The World Health Organisation indicates that over 70% of deaths worldwide is attributable to chronic non-communicable diseases (NCDs), of which biological risk factors usually emerge in adulthood. Currently, risk prediction tools are used in adults; although, many NCDs have early precursors that manifest at a young age. The identification of high-risk individuals and early behaviour change strategies could improve the primary prevention of NCDs.

SmartCHANGE uses machine-learning methods, that can train models on heterogeneous datasets (i.e. behaviour, fitness levels, biomarkers), to enable accurate risk prediction for people of various ages; while federated learning techniques are employed to ensure data privacy. These methods are leveraged to design two apps: one for health professionals and the other for citizens, which are able to communicate in a bi-directional manner and reduce long-term risk of NCDs. Engineering is responsible for the development of a web-app for health professionals, which integrates an explainable Clinical Decision Support System and for the overall architecture design and subsequent configuration, based on microservices paradigm. The project has received co-funding from the European Union's Horizon Europe programme - Contract No. 101080965.