Advancements in Diabetic Retinopathy and Cataract Identification Through Deep Learning

The increasing global prevalence of diabetic retinopathy and cataracts necessitates improved detection methods, as traditional manual examinations are often slow and error-prone, risking delayed treatment and potential vision loss. This paper presents the development of a sophisticated deep learning...

Full description

Saved in:
Bibliographic Details
Published in2024 International Conference on Data Science and Network Security (ICDSNS) pp. 1 - 5
Main Authors Raman, Ramakrishnan, Kumar, Vikram, Saini, Dimple, Rabadiya, Dhaval, Patre, Smruti, Meenakshi, R.
Format Conference Proceeding
LanguageEnglish
Published IEEE 26.07.2024
Subjects
Accuracy
Attention mechanisms
automated detection
Blindness
Cataracts
Convolutional neural networks
convolutional neural networks (cnns)
Deep learning
Diabetic retinopathy
Refining
Residual neural networks
Retina
Online AccessGet full text

Cover

More Information
Summary:The increasing global prevalence of diabetic retinopathy and cataracts necessitates improved detection methods, as traditional manual examinations are often slow and error-prone, risking delayed treatment and potential vision loss. This paper presents the development of a sophisticated deep learning model using convolutional neural networks (CNNs), which enhances the accuracy, efficiency, and accessibility of diagnosing these ocular conditions from retinal images. Our novel CNN architecture incorporates Inception modules, Residual Networks, and attention mechanisms, specifically tailored to the intricacies of automated diabetic retinopathy and cataract detection. Rigorously evaluated on a diverse dataset, the model demonstrates significant advancements over existing methods, achieving 94.5% precision, 93.2% recall, 93.8% F1 score, and 95.1% overall accuracy. These results not only highlight the model's capability in providing rapid and precise diagnosis but also suggest its potential application in widespread screening programs, especially in areas lacking specialized ophthalmological services. The findings underscore the transformative impact of deep learning in ophthalmology, offering a promising tool for early detection and management of eye diseases, thereby preventing possible blindness. Moving forward, further refinement and clinical validation of this model will be essential to realize its full potential in enhancing global ocular health.
DOI:10.1109/ICDSNS62112.2024.10691108