Automated slice-specific z-shimming for fMRI of the human spinal cord

Functional magnetic resonance imaging (fMRI) of the human spinal cord faces many challenges, one of which is signal loss due to local magnetic field inhomogeneities. This issue can be addressed with slice-specific z-shimming, which compensates for the dephasing effect of the inhomogeneities using a single, slice-specific gradient pulse. Since the original demonstration of its utility, this technique has already been employed in several spinal fMRI studies. Here, we aim to address two outstanding issues regarding this technique: on the one hand, we evaluate its effects on several parameters that are directly relevant for spinal fMRI (but have not yet been assessed) and on the other hand, we improve upon the manual selection of slice-specific z-shims by developing automated procedures. First, we demonstrate that the beneficial effects of z-shimming i) are apparent across a large range of echo times, ii) hold true for both the dorsal and ventral horn gray matter, and iii) are also clearly apparent in the temporal signal-to-noise ratio (tSNR) of gradient-echo EPI time-series data. Second, and more importantly, we address the time-consuming and subjective nature of manual selection of slice-specific z-shims by developing two automated approaches: one is based on finding the z-shim that maximizes spinal cord signal intensity in each slice of an EPI z-shim reference-scan and the other is based on finding the strength of the gradient-field that compensates the through-slice inhomogeneity in field map data. Both automated approaches i) were much faster than the manual approach, ii) lead to significant improvements in spinal cord gray matter tSNR compared to no z-shimming and iii) resulted in beneficial effects that were stable across time. While the field map-based approach performed slightly worse than the manual approach, the EPI-based approach performed at least as well as the manual one and was furthermore validated on an independently acquired corticospinal data-set (N > 100). Together, we believe that automated z-shimming will improve the data quality of future spinal fMRI studies and -- by removing the subjective step of manual z-shim selection -- may also lead to increased reproducibility in longitudinal studies.