Advertisement

Utility of a public-available artificial intelligence in diagnosis of polypoidal choroidal vasculopathy

  • Jingyuan Yang
  • Chenxi Zhang
  • Erqian Wang
  • Youxin ChenEmail author
  • Weihong Yu
Retinal Disorders
  • 46 Downloads

Abstract

Purpose

To investigate the feasibility of training an artificial intelligence (AI) on a public-available AI platform to diagnose polypoidal choroidal vasculopathy (PCV) using indocyanine green angiography (ICGA).

Methods

Two methods using AI models were trained by a data set including 430 ICGA images of normal, neovascular age-related macular degeneration (nvAMD), and PCV eyes on a public-available AI platform. The one-step method distinguished normal, nvAMD, and PCV images simultaneously. The two-step method identifies normal and abnormal ICGA images at the first step and diagnoses PCV from the abnormal ICGA images at the second step. The method with higher performance was used to compare with retinal specialists and ophthalmologic residents on the performance of diagnosing PCV.

Results

The two-step method had better performance, in which the precision was 0.911 and the recall was 0.911 at the first step, and the precision was 0.783, and the recall was 0.783 at the second step. For the test data set, the two-step method distinguished normal and abnormal images with an accuracy of 1 and diagnosed PCV with an accuracy of 0.83, which was comparable to retinal specialists and superior to ophthalmologic residents.

Conclusion

In this evaluation of ICGA images from normal, nvAMD, and PCV eyes, the models trained on a public-available AI platform had comparable performance to retinal specialists for diagnosing PCV. The utility of public-available AI platform might help everyone including ophthalmologists who had no AI-related resources, especially those in less developed areas, for future studies.

Keywords

Artificial intelligence Deep learning Diagnosis Indocyanine green angiography Machine learning Polypoidal choroidal vasculopathy 

Notes

Acknowledgments

Thanks are due to Xiao Zhang, Huan Chen, Ruoan Han, Bilei Zhang, Yuelin Wang, and Shan Wu for making diagnosis on data set and supporting this study.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical approval

All applicable international, national, and/or institutional guidelines for the care and use of animals were followed. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. For this type of study formal consent is not required.

Supplementary material

417_2019_4493_Fig2_ESM.png (260 kb)
ESM 1

Supplementary Fig. Schematic diagram of the methods. ICGA images were uploaded to the cloud platform, and each image was given a label of “Normal”, “Typical nvAMD”, or “PCV” based on the truth diagnosis. Then the models used in the two methods were trained and self-evaluated by the platform using corresponding ICGA images. Finally, the method of better performance was compared with human on test images (PNG 259 kb)

417_2019_4493_MOESM1_ESM.tif (2.5 mb)
High resolution image (TIF 2574 kb)

References

  1. 1.
    Cheung CMG, Lai TYY, Ruamviboonsuk P, Chen SJ, Chen Y, Freund KB, Gomi F, Koh AH, Lee WK, Wong TY (2018) Polypoidal choroidal vasculopathy: definition, pathogenesis, diagnosis, and management. Ophthalmology 125:708–724.  https://doi.org/10.1016/j.ophtha.2017.11.019 CrossRefGoogle Scholar
  2. 2.
    Cheung CMG, Lee WK, Koizumi H, Dansingani K, Lai TYY, Freund KB (2019) Pachychoroid disease. Eye (London, England) 33:14–33.  https://doi.org/10.1038/s41433-018-0158-4 CrossRefGoogle Scholar
  3. 3.
    Koh AH, Chen LJ, Chen SJ, Chen Y, Giridhar A, Iida T, Kim H, Yuk Yau Lai T, Lee WK, Li X, Han Lim T, Ruamviboonsuk P, Sharma T, Tang S, Yuzawa M (2013) Polypoidal choroidal vasculopathy: evidence-based guidelines for clinical diagnosis and treatment. Retina 33:686–716.  https://doi.org/10.1097/IAE.0b013e3182852446 CrossRefGoogle Scholar
  4. 4.
    Ozkaya A, Alagoz C, Garip R, Alkin Z, Perente I, Yazici AT, Taskapili M (2016) The role of indocyanine green angiography imaging in further differential diagnosis of patients with nAMD who are morphologically poor responders to ranibizumab in a real-life setting. Eye (Lond) 30:958–965.  https://doi.org/10.1038/eye.2016.71 CrossRefGoogle Scholar
  5. 5.
    Broadhead GK, Hong T, Chang AA (2014) Treating the untreatable patient: current options for the management of treatment-resistant neovascular age-related macular degeneration. Acta Ophthalmol 92:713–723.  https://doi.org/10.1111/aos.12463 CrossRefPubMedGoogle Scholar
  6. 6.
    Tan CS, Ngo WK, Lim LW, Tan NW, Lim TH (2016) EVEREST study report 3: diagnostic challenges of polypoidal choroidal vasculopathy. Lessons learnt from screening failures in the EVEREST study. Graefes Arch Clin Exp ophthalmol 254:1923–1930.  https://doi.org/10.1007/s00417-016-3333-y CrossRefPubMedPubMedCentralGoogle Scholar
  7. 7.
    Schmidt-Erfurth U, Sadeghipour A, Gerendas BS, Waldstein SM, Bogunovic H (2018) Artificial intelligence in retina. Prog Retin Eye Res 67:1–29.  https://doi.org/10.1016/j.preteyeres.2018.07.004 CrossRefPubMedGoogle Scholar
  8. 8.
    Lu W, Tong Y, Yu Y, Xing Y, Chen C, Shen Y (2018) Applications of artificial intelligence in ophthalmology: general overview. J Ophthalmol 2018:5278196.  https://doi.org/10.1155/2018/5278196 CrossRefPubMedPubMedCentralGoogle Scholar
  9. 9.
    Yann L, Leon B, Yoshua B, Patrick H (1998) Gradient-based learning applied to document recognition. Proc IEEE 86:2278–2324CrossRefGoogle Scholar
  10. 10.
    Anwar SM, Majid M, Qayyum A, Awais M, Alnowami M, Khan MK (2018) Medical image analysis using convolutional neural networks: a review. J Med Syst 42:226.  https://doi.org/10.1007/s10916-018-1088-1 CrossRefPubMedGoogle Scholar
  11. 11.
    Zoph B, Le QV (2016) Neural architecture search with reinforcement learning. https://arxiv.org/abs/1611.01578
  12. 12.
    Zoph B, Vasudevan V, Shlens J, Le QV (2018) Learning transferable architectures for scalable image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 8697–8710Google Scholar
  13. 13.
    Gulshan V, Peng L, Coram M, Stumpe MC, Wu D, Narayanaswamy A, Venugopalan S, Widner K, Madams T, Cuadros J, Kim R, Raman R, Nelson PC, Mega JL, Webster DR (2016) Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. JAMA 316:2402–2410.  https://doi.org/10.1001/jama.2016.17216 CrossRefPubMedGoogle Scholar
  14. 14.
    Burlina PM, Joshi N, Pekala M, Pacheco KD, Freund DE, Bressler NM (2017) Automated grading of age-related macular degeneration from color fundus images using deep convolutional neural networks. JAMA ophthalmology 135:1170–1176.  https://doi.org/10.1001/jamaophthalmol.2017.3782 CrossRefPubMedPubMedCentralGoogle Scholar
  15. 15.
    Treder M, Lauermann JL, Eter N (2018) Automated detection of exudative age-related macular degeneration in spectral domain optical coherence tomography using deep learning. Graefes Arch Clin Exp Ophthalmol 256:259–265.  https://doi.org/10.1007/s00417-017-3850-3 CrossRefPubMedGoogle Scholar
  16. 16.
    Yuzawa M (2015) Two subtypes of polypoidal choroidal vasculopathy: idiopathic disease or age-related macular degeneration. Invest Ophthalmol Vis Sci 56:3998.  https://doi.org/10.1167/iovs.15-17207 CrossRefPubMedGoogle Scholar
  17. 17.
    Treder M, Lauermann JL, Eter N (2018) Deep learning-based detection and classification of geographic atrophy using a deep convolutional neural network classifier. Graefes Arch Clin Exp Ophthalmol 256:2053–2060.  https://doi.org/10.1007/s00417-018-4098-2 CrossRefPubMedGoogle Scholar
  18. 18.
    Saleh E, Blaszczynski J, Moreno A, Valls A, Romero-Aroca P, de la Riva-Fernandez S, Slowinski R (2018) Learning ensemble classifiers for diabetic retinopathy assessment. Artif Intell Med 85:50–63.  https://doi.org/10.1016/j.artmed.2017.09.006 CrossRefPubMedGoogle Scholar
  19. 19.
    Bogunovic H, Waldstein SM, Schlegl T, Langs G, Sadeghipour A, Liu X, Gerendas BS, Osborne A, Schmidt-Erfurth U (2017) Prediction of anti-VEGF treatment requirements in neovascular AMD using a machine learning approach. Invest Ophthalmol Vis Sci 58:3240–3248.  https://doi.org/10.1167/iovs.16-21053 CrossRefPubMedGoogle Scholar
  20. 20.
    Rohm M, Tresp V, Muller M, Kern C, Manakov I, Weiss M, Sim DA, Priglinger S, Keane PA, Kortuem K (2018) Predicting visual acuity by using machine learning in patients treated for neovascular age-related macular degeneration. Ophthalmology 125:1028–1036.  https://doi.org/10.1016/j.ophtha.2017.12.034 CrossRefPubMedGoogle Scholar
  21. 21.
    Guo J, Li B (2018) The application of medical artificial intelligence technology in rural areas of developing countries. Health Equity 2:174–181.  https://doi.org/10.1089/heq.2018.0037 CrossRefPubMedPubMedCentralGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2019

Authors and Affiliations

  1. 1.Department of Ophthalmology, Peking Union Medical College HospitalChinese Academy of Medical SciencesBeijingChina

Personalised recommendations