Lung Disease Classification

A deep learning project for classifying lung diseases from chest X-ray images using transfer learning and attention mechanisms.

Features

• Multi-class classification of lung diseases:
  • Bacterial Pneumonia
  • COVID-19
  • Tuberculosis
  • Viral Pneumonia
  • Normal
• Support for multiple pre-trained architectures:
  • DenseNet201
  • ResNet152V2
  • VGG16
• Transfer learning with EfficientNetV2
• Attention mechanism for better feature extraction
• Advanced data augmentation and preprocessing
• Model interpretation using Grad-CAM
• Comprehensive evaluation metrics and visualizations
• Experiment tracking and logging
• Automated testing and CI/CD

Installation

1. Clone the repository:

git clone https://github.com/yourusername/lung-disease-classification.git
cd lung-disease-classification

2. Create and activate a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install the package:

pip install -e ".[dev]"  # For development
# or
pip install -e .  # For production

Dataset

Follow the instructions in data/README.md to download and set up the dataset.

Usage

Training

python scripts/train.py \
    --config configs/training_config.yaml \
    --data-dir data/processed \
    --output-dir output

Prediction

python scripts/predict.py \
    --model-path output/best_model.h5 \
    --image-path path/to/image.jpg

Evaluation

python scripts/evaluate.py \
    --model-path output/best_model.h5 \
    --test-dir data/processed/test

Training a Model (Python API)

from src.data import DataLoader
from src.model import LungDiseaseModel
from src.evaluate import ModelEvaluator

# Load and preprocess data
data_loader = DataLoader("data/")
train_gen, val_gen = data_loader.load_and_preprocess()

# Create and train model
model = LungDiseaseModel(model_name='densenet')
history = model.train(train_gen, val_gen, epochs=50)

# Evaluate model
evaluator = ModelEvaluator(model.model, class_names=data_loader.classes)
evaluator.plot_training_history(history)
evaluator.compute_confusion_matrix(val_gen)
print(evaluator.generate_classification_report(val_gen))

Generating Grad-CAM Visualizations

evaluator.generate_gradcam("path/to/image.jpg")

Development

1. Install development dependencies:

pip install -e ".[dev]"

2. Run tests:

pytest tests/

3. Run linting:

flake8 src tests
black src tests

Project Structure

lung-disease-classification/
├── configs/                     # Configuration files
│   ├── model_config.yaml       # Model architecture configuration
│   └── training_config.yaml    # Training hyperparameters
├── data/                       # Dataset directory
│   ├── raw/                   # Original dataset
│   ├── processed/             # Preprocessed data
│   └── README.md             # Dataset documentation
├── notebooks/                  # Jupyter notebooks for analysis
│   ├── exploratory_analysis.ipynb
│   └── model_experiments.ipynb
├── src/                       # Source code
│   ├── data/                 # Data handling modules
│   │   ├── __init__.py
│   │   ├── dataset.py       # Dataset class
│   │   └── preprocessing.py # Data preprocessing
│   ├── models/              # Model architecture modules
│   │   ├── __init__.py
│   │   ├── base.py         # Base model class
│   │   └── attention.py    # Attention mechanism
│   ├── training/           # Training modules
│   │   ├── __init__.py
│   │   ├── trainer.py     # Training loop
│   │   └── callbacks.py   # Custom callbacks
│   ├── evaluation/        # Evaluation modules
│   │   ├── __init__.py
│   │   ├── metrics.py    # Custom metrics
│   │   └── visualization.py # Plotting utilities
│   └── utils/            # Utility functions
│       ├── __init__.py
│       └── helpers.py    # Helper functions
├── tests/                # Unit tests
│   ├── test_data.py
│   ├── test_models.py
│   └── test_training.py
├── scripts/             # Command-line scripts
│   ├── train.py        # Training script
│   ├── predict.py      # Prediction script
│   └── evaluate.py     # Evaluation script
├── docs/               # Documentation
│   ├── api/           # API documentation
│   └── examples/      # Usage examples
├── .github/           # GitHub configuration
│   └── workflows/     # CI/CD workflows
├── requirements/      # Project dependencies
│   ├── base.txt      # Base requirements
│   ├── dev.txt       # Development requirements
│   └── test.txt      # Testing requirements
├── setup.py          # Package setup
├── pyproject.toml    # Project metadata
├── .pre-commit-config.yaml  # Pre-commit hooks
├── .gitignore
├── LICENSE
└── README.md

Contributing

1. Fork the repository
2. Create a feature branch: git checkout -b feature-name
3. Commit changes: git commit -am 'Add feature'
4. Push to the branch: git push origin feature-name
5. Submit a pull request

License

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

Acknowledgments

• Dataset: Chest X-Ray Images (Pneumonia)
• Model architecture: EfficientNetV2
• Pre-trained models: Keras Applications