Related Experiment Video
Updated: Jun 30, 2025
08:36
Development of an Uncomplicated Mild Traumatic Brain Injury Model Modified by Weight-Drop Method and Evidenced by Magnetic Resonance Imaging
Published on: April 11, 2025
255
ComBating inter-site differences in field strength: harmonizing preclinical traumatic brain injury MRI data
Mara Quach1,2,3, Idrish Ali1, Sandy R Shultz1,4,5
1Department of Neuroscience, School of Translational Medicine, Monash University, Melbourne, Victoria, Australia.
NMR in Biomedicine
|March 18, 2024
Summary
Harmonizing multi-site preclinical MRI data using ComBat effectively reduces inter-site variability. This method preserves key imaging biomarkers of traumatic brain injury (TBI) in rats, enabling more powerful neuroimaging studies.
Area of Science:
- Neuroimaging
- Preclinical Research
- Data Harmonization
Background:
- Multi-site neuroimaging studies offer increased statistical power but face challenges with inter-site confounds.
- Standardizing data acquisition and analysis across different locations and scanner types is crucial for reliable results.
- Traumatic brain injury (TBI) research requires robust methodologies to detect subtle neurobiological changes.
Purpose of the Study:
- To evaluate the effectiveness of ComBat, an empirical Bayes approach, for harmonizing longitudinal preclinical MRI data.
- To assess ComBat's ability to mitigate inter-site confounds in MRI data from two different Australian sites.
- To determine if ComBat harmonization preserves TBI-induced phenotypical changes in rat brain imaging metrics.
Main Methods:
- Longitudinal MRI data (diffusion-weighted and multi-gradient-echo) were acquired from Sprague Dawley rats at 4.7T and 9.4T across two sites.
- ComBat harmonization was applied to quantitative susceptibility mapping (QSM), fractional anisotropy (FA), and apparent diffusion coefficient (ADC) metrics.
- Mixed-effects modeling and Bland-Altman plots were used to assess inter-site variability and the impact of ComBat.
Main Results:
- ComBat successfully removed significant site effects on FA and ADC values in the corpus callosum (CC).
- Bland-Altman analysis showed reduced variability across sites for QSM, FA, and ADC after ComBat harmonization.
- Harmonized data revealed significant TBI-related changes, including reduced FA and increased susceptibility in the ipsilateral CC, and altered ADC over time.
Conclusions:
- ComBat is an effective method for harmonizing multi-site preclinical MRI data, reducing inter-site variability.
- Data harmonization using ComBat successfully preserves crucial phenotypical changes associated with TBI.
- This approach enhances the reliability and statistical power of multi-site preclinical neuroimaging studies.
