Integrating large-scale neuroimaging research datasets: Harmonisation of white matter hyperintensity measurements across Whitehall and UK Biobank datasets.

摘要: Large scale neuroimaging datasets present the possibility of providing normative distributions for a wide variety markers, which would vastly improve clinical utility these measures. However, major challenge is our current poor ability to integrate measures across different large-scale datasets, due inconsistencies in imaging and non-imaging protocols populations. Here we explore harmonisation white matter hyperintensity (WMH) two studies healthy elderly populations, Whitehall II sub-study UK Biobank. We identify pre-processing strategies that maximise consistency utilise multivariate regression characterise study sample differences contributing WMH variations studies. also parser harmonise WMH-relevant variables datasets. show can provide highly calibrated from with: (1) inclusion number specific standardised processing steps; (2) appropriate modelling through alignment demographic, cognitive physiological variables. These results open up range applications WMHs other markers extensive databases data.