🩺 Diabetes Risk Diagnostic Tool

A machine learning-powered web application that estimates diabetes risk from clinical health metrics, built with a FastAPI backend and a Next.js frontend. Originally a Streamlit app, rebuilt into a full separate backend/frontend architecture.

---

🚀 Live Demo

Try the live app

---

✨ Features

• AI-powered predictions — a Random Forest Classifier (scikit-learn) trained on the Pima Indians Diabetes Database estimates risk from 8 clinical inputs.
• Population comparison — each input shows the dataset's actual average alongside your entry, for context.
• Session history — every check you run stays visible in a running list for the current session.
• Downloadable report — generates a plain-text summary of your inputs and result.
• Location-aware doctor recommendation — if a result reads High Risk, a button finds nearby endocrinologists (Oladoc in Pakistan, Google Maps elsewhere), based on your browser's location.
• Animated ECG pulse visual — a live-drawing heartbeat-style line that changes color depending on the result.

🛠️ Tech Stack

• Backend: FastAPI, scikit-learn, pandas, deployed on Render
• Frontend: Next.js (App Router, JavaScript, Tailwind CSS), deployed on Vercel
• Model: Random Forest Classifier, ~80.5% accuracy, trained on the Pima Indians Diabetes Database

🚀 Installation & Local Setup

Backend

cd backend
python -m venv venv
venv\Scripts\Activate.ps1
pip install -r requirements.txt
uvicorn app.main:app --reload

Runs at http://127.0.0.1:8000.

Frontend

cd frontend
npm install
npm run dev

Runs at http://localhost:3000. Requires a .env.local file containing:

NEXT_PUBLIC_API_URL=http://127.0.0.1:8000

📂 Project Structure

DiabetesDetector/
├── backend/
│   ├── app/
│   │   ├── main.py          # FastAPI app: loads model, exposes /predict
│   │   └── model.pkl        # Pre-trained Random Forest model
│   └── requirements.txt
├── frontend/
│   ├── app/
│   │   ├── layout.js        # Fonts, metadata
│   │   ├── page.js          # Main UI: form, pulse animation, results
│   │   └── globals.css      # Design tokens (colors, fonts)
│   └── package.json
├── Notebook/
│   ├── explore.ipynb        # Data analysis & model training
│   └── model.pkl
├── assets/
│   ├── confusion_matrix.png
│   └── DiabetesDetector.gif
└── model.py                 # Alternate training script (Pipeline + scaler)

🧠 Model Insights

The model achieves ~80.5% accuracy. Confusion matrix on held-out test data:

The model is particularly strong at correctly identifying healthy patients, with a focus on minimizing false negatives.

---

👤 Author

Syed Ali Faraz — GitHub Profile

If you found this tool insightful, please give the repository a ⭐!