Streamlit + SHAP: building fall risk app for Parkinson’s patients

Show the Community!
1

Hi everyone! :waving_hand:

I’m building a Streamlit-based web app that predicts fall risk in Parkinson’s patients using wearable sensors and explainable AI (SHAP values and SHAPSet plots).

I’m curious to learn:

  • How do you handle explainability in your Streamlit-based AI or healthcare apps?
  • What techniques or layouts help make the insights clearer for non-technical users like clinicians?

Here’s the GitHub repo & a demo:

Thanks in advance, really excited to hear your best practices!