🩺 Pneumonia Detection using Deep Learning Models 🧠📊

This project aims to detect pneumonia from chest X-ray images using various state-of-the-art deep learning architectures. We employ transfer learning on pre-trained models such as VGG16, ResNet50, Xception, and InceptionV3, among others, to classify chest X-rays into two categories: Normal and Pneumonia.

📁 Dataset

We use the publicly available Chest X-Ray dataset from Kaggle. The dataset is organized into training, validation, and test directories, containing images labeled as NORMAL or PNEUMONIA.

Dataset structure:

• train/: Contains training images.
• val/: Contains validation images.
• test/: Contains test images.

🏗️ Architectures Implemented

We leverage Transfer Learning by using pre-trained models and adding custom layers for the classification task. The following architectures have been implemented:

1. VGG16 🧑‍💻
2. ResNet50 🧑‍🔬
3. InceptionV3 🧙‍♂️
4. MobileNetV2 🦸‍♀️

Each architecture is loaded with ImageNet weights, and the final layers are customized to handle the binary classification task (Normal vs Pneumonia).

🧪 Model Training

• Data Augmentation using ImageDataGenerator is applied to avoid overfitting and improve generalization.
• Early Stopping is used to prevent overtraining the model.
• Adam Optimizer with a learning rate of 0.0001 is used to train the models.

📊 Evaluation Metrics

For each architecture, we evaluate the model using the following metrics:

• Accuracy 🎯
• Confusion Matrix 📊
• Precision, Recall, and F1-Score 📏
• AUC-ROC Curve 🟠

Example: ROC Curve (Xception)

fpr, tpr, _ = roc_curve(y_true, y_pred_probs)
roc_auc = auc(fpr, tpr)

plt.plot(fpr, tpr, color='darkorange', lw=2, label=f'ROC curve (area = {roc_auc:.2f})')

Example: Confusion Matrix (Xception)

conf_matrix = confusion_matrix(y_true, y_pred_classes)
sns.heatmap(conf_matrix, annot=True, fmt='d', cmap='Blues')

📈 Training History

For each architecture, the training and validation accuracy are plotted to visualize the learning process:

plt.plot(history.history['accuracy'], label='accuracy')
plt.plot(history.history['val_accuracy'], label='val_accuracy')
plt.title('Training and Validation Accuracy')

🛠️ How to Use

1. Clone the repository:

   git clone https://github.com/yourusername/pneumonia-detection.git
   cd pneumonia-detection

2. Install the required dependencies:

   pip install -r requirements.txt

3. Download the dataset from Kaggle and extract it to the data/ folder.

4. Train the model:

   python train.py --model Xception

5. Evaluate the model:

   python evaluate.py --model Xception

🔍 Results

Model	Accuracy	Precision	Recall	F1-Score	AUC
VGG16	94.5%	93.2%	95.0%	94.1%	0.96
ResNet50	95.1%	94.6%	95.7%	95.1%	0.97
InceptionV3	95.8%	94.9%	96.0%	95.4%	0.97
MobileNetV2	94.0%	92.8%	94.2%	93.5%	0.95

📝 License

This project is licensed under the MIT License.