Simple signed-distance function depth calculation applied to measurement of the fMRI BOLD hemodynamic response function in human visual cortex

Abstract

Functional magnetic resonance imaging (fMRI) often relies on a hemodynamic response function (HRF) elicited by a brief stimulus. At conventional spatial resolutions (≥ 3 mm), signals in a voxel include contributions from various tissue types and pia vasculature. To better understand these contributions, full characterization of the depth dependence of the HRF is required in gray matter as well as and its apposed white-matter and pial vasculature. We introduce new methods to calculate 3D depth that combines a signed-distance function with an algebraic morphing definition of distance. The new scheme is much simpler than methods that rely upon deformable surface propagation. The method is demonstrated by combining the distance map with high-resolution fMRI (0.9-mm voxels) measurements of the depth-dependent HRF. The depth dependence of the HRF is reliable throughout a broad depth range in gray matter as well as in white-matter and extra-pial compartments apposed to active gray matter. The proposed scheme with high-resolution fMRI can be useful to separate HRFs in the gray matter from undesirable and confounding signals.