Diabetic Retinopathy Severity Level Detection Using Deep Learning
No Thumbnail Available
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
In recent years, Artificial Intelligence (AI) has revolutionized numerous fields of scientific research by
offering innovative solutions to complex challenges. The healthcare sector is no exception. Diabetes, a
prevalent condition both globally and in Algeria, often leads to a severe complication known as diabetic
retinopathy. This condition can blur or distort vision and is a major cause of blindness. Early detection
of diabetic retinopathy is crucial for effective treatment. Recent advancements in AI, especially in deep
learning, provide promising solutions for predicting, diagnosing, and managing various diseases in their
early stages.
In our master project, we develop a robust system for detecting the levels of diabetic retinopathy
using Convolutional Neural Networks (CNNs) enhanced with specific adaptations to improve performance.
Leveraging a pre-trained CNN model, we introduce tailored adjustments to parameters that optimize
the network for our dataset and the particular nuances of DR detection. Furthermore, we implement
multi-label classification to address the complex and overlapping features of diabetic retinopathy stages.
This approach allows our model to capture and represent the intricate patterns in retinal images more
effectively than traditional single-label methods. Our experimental results demonstrate that this combined
strategy significantly enhances the accuracy and reliability of Diabetic Retinopathy (DR) level detection.
This project emphasizes the Importance of Machine learning and Deep learning techniques in healthcare,
offering a promising tool for clinicians in the fight against diabetic eye diseases.
.