Emulation of dynamic multi–output simulator of risk of type-2 diabetes

We have recently developed and validated multilevel patient-specific model able to integrate metabolic, nutritional and lifestyle data for the prediction of the metabolic and inflammatory processes underlying the development of type-2 diabetes in the absence of familiarity. Given the incidence of type-2 diabete, which accounts for 85-90% of all cases of diabetes in the world, the implementation of this predictive model on mobile devices could provide a useful instrument to assess the risk of type-2 diabete by informed and aware individuals. However, give the high computational cost of this model, being a mixture of agent-based and ordinary differential equations and providing a dynamic multivariate output, it can run only on powerful workstations but not on mobile devices. The aim of the present paper it to construct an emulator model, using a machine learning approach, with reduced computational cost so to run on mobile devices to provide real time self-monitoring.