Official Repository for "Ultra-efficient causal deep learning for Dynamic CSA-AKI Detection Using Minimal Variables"
REACT (Real-time Evaluation and Anticipation with Causal disTillation): a causal deep learning approach that combines the universal approximation abilities of neural networks with causal discovery to develop REACT, a reliable and generalizable model to predict a patient's risk of developing CSA-AKI within the next 48 hours.
Try dynamic early alerts of CSA-AKI at web-based platform.
Run our example at Google Colab to see how REACT works on simulated data.
git clone [email protected]:jarrycyx/UNN.git
cd UNN/REACT
conda create -n react_env python=3.8
conda activate react_env
pip install -r requirements.txt
You can run the notebook run_example.ipynb to see how REACT works on simulated data.
If you use this code, please consider citing our work.