The results of the study conducted together with the University Hospital of Liège, Belgium (CHU Liège), show that AI significantly increases efficiency in predicting the risk of deterioration of the COVID-19 patient’s condition. This study demonstrates that AI could help better manage the patient flow in hospitals, relieve the workload of radiologists, be used as a clinical decision support system for patients’ triage and promptly guide the indicated therapy.

Ever since the start of the pandemic, CT scans have been a sure way to identify COVID-19 infections even in the earliest stages. However, the usual way of assessing the severity of the disease through CT scans is time-consuming and can be especially error-prone when the physician workload is heavy.

CAD4COVID-CT

Based on a standardized measurement of pulmonary involvement of COVID-19, Thirona’s CAD4COVID-CT conducts automatic assessments of CT scans and provides a quantitative measure of COVID-19 suspicion, as well as the severity of the infection based on the percentage of affected lung area. The results of the algorithm were cross-checked with CT assessments by experienced radiologists, confirming its high prognostic performance in predicting disease severity. Developed in collaboration with our partners at Radboud UC, Fraunhofer MEVIS, and HT Medica, CAD4COVID-CT was created to help triage COVID-19 suspects and inform on disease progression and recovery.

How can AI help?

CHU Liège clinical experts worked with Thirona’s CAD4COVID-CT AI-based software package to explore the prognostic value of its results on hospitalization severity of patients with COVID-19. The image analysis indeed shows high correlation between the degree of lung injury detected by the software and the likelihood of ICU admission and ventilation. This is yet another benefit of AI, as these are the key factors when caring for people hospitalized with COVID-19, as there is a limited number of ICU units, ventilators and need for specialized staff to monitor severe cases.

The results of the study show that AI significantly increases efficiency in predicting the risk of deterioration of the patient’s condition. This indicates that AI could help better manage the patient flow in hospitals, relieve the workload of radiologists, be used as a clinical decision support system for patients’ triage and promptly guide the indicated therapy.

DRAGON

This research was conducted as part of DRAGON - an Innovative Medicines Initiative funded project started in the midst of the pandemic back in October 2020. The project is focused on partnering up with high tech SMEs, academia, BioTech and Pharma with the participation of the European Lung Foundation and the European Respiratory Society. Its purpose is to go beyond flattening the curve in order to dramatically elevate the response capacity of the health system to the coronavirus pandemics. Within DRAGON, Thirona develops AI tools that will allow for precise diagnosis and prediction of disease progression and recovery, building on its existing CAD4COVID-CT software.

For more information on the research: https://www.frontiersin.org/articles/10.3389/fmed.2022.930055/full

Learn more on the DRAGON project: https://europeanlung.org/dragon/about-the-project/

The DRAGON project has received funding from the Innovative Medicines Initiative 2 Joint Undertaking (JU) under grant agreement No 101005122. The JU receives support from the European Union’s Horizon 2020 research and innovation programme and EFPIA.