Utility of Harmonisation for Fixel‐Based Metrics in Travelling Subjects and Alzheimer's Disease Data
Open Access
- 12 November 2025
- journal article
- Published by Wiley in Human Brain Mapping
- Vol. 46 (16) , e70408
Abstract
Fixel‐based analysis (FBA) is an advanced diffusion MRI analysis technique that facilitates the evaluation of white matter microstructure within ‘fixels’ (specific fibre populations within a voxel). In recent years, FBA has gained prominence for its ability to better characterise fibre tract‐specific changes than the more conventional diffusion MRI approaches and has shown promise in elucidating the pathophysiology of psychiatric and neurological diseases. However, FBA has been predominantly limited to single‐centre studies, minimising the generalisability of the technique. In this study, the popular ComBat harmonisation technique was adapted for whole‐brain FBA of diffusion MRI data. The study evaluates the effectiveness of ComBat in harmonising FBA metrics of fibre density, fibre cross‐section and the combined metric of fibre density and cross‐section in a large travelling subject dataset ( n = 49, scan = 162). Participants were scanned across multiple centres, using different scanner models and imaging protocols, and FBA metrics were compared under these varying conditions before and after harmonisation. In addition, the impact of ComBat harmonisation on disease‐related findings was evaluated in an independent multi‐centre Alzheimer's disease (AD) dataset, by comparing the same fixel‐based measures in patients with AD ( n = 27) to those in cognitively normal control participants ( n = 29) before and after ComBat harmonisation. We demonstrated that ComBat harmonisation effectively mitigated variability across scanner sites, scanner models, and protocols, in the travelling subject dataset, thus enhancing the comparability of FBA metrics. Notably, ComBat harmonisation improved the detection of AD‐related changes in the fornix, a critical white matter tract associated with cognitive function, and strengthened the correlations between FBA metrics and cognitive scores. These results underscore the potential of ComBat harmonisation in enhancing the reliability of multi‐centre neuroimaging studies, supporting the use of harmonisation techniques for accurate detection of disease‐specific changes in neurodegenerative conditions. The ability to perform ComBat harmonisation within the whole‐brain FBA pipeline may help further this fibre‐specific technique into large‐scale multi‐centre studies.Keywords
Funding Information
- Japan Society for the Promotion of Science (23H02865)
- Japan Society for the Promotion of Science (23K27556)
- Japan Society for the Promotion of Science (23K14927)
- Japan Agency for Medical Research and Development (JP18dm0307004)
- Japan Agency for Medical Research and Development (JP19dm0307101)
- Japan Agency for Medical Research and Development (JP21wm0425006)
- Japan Agency for Medical Research and Development (JP24wm0625310)
This publication has 57 references indexed in Scilit:
- Multicenter Collaboration in Observational Research: Improving Generalizability and EfficiencyJournal of Bone and Joint Surgery, 2009
- Bayesian analysis of neuroimaging data in FSLNeuroImage, 2009
- Robust determination of the fibre orientation distribution in diffusion MRI: Non-negativity constrained super-resolved spherical deconvolutionNeuroImage, 2007
- Probabilistic diffusion tractography with multiple fibre orientations: What can we gain?NeuroImage, 2007
- Adjusting batch effects in microarray expression data using empirical Bayes methodsBiostatistics, 2006
- The Montreal Cognitive Assessment, MoCA: A Brief Screening Tool For Mild Cognitive ImpairmentJournal of the American Geriatrics Society, 2005
- Advances in functional and structural MR image analysis and implementation as FSLNeuroImage, 2004
- Nonparametric permutation tests for functional neuroimaging: A primer with examplesHuman Brain Mapping, 2001
- Cortical Surface-Based AnalysisNeuroImage, 1999
- “Mini-mental state”Journal of Psychiatric Research, 1975