Functional Data Analysis: Novel Statistical Methods and Applications in Alzheimer’s Disease Research

Seminar Series

Thursday, April 1, 2021 - 12:00
Zoom Conference
Sheng Luo, PhD

Abstract: Alzheimer's disease (AD) is a progressive neurodegenerative disease that causes impairment in multiple domains (e.g., cognition and behavior) and progresses heterogeneously in time and across domains and individuals. AD studies collect data from multiple sources: longitudinal clinical data (e.g., neuropsychological, functional, and behavioral assessments), neuroimaging (e.g., MRI, and PET), genetic (e.g., GWAS), and various biomarkers from biofluids (e.g., Aβ and pathologic tau from cerebrospinal fluid), together generating multimodal data. These multimodal data are predictive of AD related dementia occurrence and disease progression. In this talk, I will discuss a series of our recent papers based on functional data analysis techniques to analyze these multimodal data. Specifically, we developed a multivariate functional mixed model (MFMM) framework which models the longitudinal outcomes as multivariate sparse funcitonal data linked to event time data via a functional joint model. The predictive models provide accurate personliazed dynamically updated predictions of disease progression. The proposed methods have been applied to and validated in several large AD studies. 

Speaker:  Dr. Luo is professor of Biostatistics & Bioinformatics at Duke University. He is Fellow of American Statistical Association (ASA) and NIH NAME Study Section Chartered Member. Dr. Luo is a leading expert in multivariate longitudinal data analysis, functional data analysis methods, survival analysis, and medical imaging analysis. Dr. Luo has served as PI of multiple NIH grants on statistical methodology development. He is also PI of multiple projects regarding novel latent variable modeling and various neurodegenerative disorders rating scale development, validation, and translation funded by Parkinson’s Foundation, CHDI Foundation, and Movement Disorders Society.



Passcode: 375271