Case Study: Banner Health advances COVID CT diagnosis research with Google for Education

Banner Health uses Google for Education to analyze 300 CT scans and find COVID lung injury patterns 10x more often

Banner Health, through its cardiothoracic radiologist Dr. Kathryn Olsen, faced the challenge of analyzing a large volume of CT scans to determine if unique lung injury patterns could help diagnose COVID-19 and predict patient outcomes. The project required a secure and scalable infrastructure to manage and process sensitive medical imaging data from multiple collaborating healthcare institutions. The solution was implemented using the Flywheel data management platform, which is hosted on Google for Education's Google Cloud.

Using Flywheel on Google Cloud, Dr. Olsen's team established a centralized repository to securely store, de-identify, and manage the complex imaging datasets. This enabled a blinded reader study involving multiple radiologists and later, AI analysis from Imbio. The project confirmed that COVID-19 patients are over ten times more likely to show a specific lung injury pattern on CT scans. Google for Education's cloud infrastructure facilitated this unprecedented collaboration, leading to preliminary findings that can improve diagnostic criteria and potentially predict clinical outcomes.

Banner Health