The abstract is read at the publisher; the summary is JClub's.
In ADNI baseline [18F]flortaucipir tau PET, integrating multiple imputation with harmonization preserved cognitively impaired vs unimpaired group differences and improved precision (median SE ratios 0.89/0.93).
- Why it matters: Multisite harmonization for Alzheimer’s disease tau PET has limited evidence on how missing data handling interacts with site-effect reduction and downstream biological inference.
- What they did: Using ADNI-preprocessed baseline [18F]flortaucipir tau PET, the study compared complete-case vs MICE-imputed datasets (m = 5) with ComBat (with/without covariates), quantile normalization, and quantile regression, then assessed site variability and biological effect preservation across six Braak regions.
- The result: Post-processing harmonization reduced site-level variability, and overall harmonization quality stayed comparable after MI; disease group differences remained statistically significant with a consistent regional pattern (strongest Braak I, weakest Braak VI), while MI modestly increased effect and coefficient estimates with improved precision.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.