🫁 Pneumonia Detection using Teachable Machine + Streamlit

A lightweight and effective chest X-ray classifier built using Teachable Machine (no-code) and deployed using Streamlit. The model predicts whether an uploaded X-ray image shows signs of Pneumonia or is Normal.

---

🔧 Tools Used

• Model Training: Teachable Machine
• Frameworks: TensorFlow, Keras
• Deployment: Streamlit
• Languages: Python

---

📊 Model Performance

Metric	Score
Accuracy	98%
Precision	98.1%
Recall	95.2%
F1-Score	96.6%

📌 Visuals included:

• Confusion matrix
• Accuracy/loss per epoch graphs

---

📁 Folder Structure

PneumoniaClassification/
├── Code/
│ ├── app.py #Streamlit app
│ ├── main.py #Prediction script
│ ├── keras_model.h5 #Trained model
│ ├── labels.txt #Class names
├── metrics/ #Evaluation visuals
├── requirements.txt #Python dependencies
└── README.md #This file

---

⚙️ Setup & Run

# Step 1: Clone the repo
git clone https://github.com/yourusername/pneumonia-detection-teachable.git
cd pneumonia-detection-teachable

# Step 2: Create virtual environment (optional)
python -m venv env
source env/bin/activate   # Windows: .\env\Scripts\activate

# Step 3: Install required packages
pip install -r requirements.txt

# Step 4: Run the Streamlit app
streamlit run Code/app.py

---

📦 requirements.txt

tensorflow==2.12.1
keras==2.12.0
opencv-python
numpy
Pillow
streamlit

---

▶️ How It Works

1.Upload an X-ray image in the Streamlit UI

2.The model processes the image

3.You'll get a prediction (Pneumonia/Normal) with a confidence score

---

✅ Next Improvements

• Add Grad-CAM for explainability
• Support more medical image types
• Convert model to TFLite for mobile usage
• Train on larger, real-world datasets

---

👤 Author

• Ahamed Ayyash
• Computer Engineering Student | Passionate about AI for Healthcare

📬LinkedIn: Ayyash Fous 💡Open for collaboration or internship opportunities

---

📜 License

MIT License — free to use, modify, and share with credit.