Integrating Event‐Based and Network Diffusion Models to Predict Individual Tau Progression in Alzheimer's Disease

Our hybrid modeling approach enables individual-level prediction of tau spread patterns from cross-sectional data, outperforming previous attempts. Integrating EBMs and NDMs leverages the strengths and tackles the limitations of each technique: EBMs provide longitudinal information, while NDMs generate a biophysical basis for pathology progression. Our findings suggest AD encompasses more diverse tau spread patterns than classical Braak staging, with heterogeneity stemming from distinct seeding sites that converge over time. This framework could guide personalized therapeutic strategies and be applied to other neurodegenerative diseases, potentially improving early diagnosis and treatment selection.