SBC-Based Diabetic Retinopathy and Diabetic Macular Edema Classification System using Deep Convolutional Neural Network
Anicia Coleen S. Reyes1, Gem Ryan C. Milan2, James Marvin M. Quilaton3, Bryant Exel G. Sigue4, Steven Valentino E. Arellano5, Kenneth C. Karamihan6
1Anicia Coleen S. Reyes*, Computer Engineering Department, College of Engineering, University of Perpetual Help System Laguna, City of Biñan, Laguna, 4024 Philippines.
2Gem Ryan C. Milan, Computer Engineering Department, College of Engineering, University of Perpetual Help System Laguna, City of Biñan, Laguna, 4024 Philippines.
3James Marvin M. Quilaton, Computer Engineering Department, College of Engineering, University of Perpetual Help System Laguna, City of Biñan, Laguna, 4024 Philippines.
4Bryant Exel G. Sigue, Computer Engineering Department, College of Engineering, University of Perpetual Help System Laguna, City of Biñan, Laguna, 4024 Philippines.
5Steven Valentino E. Arellano, Computer Engineering Department, College of Engineering, and Graduate School, University of Perpetual Help System Laguna, City of Biñan, Laguna, Philippines.
6Kenneth C. Karamihan, Computer Engineering Department, College of Engineering, University of Perpetual Help System Laguna, City of Biñan, Laguna, Philippines.
Manuscript received on August 01, 2020. | Revised Manuscript received on August 05, 2020. | Manuscript published on September 30, 2020. | PP: 9-16 | Volume-9 Issue-3, September 2020. | Retrieval Number: 100.1/ijrte.C4195099320 | DOI: 10.35940/ijrte.C4195.099320
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: This Raspberry Pi Single-Board Computer-Based Diabetic Retinopathy (DR) and Diabetic Macular Edema (DME) Classification System using Deep Convolutional Neural Network through Inception v3 Transfer Learning and MATLAB digital image processing paradigm based on International Clinical DR and DME Disease Severity Scale with Python application, which would capture the image of the retina of diabetic patients to classify the grade, severity, and types of DR; and the grade of DME without using dilating drops. It would also display, save, search and print the partial diagnosis that can be done to the patients. Diabetic patients, endocrinologists and ophthalmologists of one of the medical centers in City of San Pedro, Laguna, Philippines tested the system. Obtained results indicated that the classification of DR and DME, and its characteristics using the system were accurate and reliable, which could be an assistive device for endocrinologists and ophthalmologists.
Keywords: Diabetic retinopathy, diabetic macular edema, deep convolutional neural network, digital image processing, transfer learning.