Automated screening for retinopathy
https://doi.org/10.18008/1816-5095-2013-1-4-7
Abstract
Retinal pathology is a common cause of an irreversible decrease of central vision commonly found amongst senior population. Detection of the earliest signs of retinal diseases can be facilitated by viewing retinal images available from the telemedicine networks. To facilitate the process of retinal images, screening software applications based on image recognition technology are currently on the various stages of development.
Purpose: To develop and implement computerized image recognition software that can be used as a decision support technology
for retinal image screening for various types of retinopathies.
Methods: The software application for the retina image recognition has been developed using C++ language. It was tested on dataset of 70 images with various types of pathological features (age related macular degeneration, chorioretinitis, central serous chorioretinopathy and diabetic retinopathy).
Results: It was shown that the system can achieve a sensitivity of 73 % and specificity of 72 %.
Conclusion: Automated detection of macular lesions using proposed software can significantly reduce manual grading workflow. In addition, automated detection of retinal lesions can be implemented as a clinical decision support system for telemedicine screening. It is anticipated that further development of this technology can become a part of diagnostic image analysis system for the electronic health records.
About the Authors
A. S. RodinRussian Federation
V. S. Akopyan
Russian Federation
N. S. Semenova
Russian Federation
A. S. Krylov
Russian Federation
References
1. Whited J. D., Datta S. K., Aiello L. M. A modeled economic analysis of a digital tele-ophthalmology system as used by three federal health care agencies for detecting proliferative diabetic retinopathy // Telemed J E Health. 2005. V. 1.P. 641‑651.
2. Dua S., Acharya R. Computational analysis of the human eye with applications 2011, World Scientific, Singapore.
3. Антонов Е. А., Акопян В. С., Родин А. С. и др. Методы автоматизированной обработки изображений сетчатки в телемедицинском скрининге диабетической ретинопатии. Обзор литературы // Офтальмология. 2011. Т. 8, № 3. С.4‑8.
4. Chernomorets A. A., Krylov A. S., Nasonov A. V., et al. Automated processing of retinal images // In: 21‑th International Conf. Computer Graphics GraphiCon’2011. Moscow, Russia, 2011. P. 78−81.
5. National Institute of Diabetes and Digestive and Kidney Diseases. National Diabetes Statistics. http://diabetes.niddk.nih.gov / dm / pubs / statistics / index. htm#12. (Дата обращения 11.07.2012).
6. Cavallerano A. A., Cavallerano J. D., Katalinic P. Use of Joslin Vision Network digitalvideo nonmydriatic retinal imaging to assess diabetic retinopathy in a clinical programme // Retina. 2003. V. 23. P. 215‑223.
7. Lin D. Y., Blumenkranz M. S., Brothers R. J. The sensitivity and specificity of singlefield nonmydriatic monochromatic digital fundus photography with remote interpretation for diabetic retinopathy screening: a comparison with ophthalmoscopy and standardized mydriatic color photography // Am J Ophthalmol. 2002. V. 134. P. 204‑213.
8. Zimmer-Galler I. E., Zeimer R. Feasibility of screening for high-risk age-related macular degeneration with an Internet-based automated fundus camera // Ophthalmic Surg Lasers Imaging. 2005. V. 36. P. 228‑236.
9. Mason J. National screening for diabetic retinopathy: Clear vision needed // Diabet. Med. 2003. V. 20. P. 959‑961.
Review
For citations:
Rodin A.S., Akopyan V.S., Semenova N.S., Krylov A.S. Automated screening for retinopathy. Ophthalmology in Russia. 2013;10(1):4-7. https://doi.org/10.18008/1816-5095-2013-1-4-7
Email this article 