Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Deep-learning-based prediction of late age-related macular degeneration progression

Abstract

Both genetic and environmental factors influence the etiology of age-related macular degeneration (AMD), a leading cause of blindness. AMD severity is primarily measured by images of the fundus of the retina and recently developed machine learning methods can successfully predict AMD progression using image data. However, none of these methods have used both genetic and image data for predicting AMD progression. Here we used both genotypes and fundus images to predict whether an eye had progressed to late AMD with a modified deep convolutional neural network. In total, we used 31,262 fundus images and 52 AMD-associated genetic variants from 1,351 subjects from the Age-Related Eye Disease Study, which provided disease severity phenotypes and fundus images available at baseline and follow-up visits over a period of 12 years. Our results showed that fundus images coupled with genotypes could predict late AMD progression with an averaged area-under-the-curve value of 0.85 (95% confidence interval 0.83–0.86). The results using fundus images alone showed an averaged area under the receiver operating characteristic curve value of 0.81 (95% confidence interval 0.80–0.83). We implemented our model in a cloud-based application for individual risk assessment.

A preprint version of the article is available at ArXiv.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Fig. 1: Receiver operator characteristic curves of the prediction of late AMD progression time exceeding the inquired years for four models.
Fig. 2: Saliency maps for left eye of subject 1 over 5.9 years.
Fig. 3: Saliency maps for left eye of subject 2 over the first 5.8 years.
Fig. 4: Saliency maps for left eye of subject 3 over 12 years.
Fig. 5

Data availability

All the phenotype data and fundus images of AREDS participants required are available from dbGap (accession phs000001.v3.p1). The genotype data on AREDS subjects have been reported earlier9 and are available from dbGap (accession phs001039.v1.p1). The UK Biobank test dataset was obtained from UK Biobank (application number 43252).

Code availability

The prediction models with Python implementation and a detailed tutorial are available at https://github.com/QiYanPitt/AMDprogressCNN and a web-based graphical user interface is also available at http://www.pitt.edu/~qiy17/amdprediction.html.

References

  1. 1.

    Stark, K. et al. The German AugUR study: study protocol of a prospective study to investigate chronic diseases in the elderly. BMC Geriatr. 15, 130 (2015).

    Article  Google Scholar 

  2. 2.

    Congdon, N. et al. Causes and prevalence of visual impairment among adults in the United States. Arch. Ophthalmol. 122, 477–485 (2004).

    Article  Google Scholar 

  3. 3.

    Swaroop, A., Chew, E. Y., Rickman, C. B. & Abecasis, G. R. Unraveling a multifactorial late-onset disease: from genetic susceptibility to disease mechanisms for age-related macular degeneration. Annu. Rev. Genom. Hum. Genet. 10, 19–43 (2009).

    Article  Google Scholar 

  4. 4.

    Fritsche, L. G. et al. Age-related macular degeneration: genetics and biology coming together. Annu. Rev. Genom. Hum. Genet. 15, 151–171 (2014).

    Article  Google Scholar 

  5. 5.

    Ratnapriya, R. & Chew, E. Y. Age-related macular degeneration-clinical review and genetics update. Clin. Genet. 84, 160–166 (2013).

    Article  Google Scholar 

  6. 6.

    Grassmann, F., Fauser, S. & Weber, B. H. The genetics of age-related macular degeneration (AMD)–novel targets for designing treatment options? Eur. J. Pharm. Biopharm. 95, 194–202 (2015).

    Article  Google Scholar 

  7. 7.

    Holz, F. G. et al. Progression of geographic atrophy and impact of fundus autofluorescence patterns in age-related macular degeneration. Am. J. Ophthalmol. 143, 463–472 (2007).

    Article  Google Scholar 

  8. 8.

    Fritsche, L. G. et al. Seven new loci associated with age-related macular degeneration. Nat. Genet. 45, 439 (2013).

    Google Scholar 

  9. 9.

    Fritsche, L. G. et al. A large genome-wide association study of age-related macular degeneration highlights contributions of rare and common variants. Nat. Genet. 48, 134–143 (2016).

    Article  Google Scholar 

  10. 10.

    Yan, Q. et al. Genome-wide analysis of disease progression in age-related macular degeneration. Hum. Mol. Genet. 27, 929–940 (2018).

    Article  Google Scholar 

  11. 11.

    Chen, W. et al. Genetic variants near TIMP3 and high-density lipoprotein-associated loci influence susceptibility to age-related macular degeneration. Proc. Natl Acad. Sci. USA 107, 7401–7406 (2010).

    Article  Google Scholar 

  12. 12.

    The Age-Related Eye Disease Study Research Group. The Age-Related Eye Disease Study (AREDS): design implications. AREDS Report No. 1. Control Clin. Trials 20, 573–600 (1999).

    Article  Google Scholar 

  13. 13.

    Abramoff, M. D. et al. Improved automated detection of diabetic retinopathy on a publicly available dataset through integration of deep learning. Invest. Ophthalmol. Vis. Sci. 57, 5200–5206 (2016).

    Article  Google Scholar 

  14. 14.

    Gulshan, V. et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. J. Am. Med. Assoc. 316, 2402–2410 (2016).

    Article  Google Scholar 

  15. 15.

    Kermany, D. S. et al. Identifying medical diagnoses and treatable diseases by image-based deep learning. Cell 172, 1122–1131 (2018).

    Article  Google Scholar 

  16. 16.

    Devalla, S. K. et al. A deep learning approach to digitally stain optical coherence tomography images of the optic nerve head. Invest. Ophthalmol. Vis. Sci. 59, 63–74 (2018).

    Article  Google Scholar 

  17. 17.

    Grassmann, F. et al. A deep learning algorithm for prediction of age-related eye disease study severity scale for age-related macular degeneration from color fundus photography. Ophthalmology 125, 1410–1420 (2018).

    Article  Google Scholar 

  18. 18.

    Burlina, P. M. et al. Automated grading of age-related macular degeneration from color fundus images using deep convolutional neural networks. JAMA Ophthalmol. 135, 1170–1176 (2017).

    Article  Google Scholar 

  19. 19.

    Poplin, R. et al. Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning. Nat. Biomed. Eng. 2, 158–164 (2018).

    Article  Google Scholar 

  20. 20.

    Peng, Y. et al. DeepSeeNet: a deep learning model for automated classification of patient-based age-related macular degeneration severity from color fundus photographs. Ophthalmology 126, 565–575 (2019).

    Article  Google Scholar 

  21. 21.

    Abramoff, M. D., Garvin, M. K. & Sonka, M. Retinal imaging and image analysis. IEEE Rev. Biomed. Eng. 3, 169–208 (2010).

    Article  Google Scholar 

  22. 22.

    Yousefi, S. et al. Learning from data: recognizing glaucomatous defect patterns and detecting progression from visual field measurements. IEEE Trans. Biomed. Eng. 61, 2112–2124 (2014).

    Article  Google Scholar 

  23. 23.

    Chaudhuri, S., Chatterjee, S., Katz, N., Nelson, M. & Goldbaum, M. Detection of blood vessels in retinal images using two-dimensional matched filters. IEEE Trans. Med. Imaging 8, 263–269 (1989).

    Article  Google Scholar 

  24. 24.

    Burlina, P. M. et al. Use of deep learning for detailed severity characterization and estimation of 5-year risk among patients with age-related macular degeneration. JAMA Ophthalmol. 136, 1359–1366 (2018).

    Article  Google Scholar 

  25. 25.

    Youden, W. J. Index for rating diagnostic tests. Cancer 3, 32–35 (1950).

    Article  Google Scholar 

  26. 26.

    Sudlow, C. et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 12, e1001779 (2015).

    Article  Google Scholar 

  27. 27.

    Schmidt-Erfurth, U. et al. Prediction of individual disease conversion in early AMD using artificial intelligence. Invest. Ophthalmol. Vis. Sci. 59, 3199–3208 (2018).

    Article  Google Scholar 

  28. 28.

    Hao, J., Kim, Y., Mallavarapu, T., Oh, J. H. & Kang, M. Cox-PASNet: pathway-based sparse deep neural network for survival analysis. In 2018 IEEE Int. Conf. on Bioinformatics and Biomedicine (BIBM) 381–386 (IEEE, 2018).

  29. 29.

    The Age-Related Eye Disease Study Research Group. The Age-Related Eye Disease Study system for classifying age-related macular degeneration from stereoscopic color fundus photographs. AREDS Report No. 6. Am. J. Ophthalmol. 132, 668–681 (2001).

    Article  Google Scholar 

  30. 30.

    Davis, M. D. et al. The Age-Related Eye Disease Study severity scale for age-related macular degeneration. AREDS Report No. 17. Arch. Ophthalmol. 123, 1484–1498 (2005).

    Article  Google Scholar 

  31. 31.

    Swaroop, A., Branham, K. E., Chen, W. & Abecasis, G. Genetic susceptibility to age-related macular degeneration: a paradigm for dissecting complex disease traits. Hum. Mol. Genet. 16, R174–R182 (2007).

    Article  Google Scholar 

  32. 32.

    Ferris, F. L. 3rd et al. Clinical classification of age-related macular degeneration. Ophthalmology 120, 844–851 (2013).

    Article  Google Scholar 

  33. 33.

    Angermueller, C., Parnamaa, T., Parts, L. & Stegle, O. Deep learning for computational biology. Mol. Syst. Biol. 12, 878 (2016).

    Article  Google Scholar 

  34. 34.

    LeCun, Y., Bengio, Y. & Hinton, G. Deep learning. Nature 521, 436–444 (2015).

    Article  Google Scholar 

  35. 35.

    Szegedy, C., Vanhoucke V., Ioffe S., Shlens J. & Wojna Z. Rethinking the inception architecture for computer vision. In Proc. IEEE Conf. on Computer Vision and Pattern Recognition (IEEE, 2016).

  36. 36.

    Christopher, M. et al. Performance of deep learning architectures and transfer learning for detecting glaucomatous optic neuropathy in fundus photographs. Sci. Rep. 8, 16685 (2018).

    Article  Google Scholar 

  37. 37.

    Russakovsky, O. et al. ImageNet large scale visual recognition challenge. Int. J. Comp. Sci. 115, 211–252 (2015).

    MathSciNet  Google Scholar 

  38. 38.

    Ding, Y. et al. Bivariate analysis of age-related macular degeneration progression using genetic risk scores. Genetics 206, 119–133 (2017).

    Article  Google Scholar 

  39. 39.

    Chollet, F. Deep learning for humans: keras. GitHub https://github.com/keras-team/keras (2015).

  40. 40.

    Graf, E., Schmoor, C., Sauerbrei, W. & Schumacher, M. Assessment and comparison of prognostic classification schemes for survival data. Stat. Med. 18, 2529–2545 (1999).

    Article  Google Scholar 

  41. 41.

    Kotikalapudi, R. et al. Neural network visualization toolkit for keras. GitHub https://github.com/raghakot/keras-vis (2017).

  42. 42.

    Ribeiro, M. T., Singh, S. & Guestrin, C. “Why should I trust you?”: explaining the predictions of any classifier. In Proc. 22nd ACM SIGKDD Int. Conf. on Knowledge Discovery and Data Mining 1135–1144 (ACM, 2016).

  43. 43.

    DeLong, E. R., DeLong, D. M. & Clarke-Pearson, D. L. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 44, 837–845 (1988).

    Article  Google Scholar 

Download references

Acknowledgements

H.H. was partially supported by NSF IIS 1852606 and IIS 1837956. A.S. was supported by NEI Intramural Research Program grant number ZIAEY000546. The independent validation has been conducted using the UK Biobank Resource under Application Number 43252.

Author information

Affiliations

Authors

Contributions

Q.Y., Y.D. and W.C. conceived the project and supervised it. Q.Y. performed data processing and analysis. Q.Y. and H.X. designed the study. Q.Y., A.S. and E.Y.C. interpreted the data. Q.Y. and D.E.W. wrote the paper and Y.D., W.C., H.H., A.S. and E.Y.C. commented on the manuscript.

Corresponding authors

Correspondence to Qi Yan or Ying Ding or Wei Chen.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

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

Supplementary information

Supplementary Information

Supplementary Figs. 1–11 and Tables 1–3

Reporting Summary

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Yan, Q., Weeks, D.E., Xin, H. et al. Deep-learning-based prediction of late age-related macular degeneration progression. Nat Mach Intell 2, 141–150 (2020). https://doi.org/10.1038/s42256-020-0154-9

Download citation

Further reading

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing