Diabetic Retinopathy Detection

A deep learning project for detecting and classifying diabetic retinopathy using PyTorch. The project implements both multiclass (5-class) and binary classification approaches using various CNN architectures.

Project Overview

This project aims to detect diabetic retinopathy from retinal images using different CNN architectures:

• Custom CNN
• AlexNet
• VGG

The implementation includes both multiclass classification (No_DR, Mild, Moderate, Severe, Proliferate_DR) and binary classification (No_DR vs DR) approaches.

Requirements

• Python 3.x
• PyTorch
• torchvision
• pandas
• numpy
• scikit-learn
• PIL (Python Imaging Library)
• matplotlib
• seaborn
• tqdm

Dataset Structure

The dataset should be organized as follows:

root_path/
│
├── train.csv
└── gaussian_filtered_images/
    └── gaussian_filtered_images/
        ├── No_DR/
        ├── Mild/
        ├── Moderate/
        ├── Severe/
        └── Proliferate_DR/

Features

Multiclass Classification

• Implementation of three CNN architectures:
  • Custom CNN
  • AlexNet
  • VGG
• Data augmentation techniques
• Training with metrics tracking
• Confusion matrix visualization
• Performance evaluation using accuracy, precision, recall, and F1-score

Binary Classification

• Custom CNN implementation
• Hyperparameter tuning
• Model evaluation
• Prediction visualization
• Confusion matrix analysis

Model Architectures

Custom CNN

- Input Layer (224x224x3)
- Conv2D + ReLU + MaxPool2D + BatchNorm
- Conv2D + ReLU + MaxPool2D + BatchNorm
- Conv2D + ReLU + MaxPool2D + BatchNorm
- Fully Connected Layers
- Output Layer (5 classes for multiclass, 2 for binary)

AlexNet

• Modified AlexNet architecture with:
  • 5 convolutional layers
  • 3 fully connected layers
  • Dropout for regularization
  • BatchNorm for better training stability

VGG

• Modified VGG architecture with:
  • 13 convolutional layers
  • 3 fully connected layers
  • Extensive use of 3x3 convolutions
  • MaxPooling layers

Training

Multiclass Training

python train_multiclass.py --model [custom|alexnet|vgg] --epochs 10 --batch_size 64

Binary Training

python train_binary.py --epochs 10 --batch_size 64

Model Performance

The models are evaluated using:

• Accuracy
• Precision
• Recall
• F1-score
• Confusion Matrix

Hyperparameter Tuning

The binary classification model includes hyperparameter tuning for:

• Learning rates: [1e-4, 1e-3]
• Batch sizes: [32, 64]
• Number of epochs: [10, 15]
• Dropout rates: [0.1, 0.2]

Data Preprocessing

• Image resizing to 224x224
• Normalization
• Data augmentation (for multiclass):
  • Random horizontal flip
  • Random rotation
  • Color jitter

Model Saving and Loading

The final trained model is saved as 'diabetic_retinopathy_model.pth'.

To load the model:

model = CustomCNN(dropout_rate)
model.load_state_dict(torch.load('diabetic_retinopathy_model.pth'))

Visualization

The project includes visualization tools for:

• Training and validation loss curves
• Confusion matrices
• Sample predictions on test images

Contributing

Feel free to submit issues and enhancement requests!

License

This project is licensed under the MIT License - see the LICENSE file for details.