Neurodegen-Vis

Welcome to Neurodegen-Vis - Privacy-Preserving Visual Exploration of Healthcare Data with a Language Model Integration

Neurodegen-Vis is a visual analytics tool designed to explore correlations and dependencies in patient data while maintaining privacy through differential privacy techniques. It integrates a Large Language Model to assist users in interpreting complex healthcare data and navigating dashboard interactions.

The work was published in the Joint Proceedings of the ECAI 2025 Workshops XAI-KRKG and UCEX-XAI. The dashboard is also available on a demo server.

🚀 Try the Demo | 📄 Read the Paper | 📥 Download PDF

Key Features

• Privacy-First Analytics: Data is anonymized using Differential Privacy to protect patient confidentiality.
• Dataset: The example dataset includes demographic information, cognitive test results, and treatment data from Parkinson’s disease patients.
• LLM Assistance: An integrated chatbot acts as a co-pilot to help users understand, explore, and analyze the dashboard and the underlying data:
  • Context Awareness: The model is aware of the dashboard’s visualizations, interactions, feature descriptions, and correlation results.
  • Automatically Updated Visualizations: Interacting with the chatbot dynamically updates the UI. For example:
    • Highlighting specific heatmap cells and switching to the corresponding scatterplot when discussing correlation results.
    • Switching to the corresponding histogram view when asked about feature explanations or distributions.
  • Exploration: Using the Suggest Features function proposes feature dependencies that could be interesting to analyze and automatically updates the scatterplot accordingly.

Run Locally

If you want to test the dashboard, it is recommended to use the demo server

However, it can be also run locally by

1. Add an OPEN_API_KEY and the serverflag in the .env file
2. Change the URL base path in vite.config.ts file.
3. Create the docker build

Detailed Description

1. Add an OPEN_API_KEY and the serverflag in the .env file

You additionally have to add an Language model API in a .env file, which has to be located in the base folder. Additionally you have to define the SERVER flag as false. The .env file should look like this:

#.env
OPENAI_API_KEY=sk-proj-...
SERVER=false

2. Change the URL base path in vite.config.ts file.

In the file frontend\vite.config.ts change from

  // Deployment on Demo Server.
  base: "/neurodegen-vis",

To:

  // to run locally via Docker, uncomment the following line:
  base: "./",

3. Creating a docker build

After you completed the first 2 steps, run

docker compose up --build

to create the docker build.

If you found this work helpful, please cite:

@inproceedings{2025-neurodegen-vis,
title = "Privacy-Preserving Visual Exploration of Healthcare Data with a Language Model Integration",
author = "Michael Grabner and Peter Waldert and Benedikt Kantz and Tobias Schreck",
year = "2025",
month = oct,
day = "25",
language = "English",
booktitle = "Joint Proceedings of the ECAI 2025 Workshops XAI-KRKG and UCEX-XAI, Bologna, Italy",
note = "28th European Conference on Artificial Intelligence, ECAI 2025, ECAI ; Conference date: 25-10-2025 Through 30-10-2025",
url = "https://ecai2025.org/",
}

This work was funded by the HERIDTARY project.