/EcMOdel

Our goal is to develop a machine learning prediction algorithm for neurological injury in pediatric cardiac patients.

Primary LanguageJupyter NotebookMIT LicenseMIT

EcMOdel

Our goal is to develop a machine learning prediction algorithm for neurological injury in pediatric cardiac patients.

The biggest risk factor for Pediatric EcMo patients is stroke. We have built a framework to statistically assess actionable events leading to stroke and improve neurological outcomes.

How to use

* Required Library

  • Python 3.6
  • Keras

Installing from Github

https://github.com/NCBI-Hackathons/EcMOdel.git

Add raw data into '/data' folder.

Run 'main.ipynb' to train/test model.

Final results are saved in '/results' folder.

Contributors

  • Abdelaziz Farhat, MD - UT Southwestern, Children's Medical, Pediatric Critical Care
  • Neel Shah, MD - UT Southwestern and Children's Medical, Pediatric Critical Care
  • Ziheng Wang, PhD - UT Dallas, Mechanical Engineering
  • Jeon Lee, PhD - UT Southwestern, Bioinformatics
  • Rafe McBeth, PhD - UT Southwestern, Radiation Oncology