A shimming procedure for fMRI, optimizing the local BOLD sensitivity

Introduction In functional magnetic resonance imaging (fMRI), magnetic field inhomogeneities due to air-tissue susceptibility differences are well-known to lead to severe signal dropouts and geometric distortions in echo-planar images (EPI). Therefore, the optimization of the field homogeneity is an important step in the imaging process and various so-called “shimming” techniques, using linear and higher-order resistive shim coils, have been developed. The common overall goal of the many existing approaches is to calculate the corrective shim currents in order to compensate for the field inhomogeneities over a region of interest (ROI), by minimizing the spatial standard deviation of the magnetic field. However in fMRI, the BOLD (Blood Oxygen Level Dependent) sensitivity is the measure of interest, and it is only indirectly related to the spatial variation of the magnetic field. In particular, it depends on the EPI signal intensity and the local TE [1-3]. The analytical expression for an estimate of the BOLD sensitivity has been previously developed, allowing for the computation of BOLD sensitivity maps from EPI data and field maps [1-3]. In this study, a procedure has been developed that optimizes the BOLD sensitivity over a region of interest, while ensuring satisfying overall field homogeneity in order to avoid geometric distortions in the echo-planar images. The method is applied in vivo and compared to two other methods based on the optimization of the field homogeneity.