filo-cena/Patient_digital-twin

This Space hosts a Streamlit app that demonstrates a clinical digital twin:

• —A 3D anatomical avatar rendered via <model-viewer> (GLB models).
• —A rule-based “LLM-style” interview agent that collects demographics and lifestyle factors and updates a toy clinical twin (organ risk indices, BMI, global health score).
• —An imaging-driven organ progression viewer that:
• —Loads volumetric CT/MRI-like data (NIfTI or .npy arrays).
• —Extracts organ meshes via marching cubes.
• —Colours the mesh with a risk field derived from intensities.
• —Simulates future trajectories by deforming high-risk regions.

The goal is to provide a blueprint for human digital twins that fuse:

1. 1.Generic anatomical reference models (Human Reference Atlas organs),
2. 2.Longitudinal imaging data, and
3. 3.Simple behavioural / lifestyle parameters.

🚀 How to run this Space

This is a Streamlit Space:

1. 1.Ensure the Space SDK is set to Streamlit.
2. 2.Place your main app file at the repository root as app.py (or change app_file above).
3. 3.Commit:
4. 4.app.py
5. 5.requirements.txt
6. 6.README.md
7. 7.the assets/ folder (see below).
8. 8.The Space will automatically install dependencies and start the app.

To run locally instead:

pip install -r requirements.txt
streamlit run app_advanced.py