loading
loading
NIH grant 1R01AG097729-01A1: Statistical Approaches for Data Harmonization for PET Imaging Studies of Alzheimer's Disease
Institute: {'code': 'AG', 'abbreviation': 'NIA', 'name': 'National Institute on Aging', 'admin_org_id': '12500002', 'admin_funding_url': ''}. PI: Davneet Singh Minhas. Award: $1418192.
Start: 2026-09-30T00:00:00.
PROJECT ABSTRACT
This project focuses on the development of statistical methods for data harmonization of amyloid and tau PET
neuroimaging studies of Alzheimer’s disease. Specifically, we will focus on developing methods for amyloid and
tau PET harmonization for multiple sites and multiple tracers, for cross-sectional and longitudinal data. In Aim 1
we will collect data on a 3D-printed anatomically accurate multi-region PET phantom that will be travelling to
multiple sites, including sites participating in the Head-to-Head Harmonization of Tau Tracers in Alzheimer’s
Disease (HEAD), ADNI, ABC-DS, and BLSA studies of Alzheimer’s disease. In Aim 2 we will extend the
commonly used Surrogate Variable Analysis (SVA) Removing Unwanted Variation (RUV) and RAVEL methods
for cross-sectional and longitudinal analyses with the goal of separating the non-biological and scanner effects
for PET neuroimaging studies, as well as individual characterization of unwanted biological effects characteristic
of different PET tracers. The first part of the aim will also involve extending existing image processing and
machine learning techniques developed by our group to accommodate PET neuroimaging data. These methods
will be applied to both AD and cognitively unimpaired human participant data acquired from the HEAD, ADNI,
ABC-DS and BLSA studies. In Aim 3 we will individually analyze multi-tracer PET data from publicly available
and internal studies and combine them using meta-analyses. In Aim 4 we will develop online, freely accessible,
reproducible software tools for the assessment of validation and reproducibility of published analytic pipelines.document fetched 2026-10-15 02:48