In resting-state and task fmri it is common to regress out BOLD time series of white matter as a nuisance covariate; the rationale being that whatever is causin

In resting-state and task fmri it is common to regress out BOLD time series of white matter as a nuisance covariate; the rationale being that whatever is causing BOLD fluctuation in white matter voxels it does not reflect neural activity within the voxel; nevertheless white matter does appear to exhibit BOLD responses that aren't merely physiological but arr related to task based cognition. I am willing to entertain the idea that white matter bold is important to study, but wanted a mechanism to explain why its relevant, and not mere epiphenomenon or spatially-temporally lagged response to neural activity elsewhere. Now I wonder whether astrocyte signaling is involved.