Fighting COVID-19 requires tremendous resources and overpowers many health systems. When it comes to imaging services for initial diagnostics and follow-up, there are two specific shortages in hard-hit areas: trained radiologists and machine capacity. High quality imaging devices such as CTs are not widely available in all regions and performing CT scans on ventilated severely ill patients is extremely complex, risky and time-consuming. At the same time, it is difficult to obtain all the diagnostic information doctors need from standard x-rays.
FastRAi will involve radiology experts in support of their front-line colleagues and use the data obtained to create AI that can enhance the information obtained from lung x-rays. In the first part of the project, expert radiologists will support over-burdened hospitals dealing with COVID-19 by providing high quality reports on the medical images obtained at the hospitals. During this phase, the project will build a GDPR-compliant digital biobank of the images that are reviewed. Within six months, these images will allow the development of an AI algorithm that can project x-rays onto CT images and build up data on the progression of lung disease. Using this data, the FastRAi system will increase the information that doctors obtain from standard x-rays. The system will be able to analyse x-rays and detect features that are invisible to the human eye, in order to better estimate the progression of a patient’s lung disease.
In the first phase of the project, the impact will be improved support for radiologists in the field addressing COVID-19. Once the FastRAi system is fully developed, it will assist in improved reading of x-rays. Through both phases, doctors will know more about the existence and progression of lung disease from standard x-rays. This will allow them to give better diagnosis and treatment to patients with COVID-19.
EIT Health Partner
Technical University of Munich