A project funded by Instituto de Salud Carlos III and led by IFIC researchers is carrying out radiological imaging analysis using machine learning techniques with the aim of enhancing patient diagnose and evolution assessment.
The project, which includes other research centres of the Valencian Community such as FISABIO and UPV, has at its disposal a large dataset of chest pictures of anonymized patients. Models of the illness impact on chest images are developed by ML techniques which are then validated at IFIC with the help of radiologists and other experts.
For the development of these models the team is making intensive use of the Artemisa platform at IFIC in order to accelerate model training. Under radiologist supervision, the team expects to provide a screening tool which will enable early detection of suspicious cases by physicians, anticipating the need of tracking by the first healthcare services.
The recent huge increase in the number of COVID19 cases can make these methods an invaluable, complementary medical tool.