Article

An artificial intelligence platform for the multihospital collaborative management of congenital cataracts

Received:
Accepted:
Published online:

Abstract

Using artificial intelligence (AI) to prevent and treat diseases is an ultimate goal in computational medicine. Although AI has been developed for screening and assisted decision-making in disease prevention and management, it has not yet been validated for systematic application in the clinic. In the context of rare diseases, the main strategy has been to build specialized care centres; however, these centres are scattered and their coverage is insufficient, which leaves a large proportion of rare-disease patients with inadequate care. Here, we show that an AI agent using deep learning, and involving convolutional neural networks for diagnostics, risk stratification and treatment suggestions, accurately diagnoses and provides treatment decisions for congenital cataracts in an in silico test, in a website-based study, in a ‘finding a needle in a haystack’ test and in a multihospital clinical trial. We also show that the AI agent and individual ophthalmologists perform equally well. Moreover, we have integrated the AI agent with a cloud-based platform for multihospital collaboration, designed to improve disease management for the benefit of patients with rare diseases.

  • Subscribe to Nature Biomedical Engineering for full access:

    $99

    Subscribe
  • Purchase article full text and PDF:

    $32

    Buy now

Additional access options:

Already a subscriber? Log in now or Register for online access.

References

  1. 1.

    et al. The coming of age of artificial intelligence in medicine. Artif. Intel. Med. 46, 5–17 (2009).

  2. 2.

    et al. A smartphone dongle for diagnosis of infectious diseases at the point of care. Sci. Transl. Med. 7, 273re1 (2015).

  3. 3.

    , , , & Innovative tools for assessing risks for severe adverse events in areas of overlapping Loa loa and other filarial distributions: the application of micro-stratification mapping. Parasit. Vect. 7, 307 (2014).

  4. 4.

    et al. Clinical decision support systems for improving diagnostic accuracy and achieving precision medicine. J. Clin. Bioinf. 5, 4 (2015).

  5. 5.

    , , & Why rare diseases are an important medical and social issue. Lancet 371, 2039–2041 (2008).

  6. 6.

    , & National plans and strategies on rare diseases in Europe. Adv. Exp. Med. Biol. 686, 475–491 (2010).

  7. 7.

    & Rare diseases: what's next? Lancet 371, 1978–1979 (2008).

  8. 8.

    , , , & China’s first rare-disease registry is under development. Lancet 378, 769–770 (2011).

  9. 9.

    & Current progress in the management of rare diseases and orphan drugs in China. Intractable Rare Dis. Res. 1, 45–52 (2012).

  10. 10.

    & Pediatric cataract: challenges and future directions. Clin. Ophthal. 9, 77–90 (2015).

  11. 11.

    et al. Lens regeneration using endogenous stem cells with gain of visual function. Nature 531, 323–328 (2016).

  12. 12.

    , , & Documenting rare disease data in China. Science 349, 1064 (2015).

  13. 13.

    et al. Lanosterol reverses protein aggregation in cataracts. Nature 523, 607–611 (2015).

  14. 14.

    et al. Global challenges in the management of congenital cataract: proceedings of the 4th International Congenital Cataract Symposium. J. AAPOS 19, e1–e8 (2015).

  15. 15.

    in Proceedings of the 14th International Joint Conference on Artificial Intelligence 1137–1143 (2001).

  16. 16.

    et al. Exploiting ensemble learning for automatic cataract detection and grading. Comp. Meth. Progr. Biomed. 2016, 45–57 (2015).

  17. 17.

    , , , & A computer-aided healthcare system for cataract classification and grading based on fundus image analysis. Comp. Indust. 5, 72–80 (2015).

  18. 18.

    , , , & The morphology and natural history of childhood cataracts. Surv. Ophthalmol. 48, 125–144 (2003).

  19. 19.

    et al. Mastering the game of Go with deep neural networks and tree search. Nature 529, 484–489 (2016).

  20. 20.

    et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. JAMA 316, 2402–2410 (2016)

  21. 21.

    , & Large scale visual recognition challenge (accessed 16 April 2016).

  22. 22.

    , & ImageNet classification with deep convolutional neural networks. Adv. Neur. Inf. Proc. Syst. 25, 2012 (2012).

  23. 23.

    & Visualizing and understanding convolutional networks. Lecture Notes Comp. Sci. 8689, 818–833 (2013).

  24. 24.

    & Very deep convolutional networks for large-scale image recognition. Preprint at (2014).

  25. 25.

    , , & Gradient-based learning applied to document recognition. Proc. IEEE Inst. Electr. Electron. Eng. 86, 2278–2324 (1998).

  26. 26.

    , , & What is the best multi-stage architecture for object recognition. In Proc. IEEE 12th Int. Conf. Computer Vision 2146–2153 (2009).

  27. 27.

    , & Multi-column deep neural networks for image classification. In Proc. IEEE 25th Conf. Computer Vision and Pattern Recognition 3642–3649 (2012).

  28. 28.

    , & Convolutional networks and applications in vision. In IEEE Int. Symp. Circuits and Systems 253–256 (2010).

  29. 29.

    Large-scale machine learning with stochastic gradient descent. In Proc. 19th Int. Conf. Computational Satistics 177–186 (2010).

  30. 30.

    , , , & Improving neural networks by preventing co-adaptation of feature detectors. Computer Sci. 3, 212–223 (2012).

  31. 31.

    , , , & High-performance neural networks for visual object classification. Preprint at (2011).

  32. 32.

    , & Best practices for convolutional neural networks applied to visual document analysis. In Proc. 7th Int. Conf. Document Analysis and Recognition 958–963 (2003).

  33. 33.

    et al. Caffe: convolutional architecture for fast feature embedding. In Proc. 22nd ACM Int. Conf. Multimedia 675–678 (2014).

Download references

Author information

Author notes

    • Erping Long
    •  & Haotian Lin

    These authors contributed equally to this work.

Affiliations

  1. State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Centre, Sun Yat-sen University, Guangzhou 510060, China

    • Erping Long
    • , Haotian Lin
    • , Zhenzhen Liu
    • , Xiaohang Wu
    • , Zhuoling Lin
    • , Xiaoyan Li
    • , Jingjing Chen
    • , Jing Li
    • , Qianzhong Cao
    • , Dongni Wang
    • , Weirong Chen
    •  & Yizhi Liu
  2. School of Computer Science and Technology, Xidian University, Xi’an 710071, China

    • Liming Wang
    • , Jiewei Jiang
    • , Yingying An
    •  & Xiyang Liu
  3. School of Software, Xidian University, Xi’an 710071, China.

    • Xiyang Liu

Authors

  1. Search for Erping Long in:

  2. Search for Haotian Lin in:

  3. Search for Zhenzhen Liu in:

  4. Search for Xiaohang Wu in:

  5. Search for Liming Wang in:

  6. Search for Jiewei Jiang in:

  7. Search for Yingying An in:

  8. Search for Zhuoling Lin in:

  9. Search for Xiaoyan Li in:

  10. Search for Jingjing Chen in:

  11. Search for Jing Li in:

  12. Search for Qianzhong Cao in:

  13. Search for Dongni Wang in:

  14. Search for Xiyang Liu in:

  15. Search for Weirong Chen in:

  16. Search for Yizhi Liu in:

Contributions

H.L., E.L., X.Liu and Y.L. designed the research; E.L., H.L., J.J., Z.Liu, X.W., L.W., Z.Lin, X.Li, J.C., J.L., Q.C., D.W. and W.C. conducted the study; E.L., H.L. and J.J. collected the data; E.L., H.L., J.J. and Y.A. analysed the data; E.L. and H.L. co-wrote the manuscript; all authors discussed the results and commented on the manuscript.

Competing interests

The authors declare no competing financial interests.

Corresponding authors

Correspondence to Haotian Lin or Yizhi Liu.

Supplementary information

PDF files

  1. 1.

    Supplementary Information

    Supplementary algorithm information, figures, tables and test paper.

Videos

  1. 1.

    Supplementary Video 1

    Instructions for using the CC-Cruiser website.