Sepsis Reasoning Cue Dashboard

This repository demonstrates the model training, patient selection, and dashboard design for the CHI'26 paper "Intelligent Reasoning Cues: A Framework and Case Study of the Roles of AI Information in Complex Decisions".

Setup Instructions

To set up, cd into the main repository directory, and run pip install -r requirements.txt. Then cd into the client directory and run npm install && npm run build.

Make sure you have a service account key JSON file that can access the GCS bucket, and set the environment variable like so:

export GOOGLE_APPLICATION_CREDENTIALS="/path/to/your/service-account-key.json"

Run python -m server to start the Flask server.

Model Training and Upload

The sepsis_models directory shows how the autoencoder was trained to produce the reasoning cues in our study. See the load_reasoning_cues.ipynb notebook for the full source code. (Please note that this script requires a custom preprocessed version of MIMIC-IV to work; please contact the authors if you would like to run this code as-is.)

Setting Up Study Protocol

Upload the study protocol JSON to the GCS bucket in a file named study_protocol.json. The study protocol used in our work is at the top level of this repository.

Participant accounts can be set up locally by running python blueprints/users.py.