Betes the Stroke

Empowering Diabetics with Smart Meals and Risk Insights.

Comment

💡 Inspiration

In the U.S., diabetes and obesity are on the rise, and regular doctor consultations are expensive. We wanted to build something that could help diabetic patients manage their condition more affordably. Inspired by this need, we created an app that suggests personalized nutrition plans and predicts stroke risk—one of the most severe complications of diabetes.

📱 What it does

Betes the Stroke helps diabetic users:

  • Predict their risk of stroke using their personal health data
  • Get daily personalized macronutrient and meal recommendations
  • Receive motivational achievements based on healthy lifestyle habits

Our goal is to support users in preventing complications while making smarter everyday health decisions.

🛠️ How we built it

  • React + Vite + Tailwind CSS for the frontend
  • Flask (Python) for the backend API and AI model
  • MongoDB for storing user input data
  • Scikit-learn to train a RandomForest-based stroke prediction model
  • NHANES health dataset to develop a nutrition recommendation model
  • JWT for user authentication

🧩 Challenges we ran into

  • Extracting and cleaning real-world datasets (like NHANES)
  • Making AI predictions interpretable and actionable for users
  • Integrating multiple systems (frontend, backend, database, and AI) smoothly
  • Preventing the AI model from giving inconsistent results due to stochastic behaviors

🏆 Accomplishments that we're proud of

  • Successfully deployed a working AI-driven app within a short timeframe
  • Built a functioning nutrition recommendation system from scratch using real diabetic data
  • Designed a clean, user-friendly frontend experience
  • Implemented secure authentication and data flow across components

📚 What we learned

  • How to preprocess and extract features from public health datasets
  • How to apply prescriptive machine learning for daily health management
  • How to work collaboratively across a full-stack application pipeline
  • The importance of user experience when building health-oriented apps

🚀 What's next for Betes the Stroke

  • Integrate GPT-4 Vision to classify food or health-related images and auto-fill reports
  • Add a stroke prediction probability chart and historical tracking
  • Expand the achievement system to encourage long-term engagement
  • Launch a real-time chatbot to provide immediate health guidance

Built With

Share this project:

Updates