README.md

metadata

license: mit
language:
  - en
metrics:
  - accuracy
base_model:
  - frgfm/resnet18
pipeline_tag: image-classification
datasets:
  - Baldezo313/rsna-pneumonia-dataset

🦠 Pneumonia Classification Model

Welcome to the Pneumonia Classification Model repository! This project utilizes a deep learning approach based on the powerful ResNet-18 architecture to classify chest X-ray images and detect pneumonia with high accuracy.

🚀 Project Overview

This repository contains code for:

• Training a ResNet-18 model for pneumonia classification
• Evaluating model performance with detailed accuracy metrics
• Visualizing key features using Class Activation Mapping (CAM)

📊 Key Features

✅ ResNet-18 Backbone: Fine-tuned on chest X-ray images for accurate pneumonia detection
✅ Model Evaluation: Includes both weighted and non-weighted accuracy assessments
✅ Class Activation Mapping (CAM): Visualizes critical regions influencing the model’s predictions
✅ Efficient Training: Optimized data pipelines for fast model training and inference

---

🗂️ Repository Structure

Pneumonia-Classification-Model/
├── PneumoniaClassification.ipynb   # Jupyter Notebook for model training & evaluation
├── checkpoints/                    # Saved model weights
├── data/                           # Dataset folder (chest X-ray images)
├── outputs/                        # Generated CAM images & evaluation results
└── README.md                       # Project documentation

---

⚙️ Installation

1. Clone the Repository:

git clone https://github.com/SYEDFAIZAN1987/Pneumonia-Classification-using-resnet-18-based-model-with-evaluation-and-CAM.git
cd Pneumonia-Classification-Model

2. Install Dependencies:

pip install -r requirements.txt

3. Run the Model:

jupyter notebook PneumoniaClassification.ipynb

---

🔍 Model Performance

Non-Weighted Accuracy:

Weighted Loss Accuracy:

Class Activation Mapping (CAM):

Visual representation of regions critical to the model’s pneumonia classification:

---

📈 Evaluation Metrics

• Precision, Recall, F1-Score for performance evaluation
• Confusion Matrix for classification analysis
• Weighted vs Non-Weighted Accuracy comparison

---

🤝 Contributing

Contributions are welcome! To contribute:

1. Fork the repository
2. Create a new branch: git checkout -b feature-branch
3. Commit your changes
4. Open a pull request 🚀

---

📜 License

This project is licensed under the MIT License.

---

👨‍⚕️ Author

Developed by Syed Faizan
For any queries or collaborations, feel free to connect on GitHub.

⭐ If you found this project useful, please give it a star! ⭐