Bone Fracture Classification System

This project implements a deep learning-based bone fracture classification system using various architectures including ResNet, U-Net, Attention U-Net, DenseNet, Vision Transformers, and Hybrid ViT.

Project Structure

.
├── models/             # Model architectures
│   └── resnet.py      # ResNet implementation
├── data/              # Dataset directory
├── utils/             # Utility functions
├── train.py           # Training script
├── requirements.txt   # Project dependencies
└── README.md          # Project documentation

Setup

1. Create a virtual environment (recommended):

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

2. Install dependencies:

pip install -r requirements.txt

Current Implementation

Currently implemented:

• ResNet50-based fracture classifier with transfer learning
• Basic training pipeline with validation
• Support for binary classification (fracture/no fracture)

Usage

To train the model:

python train.py

Next Steps

1. Implement remaining architectures:

   • U-Net
   • Attention U-Net
   • DenseNet
   • Vision Transformers
   • Hybrid ViT

2. Add data loading and preprocessing for MURA dataset

3. Implement data augmentation

4. Add evaluation metrics

5. Implement synthetic image generation with CycleGAN and StyleGAN