Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Enhancing Diabetic Retinopathy Diagnosis: A Lightweight CNN Architecture for Efficient Exudate Detection in Retinal Fundus Images

Mujadded Al Rabbani Alif

TL;DR

The paper tackles early diabetic retinopathy detection by exudate markers using a lightweight CNN suitable for hardware-constrained environments. It introduces a two-block architecture with domain-specific data augmentations and regularization, achieving 4.73M parameters and an F1 score of 90%, outperforming the ResNet-18 baseline under limited data. Key contributions include empirical analysis of augmentation and regularization strategies, demonstrating that batch normalization notably enhances performance while excessive regularization can hurt generalization in simple networks. The work supports practical deployment in clinics and CPU-based settings and points to future enhancements via attention mechanisms and saliency mapping to improve interpretability and accuracy.

Abstract

Retinal fundus imaging plays an essential role in diagnosing various stages of diabetic retinopathy, where exudates are critical markers of early disease onset. Prompt detection of these exudates is pivotal for enabling optometrists to arrest or significantly decelerate the disease progression. This paper introduces a novel, lightweight convolutional neural network architecture tailored for automated exudate detection, designed to identify these markers efficiently and accurately. To address the challenge of limited training data, we have incorporated domain-specific data augmentations to enhance the model's generalizability. Furthermore, we applied a suite of regularization techniques within our custom architecture to boost diagnostic accuracy while optimizing computational efficiency. Remarkably, this streamlined model contains only 4.73 million parameters a reduction of nearly 60% compared to the standard ResNet-18 model, which has 11.69 million parameters. Despite its reduced complexity, our model achieves an impressive F1 score of 90%, demonstrating its efficacy in the early detection of diabetic retinopathy through fundus imaging.

Enhancing Diabetic Retinopathy Diagnosis: A Lightweight CNN Architecture for Efficient Exudate Detection in Retinal Fundus Images

TL;DR

The paper tackles early diabetic retinopathy detection by exudate markers using a lightweight CNN suitable for hardware-constrained environments. It introduces a two-block architecture with domain-specific data augmentations and regularization, achieving 4.73M parameters and an F1 score of 90%, outperforming the ResNet-18 baseline under limited data. Key contributions include empirical analysis of augmentation and regularization strategies, demonstrating that batch normalization notably enhances performance while excessive regularization can hurt generalization in simple networks. The work supports practical deployment in clinics and CPU-based settings and points to future enhancements via attention mechanisms and saliency mapping to improve interpretability and accuracy.

Abstract

Retinal fundus imaging plays an essential role in diagnosing various stages of diabetic retinopathy, where exudates are critical markers of early disease onset. Prompt detection of these exudates is pivotal for enabling optometrists to arrest or significantly decelerate the disease progression. This paper introduces a novel, lightweight convolutional neural network architecture tailored for automated exudate detection, designed to identify these markers efficiently and accurately. To address the challenge of limited training data, we have incorporated domain-specific data augmentations to enhance the model's generalizability. Furthermore, we applied a suite of regularization techniques within our custom architecture to boost diagnostic accuracy while optimizing computational efficiency. Remarkably, this streamlined model contains only 4.73 million parameters a reduction of nearly 60% compared to the standard ResNet-18 model, which has 11.69 million parameters. Despite its reduced complexity, our model achieves an impressive F1 score of 90%, demonstrating its efficacy in the early detection of diabetic retinopathy through fundus imaging.
Paper Structure (17 sections, 5 equations, 19 figures, 14 tables)

This paper contains 17 sections, 5 equations, 19 figures, 14 tables.

Table of Contents

  1. Introduction
  2. Literature Review
  3. Methodology
  4. Dataset
  5. Data Augmentations
  6. Oriented Based Scaling
  7. Contrast Variability
  8. Proposed Architecture
  9. Model Evaluation
  10. Hyperparameter Tuning
  11. Original Dataset Performance
  12. Augmented Dataset Performance
  13. Batch Normalization
  14. Dropout
  15. Batch Normalization and Dropout
  16. ...and 2 more sections

Figures (19)

  • Figure 1: Normal and Exudate Class Illustrations
  • Figure 2: Horizontal Augmentations: (A) Pre-augmentation, (B) Post-augmentation
  • Figure 3: Rotation Augmentation at 45°: (A) Pre-augmentation, (B, C) Post-augmentation
  • Figure 4: Rotation Augmentation at 20°: (A) Pre-augmentation, (B, C) Post-augmentation
  • Figure 5: Pixel Brightness Augmentation at 38°: (A) Pre-augmentation, (B, C) Post-augmentation
  • ...and 14 more figures