❤️ Cardiac Detection Using CNN (ResNet-18)

Welcome to the Cardiac Detection Project, where we harness the power of deep learning to identify cardiac abnormalities from medical imaging data.

This project utilizes a Convolutional Neural Network (CNN) based on ResNet-18 to detect cardiac abnormalities. The model is trained on annotated medical images to predict bounding boxes around key cardiac regions.

• Model: ResNet-18 (customized for cardiac detection)
• Framework: PyTorch Lightning ⚡
• Visualization: Class Activation Mapping (CAM)
• Dataset: Cardiac Imaging Dataset

• ✅ ResNet-18 Backbone: Pretrained model fine-tuned for cardiac image analysis
• 📊 High Accuracy: Optimized for robust detection performance
• 🔍 Class Activation Mapping (CAM): Visualizes key regions influencing predictions
• ⚡ Efficient Training: Leveraging PyTorch Lightning for fast experiments

├── CardiacDetection.ipynb   # Jupyter notebook for training & evaluation
├── logs/                    # Training logs & checkpoints
├── checkpoints/             # Model weights (.ckpt files)
└── README.md                # Project documentation

1️⃣ Clone the Repository:

git clone https://github.com/SYEDFAIZAN1987/Cardiac-Detection-using-CNN.git
cd Cardiac-Detection-using-CNN

2️⃣ Install Dependencies:

pip install -r requirements.txt

3️⃣ Run the Model:

python CardiacDetection.py  # Or open the notebook CardiacDetection.ipynb

4️⃣ Visualize Results:
Check outputs and bounding boxes in the generated logs.

• Bounding boxes correctly identify cardiac regions.

• Syed Faizan – Project Lead & Developer

For queries, feel free to reach out:

• GitHub: @SYEDFAIZAN1987
• Email: your.email@example.com

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