![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
… is based on a platform architecture of patient data in combination with research structures for analyzing these data. All German university radiology departments participate by contributing radiological COVID-19 patient data. This will comprise an extensive database of computed tomography and thoracic radiography data, linked to relevant clinical patient information regarding the disease.
Structured reporting enables the pooling of decentrally collected data under high standards of standardization and quality assurance. Such data will be the basis for high quality research studies. A dashboard early warning system will help to fight future pandemics.
At i²LAB we plan to develop AI models for assessing disease severity. Specifically, we will develop an image based quantiative COVID-19 score (iqCS) which could be used to predict the patient trajectory, specifically identifying those that are at greatest risk for treatment at the intensive care unit and high long term morbidity or mortality. Moreover we will use AI models to predict CT based risk scores from chest radiographs.
Disease severity of COVID-19 patients varies substantially and identification of risk factors for severe outcome is important to guide preventive measures. We will assess the contribution of musculo-skeletal frailty based on vertebral fracture and muscle status.