Fig. 2

Goodness of fit and relationship between \(\upsigma\) and YWD. A Model goodness of fit for participants with AD using different atrophy patterns. B HIC AD FC matrix with empirical FC matrix above diagonal and simulated FC with optimal parameters below diagonal. C Latam AD FC matrix with empirical FC matrix above diagonal and simulated FC with optimal parameters below diagonal. D Linear regression of \(\sigma\) vs. YWD for HIC AD participants. E Linear regression of \(\sigma\) vs. YWD for Latam AD participants. F Comparison of the coefficient of determination (\(\sigma\) vs. YWD) between databases and genders in the AD subgroup, showing significant differences for Latam AD. G Model goodness of fit for participants with bvFTD using different atrophy patterns. H HIC bvFTD FC with empirical FC matrix above diagonal and simulated FC with optimal parameters below diagonal. I Latam bvFTD FC matrix with empirical FC matrix above diagonal and simulated FC with optimal parameters below diagonal. J Linear regression of \(\sigma\) vs. YWD for HIC Latam bvFTD participants. K Linear regression of \(\sigma\) vs. YWD for Latam bvFTD participants. L Comparison of the coefficient of determination (\(\sigma\) vs. YWD) between databases and genders in the bvFTD subgroup, showing no significant differences. SSIM = Structural Similarity Index Measure, d = Cohen’s d, HIC-Latam AD = HIC AD model with Latam AD atrophy, Latam-HIC AD = Latam AD model with HIC AD atrophy, HIC-Latam bvFTD = HIC bvFTD model with Latam bvFTD atrophy, Latam-HIC bvFTD = Latam bvfTD model with HIC bvFTD atrophy. HIC = High Income Country database. Latam = Latin American database. YWD = Years with disease. \(\mathrm{\sigma \sigma }\) = Scaling parameter of FIC as a function of local atrophy values. HIC = High Income Country database. Latam = Latin American database. AD: Alzheimer’s disease. bvFTD: behavioral variant frontotemporal dementia