Horizon 2020. Quantitative analysis of non-contrast thorax CT scans

Horizon 2020. Quantitative analysis of non-contrast thorax CT scans

The icovid project is the continuation of the pro bono icovid initiative, resulting in the AI-based CT analysis software, icolung. The icovid project aims to further improve the software, making it of greater value as the clinical needs evolve, validating it in renowned academic centers and deploying it at large-scale across Europe. icolung is expected to have a significant societal impact by increasing confidence when making a CT diagnosis and providing accurate quantification of disease and prognostic information in patients with suspected COVID-19 disease.

Importantly, even in a low prevalence setting, icolung may have a much higher sensitivity and a comparable specificity for the diagnosis compared to the RT-PCR testing used currently.

Identifying infected patients earlier will reduce the risk of further contamination and allows the right patient management to start earlier.

icolung also predicts the risk of developing severe COVID-19 disease for which ICU admission and mechanical ventilation are required, allowing optimization of patient care.

The consortium builds further on the existing organic collaboration and contains world-renowned experts of which multiple are part of the Executive Committee of the European Society of Thoracic Imaging.