Skip to main content

Optimal matrix size of chest radiographs for computer-aided detection on lung nodule or mass with deep learning

Abstract

Objectives

To investigate the optimal input matrix size for deep learning-based computer-aided detection (CAD) of nodules and masses on chest radiographs.

Methods

We retrospectively collected 2088 abnormal (nodule/mass) and 352 normal chest radiographs from two institutions. Three thoracic radiologists drew 2758 abnormalities regions. A total of 1736 abnormal chest radiographs were used for training and tuning convolutional neural networks (CNNs). The remaining 352 abnormal and 352 normal chest radiographs were used as a test set. Two CNNs (Mask R-CNN and RetinaNet) were selected to validate the effects of the squared different matrix size of chest radiograph (256, 448, 896, 1344, and 1792). For comparison, figure of merit (FOM) of jackknife free-response receiver operating curve and sensitivity were obtained.

Results

In Mask R-CNN, matrix size 896 and 1344 achieved significantly higher FOM (0.869 and 0.856, respectively) for detecting abnormalities than 256, 448, and 1792 (0.667–0.820) (p < 0.05). In RetinaNet, matrix size 896 was significantly higher FOM (0.906) than others (0.329–0.832) (p < 0.05). For sensitivity of abnormalities, there was a tendency to increase sensitivity when lesion size increases. For small nodules (< 10 mm), the sensitivities were 0.418 and 0.409, whereas the sensitivities were 0.937 and 0.956 for masses. Matrix size 896 and 1344 in Mask R-CNN and matrix size 896 in RetinaNet showed significantly higher sensitivity than others (p < 0.05).

Conclusions

Matrix size 896 had the highest performance for various sizes of abnormalities using different CNNs. The optimal matrix size of chest radiograph could improve CAD performance without additional training data.

Key Points

• Input matrix size significantly affected the performance of a deep learning-based CAD for detection of nodules or masses on chest radiographs.

• The matrix size 896 showed the best performance in two different CNN detection models.

• The optimal matrix size of chest radiographs could enhance CAD performance without additional training data.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Abbreviations

AUC:

Area under the curve

CAD:

Computer-aided detection

CNN:

Convolutional neural network

FOM:

Figure of merit

FPN:

Feature pyramid network

JAFROC:

Jackknife alternative free-response receiver operating characteristic curve

R-CNN:

Region-based convolutional neural network

ROC:

Receiver operating characteristic

ROI:

Region of interest

References

  1. 1.

    Schalekamp S, van Ginneken B, Koedam E et al (2014) Computer-aided detection improves detection of pulmonary nodules in chest radiographs beyond the support by bone-suppressed images. Radiology 272:252–261

    Article  Google Scholar 

  2. 2.

    van Ginneken B, Hogeweg L, Prokop M (2009) Computer-aided diagnosis in chest radiography: beyond nodules. Eur J Radiol 72:226–230

    Article  Google Scholar 

  3. 3.

    Nam JG, Park S, Hwang EJ et al (2018) Development and validation of deep learning–based automatic detection algorithm for malignant pulmonary nodules on chest radiographs. Radiology 290:218–228

    Article  Google Scholar 

  4. 4.

    Zakirov A, Kuleev R, Timoshenko A, Vladimirov A (2015) Advanced approaches to computer-aided detection of thoracic diseases on chest X-rays. Appl Math Sci 9:4361–4369

    Google Scholar 

  5. 5.

    Horváth G, Orbán G, Horváth Á et al (2009) A cad system for screening x-ray chest radiography. World Congress on Medical Physics and Biomedical Engineering, September 7-12, 2009, Munich, Germany. Springer, pp 210–213

  6. 6.

    Lodwick GS, Keats TE, Dorst JP (1963) The coding of roentgen images for computer analysis as applied to lung cancer. Radiology 81:185–200

    CAS  Article  Google Scholar 

  7. 7.

    Leibstein JM, Nel AL (2006) Detecting tuberculosis in chest radiographs using image processing techniques. University of Johannesburg

  8. 8.

    Carrillo-de-Gea JM, García-Mateos G (2010) Detection of normality/pathology on chest radiographs using LBP. Int Conf BioInformatics, January 20-23, 2010, Valencia, Spain, pp 167-172

  9. 9.

    Chan Y-H, Zeng Y-Z, Wu H-C, Wu M-C, Sun H-M (2018) Effective pneumothorax detection for chest X-ray images using local binary pattern and support vector machine. J Healthc Eng 2018:1-12

  10. 10.

    Dunnmon JA, Yi D, Langlotz CP, Ré C, Rubin DL, Lungren MP (2018) Assessment of convolutional neural networks for automated classification of chest radiographs. Radiology 290(2):537-544

  11. 11.

    Chaya Devi S, Satya Savithri T (2018) Review: On segmentation of nodules from posterior and anterior chest radiographs. Int J Biomed Imaging 2018:1-11

  12. 12.

    Pesce E, Withey SJ, Ypsilantis P-P, Bakewell R, Goh V, Montana G (2019) Learning to detect chest radiographs containing pulmonary lesions using visual attention networks. Med Image Anal 53:26–38

    Article  Google Scholar 

  13. 13.

    Annarumma M, Withey SJ, Bakewell RJ, Pesce E, Goh V, Montana G (2019) Automated triaging of adult chest radiographs with deep artificial neural networks. Radiology 291:196-202

  14. 14.

    Geras KJ, Wolfson S, Shen Y, Kim S, Moy L, Cho K (2017) High-resolution breast cancer screening with multi-view deep convolutional neural networks. arXiv preprint arXiv:170307047

  15. 15.

    Wang J, Yang X, Cai H, Tan W, Jin C, Li L (2016) Discrimination of breast cancer with microcalcifications on mammography by deep learning. Sci Rep 6:27327

    CAS  Article  Google Scholar 

  16. 16.

    Park S, Lee SM, Lee KH et al (2020) Deep learning-based detection system for multiclass lesions on chest radiographs: comparison with observer readings. Eur Radiol 30:1359–1368

    Article  Google Scholar 

  17. 17.

    Irvin J, Rajpurkar P, Ko M et al (2019) Chexpert: a large chest radiograph dataset with uncertainty labels and expert comparison. arXiv preprint arXiv:190107031

  18. 18.

    He K, Gkioxari G, Dollár P, Girshick R (2017) Mask R-CNN. IEEE Int Conf. Computer Vision, October 22-29, 2017, Venice, Italy. pp 2961-2969

  19. 19.

    Lin T-Y, Goyal P, Girshick R, He K, Dollár P (2017) Focal loss for dense object detection. IEEE Trans Pattern Anal Mach Intell 42(2):318-327

  20. 20.

    Szegedy C, Vanhoucke V, Ioffe S, Shlens J, Wojna Z (2016) Rethinking the inception architecture for computer vision. IEEE Int Conf. Computer Vision and Pattern Recognition, June 26-1, 2016, Las Vegas, NV, USA. pp 2818-2826

  21. 21.

    Szegedy C, Liu W, Jia Y et al (2015) Going deeper with convolutions. IEEE Int Conf Computer Vision and Pattern Recognition, June 7-12, 2015, Boston, MA, USA. pp 1-9

Download references

Funding

This study has received funding by the Industrial Strategic Technology Development program (10072064, Development of Novel Artificial Intelligence Technologies To Assist Imaging Diagnosis of Pulmonary, Hepatic, and Cardiac Diseases and Their Integration into Commercial Clinical PACS Platforms) funded by the Ministry of Trade Industry and Energy (MI, Korea).

Author information

Affiliations

Authors

Corresponding authors

Correspondence to Sang Min Lee or Namkug Kim.

Ethics declarations

Guarantor

The scientific guarantor of this publication is Sang Min Lee.

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

Jung Bok Lee, PhD kindly provided statistical advice for this manuscript.

Informed consent

Written informed consent was waived by the Institutional Review Board.

Ethical approval

Institutional Review Board approval was obtained.

Study subjects or cohorts overlap

Study subjects have been previously reported in Park et al study (Deep learning-based detection system for multiclass lesions on chest radiographs: comparison with observer readings. Eur Radiol. 2020 Mar;30(3):1359–1368).

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.

Electronic supplementary material

ESM 1

(DOCX 20 kb)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Kim, YG., Lee, S.M., Lee, K.H. et al. Optimal matrix size of chest radiographs for computer-aided detection on lung nodule or mass with deep learning. Eur Radiol 30, 4943–4951 (2020). https://doi.org/10.1007/s00330-020-06892-9

Download citation

Keywords

  • Thoracic radiography
  • Lung
  • Computer-assisted diagnosis
  • Deep learning
  • Algorithms