個別の脳接続モデルによるタウ病理の広がり予測、アルツハイマー病の進行と不均一性に関する洞察を提供

Heterogeneity in regional tau burden limits evaluation of disease progression using one-size-fits-all approaches. Prior work using tau positron emission tomography (PET) has highlighted the important role of connectivity to epicenters of tau pathology in explaining this heterogeneity. However, previous studies using population-based epicenters or connectomes fall short of a fully individualized approach to predicting regional tau burden. We use diffusion MRI-derived structural connectomes to assess the prediction of regional tau burden using distance along individual structural connectomes from individualized epicenters of tau pathology both cross-sectionally and longitudinally. Fully individualized models of connectivity and epicenters outperformed models using either population-based connectomes or epicenters in explanation of cross-sectional and longitudinal regional tau burden, improved prediction in validation datasets, and produced stronger single-subject level prediction. Together, these findings demonstrate the power of a fully individualized approach to explain regional tau heterogeneity and provide the strongest in vivo evidence to date for network-based spread of tau pathology.