🩺 CNN for Medical X-Ray Classification

---

📌 Project Overview

This repository implements a Convolutional Neural Network (CNN) to classify chest X-ray images into NORMAL or PNEUMONIA.
The project follows a structured deep learning workflow, including dataset preparation, preprocessing, training, evaluation, and visualization of results.

---

🗂️ Dataset

We use the Chest X-Ray Images (Pneumonia) dataset from Kaggle.

📥 Dataset link: Chest X-Ray Dataset (Kaggle)

Folder Structure:

data/chest_xray/
├── train/
│   ├── NORMAL/
│   └── PNEUMONIA/
├── val/
│   ├── NORMAL/
│   └── PNEUMONIA/
└── test/
    ├── NORMAL/
    └── PNEUMONIA/

• Train: used for training the CNN
• Validation: used for tuning hyperparameters
• Test: used for final evaluation

---

⚙️ Preprocessing

Before training, the dataset undergoes image preprocessing and augmentation to improve generalization:

• ✅ Resizing images to 150×150 pixels
• ✅ Normalizing pixel values to range [0, 1]
• ✅ Data augmentation:
  • Random rotations
  • Width/height shifts
  • Zoom operations
  • Horizontal flips

This ensures the model is robust and reduces overfitting.

---

🔄 Workflow

The end-to-end workflow of the project:

1. Dataset Loading → Load train/val/test splits
2. Preprocessing & Augmentation → Resize, normalize, and augment images
3. Training → Early stopping and checkpointing for best model
4. Evaluation → Classification metrics & confusion matrix
5. Visualization → Training curves and prediction samples

---

🛠 Tech Stack

• Python 3.8+
• TensorFlow / Keras (Deep Learning Framework)
• NumPy, scikit-learn (Data Handling & Metrics)

---

📊 Results

• ✅ Model achieves high accuracy in distinguishing NORMAL vs PNEUMONIA X-rays
• ✅ Confusion matrix & classification report included
• ✅ Visualization of training curves and sample predictions

📈 Training Curves

• Accuracy and loss plots for training & validation

🖼 Sample Predictions

• Example predictions with true vs predicted labels

---

🌐 Live Demo

🚀 Try the model here: Live Demo Link

---

📂 Repository Structure

xray_classification/
├── data/                  # Chest X-ray dataset
├── models/                # Saved trained models
├── results/               # Training history, plots, evaluation
├── src/                   # Source code
│   ├── data_loader.py     # Data loading & augmentation
│   ├── xray_classifier.py # CNN model architecture
│   ├── evaluate.py        # Evaluation metrics
├── train.py               # Training script
├── run.py                 # Run evaluation & predictions
├── requirements.txt       # Dependencies
├── README.md              # Project documentation
└── LICENSE                # MIT License

---

📥 Installation & Usage

# Clone the repository
git clone https://github.com/Vedag812/Image_Classification
cd xray_classification

# Install dependencies
pip install -r requirements.txt

# Train the model
python train.py

# Run evaluation & predictions
python run.py

---

📜 License

This project is licensed under the MIT License.

---

🔬 Built with Deep Learning to advance healthcare AI solutions.