|
Activity Number:
|
121
- Emerging Statistical Methods for Structured and Multimodal Data Analysis
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, August 3, 2020 : 1:00 PM to 2:50 PM
|
|
Sponsor:
|
Section on Statistics in Imaging
|
|
Abstract #312662
|
|
|
Title:
|
Multi-State Markov Transition Models for Examining Multimodal Imaging Signatures of Alzheimer's Disease
|
|
Author(s):
|
Zoe Zhang* and Ashley Heywood and Jane Stocks and Lei Wang
|
|
Companies:
|
Drexel University and Northwestern University and Northwestern University and Northwestern University
|
|
Keywords:
|
Neuroimaging Data Analysis;
Alzheimer's Disease;
Multi-state Model
|
|
Abstract:
|
Identifying antemortem biomarkers for progression from mild cognitive impairment (MCI) to dementia of Alzheimer‘s type is crucial to detect areas particularly sensitive to neurodegeneration caused by Alzheimer’s disease (AD), in addition to potentially providing a strong diagnostic tool at early disease stages. Jointly modeling the multiple states including normal control (NC), MCI, and dementia as well as the transition probability between them provides a new approach to unravel the mechanism underlying the clinical progression of AD with enhanced statistical power. To deal with a large number of imaging features, we incorporate penalization in multi-state Markov transition models to select predictive imaging signatures of AD. We demonstrate the performance of the proposed approach using MRI and metabolism FDG-PET imaging data from the ADNI project.
|
Authors who are presenting talks have a * after their name.
