The researcher at the Hospital del Mar Medical Research Institute (IMIM) and physician at the Hospital del Mar Emergency Service, Max Hardy-Werbin, as well as the bioinformatician at the Pathological Anatomy Service of the same hospital, Joan Gibert, have led the development of a new support tool for the diagnosis of patients with suspected infection with the SARS-CoV-2 coronavirus. This tool, developed thanks to the collaboration of different areas and services of the Hospital del Mar and the IMIM, uses Artificial Intelligence to help medical personnel diagnose, from chest X-rays, if the patient is positive for COVID-19.
For the development of the project, more than 6,000 chest radiographs of patients from Hospital del Mar have been used, some positive for COVID-19, others not; but all of them, reported by radiologists. Thanks to the large sample of radiographs, the level of reliability of the tool is more or less 80%. In addition, the fact that this new tool has been trained with its own samples increases its adaptation to the reality of Hospital del Mar.
“It is a complementary tool to support the decision of clinical professionals, in addition to the rest of the existing elements, the PCR test, clinical symptoms and the radiographic image itself”
The project is at the implementation phase, so it is being incorporated into the computer system of all the Parc de Salut Mar centres; healthcare professionals have quick and easy access to it. As Hardy-Werbin says, “the implementation of this new tool comes at a fundamental moment, with an increase in negative COVID-19 cases coming to the emergency service. Having this tool will be very useful for us to manage each patient and direct them to the circuit indicated for their pathology”.
The head of the Radiology Service at the Hospital del Mar, Jose Maria Maiques, anticipates that the tool, in the case of positive patients, will also be quite helpful to develop models with prognostic value for the disease; for example, the need for mechanical ventilation.