AI Democratization in Optometry: Developing a Prototype with Azure Cognitive Services Platform


  • Shaun Baek Portola High School
  • Ryan Johnson Mentor, Portola High School
  • Claire Saunders Mentor, California Institute of Technology
  • Debora Lee Chen Mentor, School of Optometry, University of California Berkeley
  • Katherine Lai Mentor, School of Optometry, University of California Berkeley



Telemedicine, Teleoptometry, Artificial Intelligence, AI Democratization


As Artificial Intelligence (AI) technology advances, it is used in almost every aspect of our lives. However, AI is still complicated to implement without help from computer engineers. In the health care field, knowledge of medical and computer knowledge is necessary to create AI-based medical systems. Close cooperation between medical experts and computer experts is essential. For this reason, even if there has been a continuous effort to apply AI into the medical field, it has yet to be universalized. In particular, in the field of optometry and ophthalmology, more complex technology is required than in other medical fields because it is necessary to analyze an eye image to diagnose a disease. Therefore, this study explores the possibility for medical professionals with little computer knowledge in the field of ophthalmology to develop an AI-based diagnostic system without the help of computer engineers. In addition, it explores not only the possibilities but also the diagnostic accuracy of the developed system. Our results show that the diagnostic system discriminates against five common eye diseases to some extent. This study explores whether AI democratization is possible even in the field of ophthalmology that requires advanced skills and knowledge.


Download data is not yet available.

References or Bibliography

American Optometric Association. (n.d.). See the Full Picture of Your Health with an Annual Comprehensive Eye Exam. Retrieved July 22, 2021, from

Bivens, J., & Zipperer, B.. (2020, August 26). Health insurance and the COVID-19 shock: What we know so far about health insurance losses and what it means for policy. Economic Policy Institute.

Boyd, K. (2021a, June 28). What Is Diabetic Retinopathy? American Academy of Ophthalmology.

Boyd, K. (2021b, June 15). What Is Glaucoma? American Academy of Ophthalmology.

Boyd, K. (2021c, July 7). What Are Cataracts? American Academy of Ophthalmology.

Bureau, U. C. (2020, September 15). Health Insurance Coverage in the United States: 2019. Retrieved from

Chen, Y. (n.d.). Retina dataset containing 1) normal 2) cataract 3) glaucoma 4) retina disease. GitHub. Retrieved July 23, 2021,

Copeland, B. (2018, November 21). MYCIN. Encyclopedia Britannica.

Garvey, C. (2018, April). A framework for evaluating barriers to the democratization of artificial intelligence. In Thirty-Second AAAI Conference on Artificial Intelligence.

Guo, K., Ren, S., Bhuiyan, M. Z. A., Li, T., Liu, D., Liang, Z., & Chen, X. (2020). MDMaaS: Medical-Assisted Diagnosis Model as a Service With Artificial Intelligence and Trust. IEEE Transactions on Industrial Informatics, 16(3), 2102–2114.

Hendrick, B. (2011, May 19). CDC: Many Americans Are Skipping Eye Care. WebMD.

Hulsey, A. (2020). A look into the impacts of tele-optometry expansionSharma. Honors Theses.

Joint Shantou International Eye Centre. (2019, June 18). 1000 Fundus images with 39 categories. Kaggle.

Microsoft Azure. (n.d.). Cognitive Services. Retrieved July 23, 2021, from

Microsoft Azure. (2020, March 6). Select a domain for a Custom Vision project.

Mohajon, J. (2021, June 18). Confusion Matrix for Your Multi-Class Machine Learning Model. Medium.

Mrutyunjaya, & Raga, S. (2020, October 14). A Smartphone Based Application for Early Detection of Diabetic Retinopathy Using Normal Eye Extraction. IEEE Xplore.

Noronha, K., & Nayak, K. P. (2012, February 1). Fundus image analysis for the detection of diabetic eye diseases-a review. IEEE Xplore.

Optix Family Eyecare. (2020, April 16). What Services Can I Get Using Tele-Optometry?

Patry, G. G. G. (2021, February 18). Messidor-2. ADCIS.

Rotterdam Ophthalmic Data Repository. (n.d.). Datasets. Retrieved July 23, 2021, from

Primera Eye Care. (2020, November 1). 9 Things That Cause Bad Eyesight | Primera Eye Care.

Shanggong Medical Technology Co., Ltd. (2020, September 24). Ocular Disease Recognition. Kaggle.

Sharma, M., Jain, N., Ranganathan, S., Sharma, N., Honavar, S. G., Sharma, N., & Sachdev, M. S. (2020). Tele-ophthalmology: Need of the hour. Indian journal of ophthalmology, 68(7), 1328–1338.

Shortliffe E. H. (1977). Mycin: A Knowledge-Based Computer Program Applied to Infectious Diseases. Proceedings of the Annual Symposium on Computer Application in Medical Care, 66–69.

Taylor, H. R., Vu, H. T., McCarty, C. A., & Keeffe, J. E. (2004, August 1). The Need for Routine Eye Examinations. Investigative Ophthalmology & Visual Science.

Turbert, D. (2021, April 9). Nearsightedness: What Is Myopia? American Academy of Ophthalmology.



How to Cite

Baek, S., Johnson, R., Saunders, C., Lee Chen, D., & Lai, K. (2021). AI Democratization in Optometry: Developing a Prototype with Azure Cognitive Services Platform. Journal of Student Research, 10(3).



HS Research Articles