Call for Participation in COVID-19 Pneumonia CT Triage Study

Apr 21, 2020

CuraCloud is actively recruiting medical institutions to participate in a sponsored study on the clinical performance of a computer-aided triage system to support a 510(k) submission to the U.S. Food & Drug Administration (FDA). This would require IRB-approval for a proposed retrospective, de-identified clinical assessment of software for use in the analysis of adult non-contrast chest CT images of patients with suspected pneumonia. The device is intended to assist hospital networks and trained radiologists in workflow triage by flagging suspected positive findings of pneumonia characteristic of COVID-19 disease. This is not a population screening application.

Participating institutions will provide de-identified DICOM files from chest CT studies of patients with suspected pneumonia. Institutions will also provide information regarding which of the cases were confirmed by RT-PCR testing.

Based on preliminary research, our multi-disciplinary team developed a three-dimensional deep learning framework to detect COVID-19 using chest CT, and differentiate it from community acquired pneumonia and other lung diseases. This technology was deployed by our collaboration partners in China and trained from more than 3,000 patients’ data acquired from six medical centers. Published results in the peer-reviewed journal, Radiology, indicate that the model achieved high performance accuracy (above 0.95 AUC) in terms of identifying COVID-19 pneumonia from community acquired pneumonia. We and our collaborators have built upon this research work and developed the COVID-19 pneumonia CT triage software to be assessed during this study.

If you work for a healthcare institution with qualified data that you would like to include in this study, please contact us.