AI System Identifies COVID-19 Patients Who Require ICU

A new artificial intelligence (AI) system developed by researchers at the University of Waterloo and DarwinAI, an alumni-founded startup company, could help doctors efficiently utilize limited resources during the COVID-19 pandemic. The system is able to identify patients who require intensive care unit (ICU) treatment.

Determining ICU Necessity

The AI system predicts this necessity of ICU admission through the use of 200 clinical data points, which include blood test results, medical history, and vital signs.

Alexander Wong is a professor of systems design engineering and Canada Research Chair in AI and Medical Imaging at Waterloo.

“That is a very important step in the clinical decision support process for triaging patients and developing treatment plans,” Wong said.

The AI software was trained on data extracted from 400 cases at Hospital Sirio-Libanês in São Paulo, Brazil. It was based on whether doctors decided if COVID patients should be admitted for ICU care.

The neural network learned off of this data and was able to subsequently predict the need for ICU admission in new COVID cases with more than 95% accuracy. It is also able to identify the key factors that result in its predictions, which help clinicians gain a better understanding of the process.

A Tool for Health Officials

This technology is not meant to replace health officials, but rather act as a tool that can make faster and more informed decisions, which helps patients receive care when they need it.

Wong is also director of the Vision and Image Processing (VIP) Lab at Waterloo.

“The goal is to help clinicians make faster, more constituent decisions based on past patient cases and outcomes,” he said. “It’s all about augmenting their expertise to optimize the use of medical resources and individualized patient care.”

The technology is freely available to engineers and scientists so that it can continue to be improved upon. It is now being incorporated into a larger clinical decision support system, which was developed through the COVID-Net open-source initiative. This support system helps doctors determine the severity of COVID cases through AI analysis and medical images.

The research titled “COVID-Net Clinical ICU: Enhanced Prediction of ICU Admission for COVID-19 Patients via Explainability and Trust CQuantification” will be presented during a December 10 workshop at the 2021 Conference on Neural Information Processing Systems. The work also involved DarwinAI researchers Audrey Chung and Mahmoud Famouri, as well as engineering PhD student Andrew Hryniowski.