Speaker: Sheng Luo
From: Duke University, Biostatistics
Title: Integrative Modeling and Dynamic Prediction of Alzheimer’s Disease

Abstract:
Alzheimer’s disease (AD) is a devastating neurodegenerative disorder characterized by cognition and functional impairments, exhibiting heterogeneous progression over time and across individuals. AD studies leverage diverse data sources, including longitudinal clinicalassessments, neuroimaging data, genetic information, and biomarkers from biofluids. These multimodal datasets hold significant potential for predicting AD-related dementia occurrence and disease progression.

In this presentation, I introduce a novel integrative modeling framework that harnesses multimodal data to build and validate AD predictive models. Our approach accommodates diverse data modalities and patient characteristics, providing a comprehensive understanding of AD progression. Key to our method is the multivariate functional mixed model (MFMM) framework, efficiently connecting longitudinal outcomes and event time data through a functional joint model. The predictive models deliver accurate personalized dynamic predictions of disease progression.

The talk explores the real-world implementation and application of our proposed framework in various large-scale AD studies, rigorously validated and supported by Dr. Luo’s current R01 grant. Our innovative approach contributes to ongoing research on early detection and improved understanding of AD progression, bringing us closer to effective therapeutic interventions and personalized patient care.