505–514

    Article  Google Scholar 

  11. Higaki T, Nakamura Y, Tatsugami F, Nakaura T, Awai K (2019) Improvement of image quality at CT and MRI using deep learning. Jpn J Radiol 37:73–80

    Article  Google Scholar 

  12. Satyananda Kashyap MM, Karargyris A, Wu JT, Morris M, Babak Saboury ES, Syeda-Mahmood T (2019) Artificial intelligence for point of care radiograph quality assessment. Conference: SPIE Medical Imaging 2019 At: San Diego

  13. Nousiainen K, Mäkelä T, Piilonen A, Peltonen JI (2021) Automating chest radiograph imaging quality control. Phys Med 83:138–145

    Article  Google Scholar 

  14. Poggenborg J, Yaroshenko A, Wieberneit N, Harder T, Gossmann A (2021) Impact of AI-based real time image quality feedback for chest radiographs in the clinical routine. medRxiv. https://doi.org/10.1101/2021.06.10.21258326 2021.2006.2010.21258326

  15. Commission of the European Communities (1996) European guidelines on quality criteria for diagnostic radiographic images, EUR 16260, Brussels

  16. Ganten M, Radeleff B, Kampschulte A, Daniels MD, Kauffmann GW, Hansmann J (2003) Comparing image quality of flat-panel chest radiography with storage phosphor radiography and film-screen radiography. AJR Am J Roentgenol 181:171–176

    Article  Google Scholar 

  17. Ronneberger O, Fischer P, Brox T (2015) U-Net: convolutional networks for biomedical image segmentation. Springer International Publishing, Cham, pp 234–241

    Google Scholar 

  18. He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, pp 770–778

  19. Nair V, Hinton GE (2010) Rectified linear units improve restricted boltzmann machines vinod nairproceedings of the 27th International Conference on Machine Learning (ICML-10), June 21-24, 2010, Haifa, Israel

  20. Ioffe S, Szegedy C (2015) Batch normalization: accelerating deep network training by reducing internal covariate shift. http://arxiv.org/abs/1502.03167

  21. Russakovsky O, Deng J, Su H et al (2014) ImageNet large scale visual recognition challenge. In: arXiv e-prints. Available via http://arxiv.org/abs/1409.0575. Accessed 13 Jul 2021

  22. Abadi M, Barham P, Chen J et al (2016) Tensorflow: a system for large-scale machine learning12th {USENIX} symposium on operating systems design and implementation ({OSDI} 16), pp 265-283

  23. Kingma DP, Ba J (2014) Adam: a method for stochastic optimization. In: arXiv e-prints. Available via https://arxiv.org/abs/1412.6980. Accessed 13 Jul 2021

  24. Chen Y, Wang Z, Peng Y, Zhang Z, Yu G, Sun J (2017) Cascaded pyramid network for multi-person pose estimation. In: arXiv e-prints. Available via https://arxiv.org/abs/1711.07319. Accessed 13 Jul 2021

  25. Payer C, Štern D, Bischof H, Urschler M (2019) Integrating spatial configuration into heatmap regression based CNNs for landmark localization. Med Image Anal 54:207–219

    Article  Google Scholar 

  26. Koo TK, Li MY (2016) A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med 15:155–163

    Article  Google Scholar 

  27. Amdisen A (1987) Pearson’s correlation coefficient, p-value, and lithium therapy. Biol Psychiatry 22:926–928

    Article  CAS  Google Scholar 

  28. Bland JM, Altman DG (1986) Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1:307–310

    Article  CAS  Google Scholar 

  29. Larson DB, Chen MC, Lungren MP, Halabi SS, Stence NV, Langlotz CP (2018) Performance of a deep-learning neural network model in assessing skeletal maturity on pediatric hand radiographs. Radiology 287:313–322

    Article  Google Scholar 

  30. Krupinski EA (2010) Current perspectives in medical image perception. Atten Percept Psychophys 72:1205–1217

    Article  Google Scholar 

  31. Båth M, Sund P, Månsson LG (2002) Evaluation of the imaging properties of two generations of a CCD-based system for digital chest radiography. Med Phys 29:2286–2297

    Article  Google Scholar 

  32. Bier B, Goldmann F, Zaech JN et al (2019) Learning to detect anatomical landmarks of the pelvis in X-rays from arbitrary views. Int J Comput Assist Radiol Surg 14:1463–1473

    Article  Google Scholar 

  33. Berg JV, Krnke S, Gooen A, Bystrov DB, Young S (2020) Robust chest x-ray quality assessment using convolutional neural networks and atlas regularization. Proc. SPIE 11313, Medical Imaging 2020: Image Processing, 113131L (10 March 2020). https://doi.org/10.1117/12.2549541

  34. Hou W, Gao X, Tao D, Li X (2015) Blind image quality assessment via deep learning. IEEE Trans Neural Netw Learn Syst 26:1275–1286

    Article  Google Scholar 

  35. Stępień I, Obuchowicz R, Piórkowski A, Oszust M (2021) Fusion of deep convolutional neural networks for no-reference magnetic resonance image quality assessment. Sensors (Basel) 21:1043

    Article  Google Scholar 

Download references

Acknowledgements

We would like to thank all the involved professional image quality assessment practitioners (radiologists and technicians) for dedicating their time and skill to the completion of this study.

Funding

This study was funded by Key Research and Development Project of Zhejiang Province of China (2020C01058).

Author information

Authors and Affiliations

  1. The Second Clinical Medical College, Zhejiang Chinese Medical University, Hangzhou, 310053, China

    Yu Meng

  2. Rehabilitation Medcine Center, Department of Radiology, Zhejiang Provincial People’s Hospital, Affiliated People’s Hospital, Hangzhou Medical College, Hangzhou, 310014, Zhejiang, China

    Yu Meng & Xiangyang Gong

  3. Bengbu Medical College, Bengbu, 233000, China

    Jingru Ruan

  4. School of Computer Science and Information Engineering, Zhejiang Gongshang University, Hangzhou, 310018, China

    Bailin Yang & Jianqiu Jin

  5. Hangzhou Medical College, Hangzhou, 310053, China

    Yang Gao

  6. School of Statistics and Mathematics, Zhejiang Gongshang University, Hangzhou, 310018, China

    Fangfang Dong

  7. Hangzhou Jianpei Technology Company Ltd, Hangzhou, 311200, China

    Hongli Ji, Linyang He & Guohua Cheng

  8. Institute of Artificial Intelligence and Remote Imaging, Hangzhou Medical College, Hangzhou, 310014, China

    Xiangyang Gong

Authors
  1. Yu Meng
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Jingru Ruan
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Bailin Yang
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Yang Gao
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Jianqiu Jin
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Fangfang Dong
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Hongli Ji
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Linyang He
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Guohua Cheng
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Xiangyang Gong
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Xiangyang Gong.

Ethics declarations

Guarantor

The scientific guarantor of this publication is Dr. Xiangyang Gong.

Conflict of interest

Hongli Ji, Linyang He, and Guohua Cheng are employees of Hangzhou Jianpei Technology Company Ltd. The remaining authors 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

Written informed consent was waived by the Institutional Review Board.

Ethical approval

Institutional Review Board approval was obtained.

Methodology

• retrospective

• experimental study

• multicenter study

Additional information

Publisher’s note

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

Supplementary information

ESM 1

(DOC 13549 kb)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Meng, Y., Ruan, J., Yang, B. et al. Automated quality assessment of chest radiographs based on deep learning and linear regression cascade algorithms. Eur Radiol 32, 7680–7690 (2022). https://doi.org/10.1007/s00330-022-08771-x

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00330-022-08771-x

Keywords

Search

Navigation