DRAGON sets out to combine the expertise of a variety of partners in the field of AI and machine learning to develop a decision support system for precise COVID-19 diagnosis and patient outcome prediction using CT scanning. 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. The health system capacity can be increased, the pace of research and innovation can be sped up and people can be empowered.  

Screening, triage, and severity assessment of COVID-19 patients for precise diagnosis and prediction of disease progression and recovery, based on chest CT analysis with artificial intelligence. 

Within DRAGON, Thirona develops and contributes with AI tools that will allow for precise diagnosis and prediction of disease progression and recovery, building on its existing CAD4COVID-CT software. Thirona’s CAD4COVID-CT uses a standardized measurement of pulmonary involvement of COVID-19 to automatically assess CT scans and quantify the measure of COVID-19 suspicion, as well as the severity of the infection based on the percentage of affected lung area.  

The COVID-19 pandemic has illustrated significant shortcomings in lack of response capacity and efficiency in the healthcare system. The DRAGON project aims to make patients more aware of their options and provide better understanding of patients’ perspectives for the clinicians.  

CT scans were used by physicians since the very start of the pandemic and proved to be a sure way to identify infections. However, the traditional way of CT scan assessment is time-consuming and is especially prone to error in the conditions of heavy workload and high pressure on the radiologist.  

The DRAGON project focuses on partnering up with high-tech SMEs, academia, BioTech and Pharma to draw on new and existing data and sample collection, including CT scans to carry out detail patient profiling. AI is then used to transform this information into a precision medicine approach to help clinicians and patients with decision-making around treatments. 

 As a result of one of the conducted research projects within DRAGON, Thirona and CHU Liège worked together to validate the precision of CAD4COVID-CT. The research showed the prognostic powers of the algorithm, that could help predict the likelihood of ICU admission and ventilation, showing how AI could facilitate planning care for such patient cases.  

This contributes to the overarching goal of the DRAGON project, easing the workload on healthcare, resulting in a clinical decision support system for patients’ triage and prompt implementation of the indicated therapy. We continue to work together with DRAGON by developing and incorporating our deep learning capabilities and using data to achieve the goals set forth by the project.  

Subsequently, the progress made with CAD4COVID within DRAGON led us to making it available for clinical use after conducting extended validation studies and completing CE and MDR certification. The CAD4COVID software is now included in Thirona’s LungQ 2.0 Clinical Suite, supporting pulmonologists in hospitals at an international scale through AI applications platforms. 

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.