Skip to main content

Advertisement

SpringerLink
Log in

Prediction of visual field from swept-source optical coherence tomography using deep learning algorithms

  • Glaucoma
  • Published:
Graefe's Archive for Clinical and Experimental Ophthalmology Aims and scope Submit manuscript

Abstract

Purpose

To develop a deep learning method to predict visual field (VF) from wide-angle swept-source optical coherence tomography (SS-OCT) and compare the performance of three Google Inception architectures.

Methods

Three deep learning models (with Inception-ResNet-v2, Inception-v3, and Inception-v4) were trained to predict 24-2 VF from the macular ganglion cell-inner plexiform layer and the peripapillary retinal nerve fibre layer map obtained by SS-OCT. The prediction performance of the three models was evaluated by using the root mean square error (RMSE) between the actual and predicted VF. The performance was also compared among different glaucoma severities and Garway-Heath sectorizations.

Results

The training dataset comprised images of 2220 eyes from 1120 subjects, and the test dataset was obtained from another 305 subjects (305 eyes). In all subjects, the global prediction errors (RMSEs) were 4.44 ± 2.09 dB, 4.78 ± 2.38 dB, and 4.85 ± 2.66 dB for the Inception-ResNet-v2, Inception-v3, and Inception-v4 architectures, respectively, and the prediction error of Inception-ResNet-v2 was significantly lower than the other two (P < 0.001). As glaucoma progressed, the prediction error of all three architectures significantly worsened to 6.59 dB, 7.33 dB, and 7.79 dB, respectively. In the analysis of sectors, the nasal sector had the lowest prediction error, followed by the superotemporal sector.

Conclusions

Inception-ResNet-v2 achieved the best performance, and the global prediction error (RMSE) was 4.44 dB. As glaucoma progressed, the prediction error became larger. This method may help clinicians determine VF, particularly for patients who are unable to undergo a physical VF test.

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

Similar content being viewed by others

Data availability

All source codes, trained model file, and example images are available on Github website: http://github.com/climyth/VFbySS-OCT. However, original training image data cannot be made publicly available because they contain identifying patient information. Data are available upon request from the Pusan National University Yangsan Hospital.

References

  1. Quigley HA, Katz J, Derick RJ et al (1992) An evaluation of optic disc and nerve fiber layer examinations in monitoring progression of early glaucoma damage. Ophthalmology 99:19–28

    Article  CAS  Google Scholar 

  2. Artes PH, Chauhan BC (2005) Longitudinal changes in the visual field and optic disc in glaucoma. Prog Retin Eye Res 24:333–354

    Article  Google Scholar 

  3. Fogagnolo P, Sangermani C, Oddone F et al (2011) Long-term perimetric fluctuation in patients with different stages of glaucoma. Br J Ophthalmol 95:189–193

    Article  Google Scholar 

  4. Wild JM, Dengler-Harles M, Searle AET et al (1989) The influence of the learning effect on automated perimetry in patients with suspected glaucoma. Acta Ophthalmol 67:537–545. https://doi.org/10.1111/j.1755-3768.1989.tb04105.x

    Article  CAS  Google Scholar 

  5. Langerhorst CT, Van den Berg T, Van Spronsen R, Greve EL (1985) Results of a fluctuation analysis and defect volume program for automated static threshold perimetry with the scoperimeter. In: Sixth International Visual Field Symposium. Springer, pp 1–6

  6. Brenton RS, Argus WA (1987) Fluctuations on the Humphrey and Octopus perimeters. Invest Ophthalmol Vis Sci 28:767–771

    CAS  PubMed  Google Scholar 

  7. Gürses-Özden R, Teng C, Vessani R et al (2004) Macular and retinal nerve fiber layer thickness measurement reproducibility using optical coherence tomography (OCT-3). J Glaucoma 13:238

    Article  Google Scholar 

  8. Blumenthal EZ, Williams JM, Weinreb RN et al (2000) Reproducibility of nerve fiber layer thickness measurements by use of optical coherence tomography. Ophthalmology 107:2278–2282

    Article  CAS  Google Scholar 

  9. Carpineto P, Nubile M, Agnifili L et al (2012) Reproducibility and repeatability of CirrusTM HD-OCT peripapillary retinal nerve fibre layer thickness measurements in young normal subjects. OPH 227:139–145. https://doi.org/10.1159/000334967

    Article  Google Scholar 

  10. Yang Z, Tatham AJ, Zangwill LM et al (2015) Diagnostic ability of retinal nerve fiber layer imaging by swept-source optical coherence tomography in glaucoma. Am J Ophthalmol 159:193–201. https://doi.org/10.1016/j.ajo.2014.10.019

    Article  PubMed  Google Scholar 

  11. Wollstein G, Schuman JS, Price LL et al (2004) Optical coherence tomography (OCT) macular and peripapillary retinal nerve fiber layer measurements and automated visual fields. Am J Ophthalmol 138:218–225. https://doi.org/10.1016/j.ajo.2004.03.019

    Article  PubMed  Google Scholar 

  12. Sato S, Hirooka K, Baba T et al (2013) Correlation between the ganglion cell-inner plexiform layer thickness measured with cirrus HD-OCT and macular visual field sensitivity measured with microperimetry. Invest Ophthalmol Vis Sci 54:3046–3051. https://doi.org/10.1167/iovs.12-11173

    Article  PubMed  Google Scholar 

  13. Raza AS, Cho J, de Moraes CGV et al (2011) Retinal ganglion cell layer thickness and local visual field sensitivity in Glaucoma. Arch Ophthalmol 129:1529–1536. https://doi.org/10.1001/archophthalmol.2011.352

    Article  PubMed  PubMed Central  Google Scholar 

  14. Garway-Heath DF, Poinoosawmy D, Fitzke FW, Hitchings RA (2000) Mapping the visual field to the optic disc in normal tension glaucoma eyes. Ophthalmology 107:1809–1815

    Article  CAS  Google Scholar 

  15. Szegedy C, Vanhoucke V, Ioffe S, et al (2016) Rethinking the inception architecture for computer vision. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp 2818–2826

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

  17. Bianco S, Cadene R, Celona L, Napoletano P (2018) Benchmark analysis of representative deep neural network architectures. IEEE Access 6:64270–64277

    Article  Google Scholar 

  18. Foster PJ, Buhrmann R, Quigley HA, Johnson GJ (2002) The definition and classification of glaucoma in prevalence surveys. Br J Ophthalmol 86:238–242

    Article  Google Scholar 

  19. Park SC, Kung Y, Su D et al (2013) Parafoveal scotoma progression in glaucoma: Humphrey 10-2 versus 24-2 visual field analysis. Ophthalmology 120:1546–1550. https://doi.org/10.1016/j.ophtha.2013.01.045

    Article  PubMed  Google Scholar 

  20. Chollet F (2015) Keras official web site: https://keras.io. Accessed 20 March 2020

  21. Dubitzky W, Granzow M, Berrar DP (2007) Fundamentals of data mining in genomics and proteomics. Springer Science & Business Media, pp 177–180

  22. Jonas JB, Budde WM, Lang P (1998) Neuroretinal rim width ratios in morphological glaucoma diagnosis. Br J Ophthalmol 82:1366–1371

    Article  CAS  Google Scholar 

  23. Szegedy C, Ioffe S, Vanhoucke V, Alemi AA (2017) Inception-v4, Inception-ResNet and the impact of residual connections on learning. In: Thirty-First AAAI Conference on Artificial Intelligence. http://www.cs.cmu.edu/~jeanoh/16-785/papers/szegedy-aaai2017-inception-v4.pdf. Accessed 20 March 2020

  24. Srivastava RK, Greff K, Schmidhuber J (2015) Highway networks. arXiv preprint arXiv:150500387

  25. Li J, Fang F, Mei K, Zhang G (2018) Multi-scale residual network for image super-resolution. In: Proceedings of the European Conference on Computer Vision (ECCV). pp 517–532

  26. Zhang Y, Tian Y, Kong Y, et al (2018) Residual dense network for image super-resolution. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp 2472–2481

  27. Li K, Bare B, Yan B, et al (2018) HNSR: highway networks based deep convolutional neural networks model for single image super-resolution. In: 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). pp 1478–1482

  28. Lee WJ, Kim YK, Jeoung J, Park KH (2017) Can probability maps of swept-source optical coherence tomography predict visual field changes in preperimetric glaucoma? IOVS ARVO J 58:6257–6264. https://doi.org/10.1167/iovs.17-22697

    Article  Google Scholar 

  29. Hood DC, De Cuir N, Blumberg DM et al (2016) A single wide-field OCT protocol can provide compelling information for the diagnosis of early glaucoma. Transl Vis Sci Technol 5:4–4

    Article  Google Scholar 

  30. Guo Z, Kwon YH, Lee K et al (2017) Optical coherence tomography analysis based prediction of Humphrey 24-2 visual field thresholds in patients with glaucoma. Invest Ophthalmol Vis Sci 58:3975–3985. https://doi.org/10.1167/iovs.17-21832

    Article  PubMed  PubMed Central  Google Scholar 

  31. Zhu H, Crabb DP, Schlottmann PG et al (2010) Predicting visual function from the measurements of retinal nerve fiber layer structure. Invest Ophthalmol Vis Sci 51:5657–5666. https://doi.org/10.1167/iovs.10-5239

    Article  PubMed  Google Scholar 

  32. Kawano J, Tomidokoro A, Mayama C et al (2006) Correlation between hemifield visual field damage and corresponding parapapillary atrophy in normal-tension glaucoma. Am J Ophthalmol 142:40–45.e1

    Article  Google Scholar 

Download references

Acknowledgements

The authors appreciated to the professional English proofread company, “AJE”, to ensure that the language is clear and free of errors.

Funding

This study was supported by the Research Institute for the Convergence of Biomedical Science and Technology (30-2018-017), Pusan National University Yangsan Hospital.

Author information

Authors and Affiliations

  1. Department of Ophthalmology, Pusan National University College of Medicine, Busan, South Korea

    Keunheung Park

  2. Department of Ophthalmology, Pusan Medical Center, Busan, South Korea

    Keunheung Park

  3. Department of Biostatistics, Clinical Trial Center, Biomedical Research Institute, Pusan National University Hospital, Busan, South Korea

    Jinmi Kim

  4. Department of Ophthalmology, College of Medicine, Pusan National University Yangsan Hospital, Yangsan, South Korea

    Sangyoon Kim & Jonghoon Shin

  5. Department of Ophthalmology, Research Institute for Convergence of Biomedical Science and Technology, Pusan National University Yangsan Hospital, 20 Geumo-ro, Mulgeum-eup, Yangsan-si, Gyeongsangnam-do, South Korea

    Jonghoon Shin

Authors
  1. Keunheung Park
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Jinmi Kim
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Sangyoon Kim
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Jonghoon Shin
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

K Park, S Kim and J Shin participated in the design of the study and in the statistical analysis. J Kim performed the statistical analysis. K Park built artificial intelligence architecture, analysed data and wrote the paper. J Shin participated in the acquisition of dataand the drafting of the manuscript, and corrected the paper. All authors read and approved the final manuscript and agree to be accountable of all aspects of the work.

Corresponding author

Correspondence to Jonghoon Shin.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Ethics approval

This study was approved by Institutional Review Board of Pusan National University Yangsan Hospital, South Korea (05-2019-134).

Informed consent

Patient consent was waived by the Institutional Review Board due to the retrospective nature of the study.

Research involving human participants and/or animals

This study was approved by Institutional Review Board of Pusan National University Yangsan Hospital, South Korea (05-2019-134).

Additional information

Publisher’s note

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

Electronic supplementary material

Table S1

(XLSX 178 kb)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Park, K., Kim, J., Kim, S. et al. Prediction of visual field from swept-source optical coherence tomography using deep learning algorithms. Graefes Arch Clin Exp Ophthalmol 258, 2489–2499 (2020). https://doi.org/10.1007/s00417-020-04909-z

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00417-020-04909-z

Keywords

Search

Navigation