Collaborative Research Project 6

Principal investigators: Ayis Pyrros (DuPage Medical Group), Alexander Schwing (University of Illinois Urbana-Champaign), Nasir Siddiqui (DuPage Medical Group), David Forsyth  (University of Illinois Urbana-Champaign)

Safe public training dataset for COVID-19 machine learning algorithms.

Updated January 20, 2023

This collaborative research project has been completed.

Screen Shot 2021-07-23 at 1.02.35 PM.png

Investigators have created a safe, publicly available synthetic dataset of COVID and non-COVID chest x-rays that may be used for training COVID-19 machine learning models. They also delivered a package for creating synthetic data from real images so that practices with sufficient computing resources can transparently share data without violating privacy or compromising accuracy.

CRP6.jpg
Previous

CRP 5: COVID pneumonia machine learning algorithm validation and visualization
Next

CRP 7: Leveraging registry data to conduct virtual clinical trials