The repository contains the implementation for "Predicting Neurological Outcome in Comatose Patients after Cardiac Arrest with Deep Neural Networks".
We develop a multiscale deep neural networks (CNN-LSTM) using continuous EEG and admission characteristics from a multicenter cohort of more than 1000 cardiac arrest patients. The ICARE dataset contains approximately 58,000 hours of clinical EEG data, demographic information from seven hospitals in Europe and the U.S.
The seven hospitals were Medisch Spectrum Twente (Enschede, Netherlands), Rijnstate Hospital (Arnhem, Netherlands), Erasmus Hospital (Brussels, Belgium), Yale New Haven Hospital (New Haven, CT, USA), Brigham and Women’s Hospital (Boston MA, USA), Beth Israel Deaconess Medical Center (Boston, MA, USA), and Massachusetts General Hospital (Boston MA, USA).
@article{zheng2021outcome,
title = {Predicting Neurological Outcome in Comatose Patients after Cardiac Arrest with Deep Neural Networks},
author = {Zheng, Wei-Long and Amorim, Edilberto and Jing, Jin and Ge, Wendong and Hong, Shenda and Wu, Ona and Ghassemi, Mohammad and Lee, Jong Woo and Sivaraju, Adithya and Pang, Trudy and Herman, Susan T. and Gaspard, Nicolas and Ruijter, Barry J. and Sun Jimeng and Tjepkema-Cloostermans, Marleen C. and Hofmeijer, Jeannette and van Putten, Michel J. A. M. and Westover, M. Brandon},
journal={Resuscitation},
year={2021},
publisher={Elsevier}
}