Hospitals are under unprecedented pressure to accurately monitor disease progression and predict which COVID-19 patients needing ICU support. University of Copenhagen, and Copenhagen University Hospital (Rigshospitalet) have been awarded a grant by Novo Nordisk Foundation to develop an applied artificial intelligence prototype that can provide relevant stakeholders the severity of COVID-19 pneumonia based on clinical parameters and chest x-rays, contracting Cerebriu to provide chest x-ray AI and automated risk assessment scoring.
Clinical COVID-19 diseases range from asymptomatic infection, to mild symptomatic disease, severe disease requiring hospitalization and intensive care unit admission including respiratory support, and death. Hospitals need a risk stratification model based on laboratory findings, clinical findings and imaging findings associated with severe illness and mortality to support clinical decision making in critical, high stress scenarios, to ensure best quality care for patients and the best use of limited resources during this crisis.
Imaging has been used sparingly in relation to COVID-19. A mix of chest computer tomography (CT) and x-ray have been used as a second-line test due to long turnaround times of molecular testing RT-PCR. There have been mixed impacts in terms of the use of imaging, especially CT, due to long disinfection times of the machine necessary after each scan. In such circumstances, chest x-ray is preferred.
“The hope is that such models can be used by frontline staff when dealing with new COVID-19 patients as well as by planners in the health service,” says Mads Nielsen, Head of Department of Computer Science, University of Copenhagen.
Cerebriu applies expertise in time of need
At the dawn of COVID-19 implementation of recently CE-marked software solution Cerebriu Apollo is postponed as hospitals postpone all R&D activities to focus on primary care.
“No one knows how long this coronacrisis will last, but Cerebriu has a world-class AI team within medical imaging, so if we can help improve workflow and save lives, that is our priority until we’re invited to implement our solutions again,” states Robert Lauritzen, Cerebriu’s CEO and Co-Founder.
Cerebriu’s vision is to expand from current CE-marked Cerebriu Apollo brain tumor and apoplexy clinical workflow support into other clinical workflows relying on diagnostic imaging.
“The chest x-ray risk assessment modelling of COVID-19 for monitoring and triage of patients is strategically aligned with our vision to provide fundamental technology to dramatically decrease rating time for radiologists, reducing burden on healthcare professionals and increasing overall quality of risk assessment,” Robert Lauritzen continues.