Validation of cross-sectional and longitudinal ComBat harmonization methods for magnetic resonance imaging data on a travelling subject cohort

Use of Support Vector Machines Approach via ComBat Harmonized Diffusion Tensor Imaging for the Diagnosis and Prognosis of Mild Traumatic Brain Injury: A CENTER-TBI Study

“…Model performance is related to the use of either single intensities (predictive voxels) or the average voxel intensity in selected regions (predictive ROIs), as well as to the choice of harmonization approach (no harmonization or Combat harmonized FA and MD maps with compared to the other methods (Table 4, p<0.05), which is not an unexpected result since the optimal harmonization of single voxels seems to highlight the group differences by preserving the underlying biological changes due to trauma. 69 Moreover, a similar trend is demonstrated in the prediction between mTBI patients with complete or incomplete recovery. The combination of age & sex along with the use of ComBat harmonized with 3 categories FA and MD in the voxel-wise approach provided a statistically significant improvement compared to other models (Table 6, p<0.05), providing evidence that the single-voxel ComBat harmonization successfully removes scanner effects in diffusion data preserving the mTBI effects in the brain.…”

“…The combination of age & sex along with the use of ComBat harmonized with 3 categories FA and MD in the voxel-wise approach provided a statistically significant improvement compared to other models (Table 6, p<0.05), providing evidence that the single-voxel ComBat harmonization successfully removes scanner effects in diffusion data preserving the mTBI effects in the brain. 69 The findings provide evidence for the use of DTI to aid identification of patients at risk of incomplete recovery after mTBI. The prediction of recovery remains an extremely challenging and complex question, which is influenced by multiple pre-injury factors.…”

Compositional structural brain signatures capture Alzheimer’s genetic risk on brain structure along the diseasecontinuum

“…This interpretation might not be straightforward, but the methodology overcomes biases and limitations of current IG multivariate approaches when working with heterogeneous MRI data from different cohorts. For instance, CoDA surpasses traditional methodologies in its ability to integrate volumetric data from various cohorts without the stringent requirement for harmonization 48 . This flexibility facilitates the analysis of diverse datasets and strengthens the reliability of cross-study comparisons.…”

Empirical assessment of the assumptions of ComBat with diffusion tensor imaging

“…20 Richter et al validated ComBat and variations on a travelling cohort in terms of ability to remove site bias without removing true biological effect from structural and diffusion MRI. 21 Orlhac et al examined deeper than only harmonization potential and assessed different situations and use cases for ComBat in the context of harmonizing image-derived biomarkers from PET scans. 22 Parekh et al posited sample size requirements under different Mahalanobis distances between datasets for structural MRI features, with larger distances corresponding to greater site biases.…”

“…Richter et al. validated ComBat and variations on a travelling cohort in terms of ability to remove site bias without removing true biological effect from structural and diffusion MRI 21 . Orlhac et al.…”