Brain MRI Classification (VGG16 & ResNet18)

This project provides a web-based system for classifying brain MRI scans into four categories using deep learning models (VGG16 and ResNet18). It features a FastAPI backend for inference and a Streamlit frontend for user interaction and model comparison.

---

Table of Contents

• Data Source
• Project Structure
• Features
• Usage
• Setup
• Model Performance
• Run with Streamlit Application
• PostMan API Testing
• Dockerization

---

Data Source

• Link: Brain Tumor MRI Dataset (Kaggle)
• Classes: Glioma, Meningioma, Pituitary, No Tumor
• Images: This dataset contains 7023 images of human brain MRI images

---

Project Structure

brain-mri-classification/
│
├── app/                         # All application code (Streamlit & FastAPI)
│   ├── appfast.py               # FastAPI backend
│   └── new_streamlit_app.py     # Streamlit frontend
│
├── models/                      # Trained model files
│   ├── vgg16_brain_mri_model.h5
│   └── resnet18_weights.pth
│
├── requirements.txt             # Python dependencies
├── Dockerfile                   # Docker build file
│
├── sample_mri_images/           # For MRI images downloaded from the internet to test the model
│
├── notebooks/                   # Jupyter notebooks
    └── Brain_MRI.ipynb

---

Features

• FastAPI backend for RESTful model inference (appfast.py)
• Streamlit frontend for uploading images, viewing predictions, and comparing models (new_streamlit_app.py)
• Pre-trained models: VGG16 (TensorFlow/Keras) and ResNet18 (PyTorch)
• Dockerized for easy deployment

---

Usage

• Upload a brain MRI image in the Streamlit app to get predictions from both models.
• Compare model performance and view detailed metrics and explanations.

---

Model Performance

• VGG16: 97% accuracy (macro avg precision: 98%, recall: 97%)
• ResNet18: 98% accuracy (macro avg precision: 98%, recall: 98%)

---

Setup

• Clone the Repository

  git clone https://github.com/Abdelrahman-Elshahed/brain-mri-classification.git

• Create and activate a virtual environment:

    python -m venv venv
    source venv/bin/activate  # On Windows use: venv\Scripts\activate

• Set up dependencies

  Install Python dependencies: To install all the required Python packages for the project, run the following command:

  pip install -r requirements.txt

---

Run with Streamlit Application

• Run FastAPI In one terminal

  cd app
  uvicorn appfast:app --reload --host 0.0.0.0 --port 8000

• Run Streamlit In another terminal

  cd app
  streamlit run new_streamlit_app.py

---

PostMan API Testing

• Health Check: Send a GET request to http://localhost:8000/health to verify the API is running.

• Model Prediction: Send a POST request to /predict/vgg or /predict/resnet with a brain MRI image file using the form-data key file.

Below is a screenshots of a successful prediction request in Postman:

---

Dockerization

• Build the Docker image with:

  docker build -t brain-mri-app .

• Run the container with:

  docker run -p 8000:8000 -p 8501:8501 brain-mri-app

• FastAPI docs: http://localhost:8000/docs
• Streamlit app: http://localhost:8501

---