The virtual aging brain model as a digital twin for making individualized predictions

Understanding individual trajectories of age-associated effects on the brain is a prerequisite for developing specific preventive or treatment strategies. Using a biologically plausible modeling approach based on The Virtual Brain (TVB) framework, we virtually simulated effects of specific age-related structural connectivity decline on the functional architecture of the subjects based on their neuroimaging data and related it to their cognitive performance. Using simulation-based inference (SBI), we could demonstrate that the simulated age effects closely resembled those observed empirically in a large group of subjects from a population-based cohort sample. The fluidity of the whole brain network decreased with aging, shifting the optimal network working point in the empirical as well as simulated setup. It could be shown that subjects with lower cognitive performance exhibited more pronounced changes in the optimal network working point than subjects with better preserved cognitive performance. This virtual aging model demonstrated the potential to use the digital twin technology to simulate effects on the brain’s health beyond those shown for specific disease states.