The objective of this study was to investigate effects of varying readout bandwidths on the arterial spin labeling (ASL)-perfusion MRI measurements at a high magnetic field MRI system. Brain perfusion studies were performed on nine volunteers (four males, five females) using flow sensitive alternating inversion recovery (FAIR) ASL single-shot echo-planar imaging (EPI)-MRI. To investigate EPI bandwidth effects on the time-series perfusion-weighted imaging (PWI) data, two regions-of-interest (ROI) were placed outside the brain to determine the level of noise and another ROI inside the brain to determine the level of signal. Coefficients of variations (CoV) were calculated for the time-series PWI data. One-way analysis of variance (ANOVA) was used to investigate voxel-wise differences in the time-series PWI data between two different bandwidth values. At the level of ROI, there was no significant effect of changing EPI bandwidths on the time-series PWI data in any of the volunteers (P > 0.031). In contrast, CoV values over the dynamic PWI data varied with depending on selecting EPI bandwidths and voxel-based tests showed that N2 ghosting, modulated by EPI bandwidth, can appear in some brain regions, especially in areas that overlap with the spatial distribution of N2 ghosting artifacts. Although N2 ghosting can be reduced by adjusting the bandwidth of EPI on the time-series of PWI data, the effects cannot be entirely eliminated. In particular, N2 ghosting can bias CBF quantification if EPI control scans to determine the equilibrium-state signal are confounded by N2 ghosting. Therefore, careful tuning of the bandwidth of EPI is necessary to avoid artifacts in the ASL signal from N2-ghosting.

Introduction

The objective of the study was to explore the impact of the background gradients on diffusion tensor (DT) magnetic resonance imaging (DT-MRI) in patients with Alzheimer's disease (AD), mild cognitive impairment (MCI), or cognitively normal (CN) aging.

Methods

Two DT-MRI sets with positive and negative polarities of the diffusion-sensitizing gradients were obtained in 15 AD patients, 18 MCI patients, and 16 CN control subjects. The maps of mean diffusivity (MD) and fractional anisotropy (FA) were computed separately for positive (p: pMD and pFA) and negative (n: nMD and nFA) polarities, and we computed the geometric mean (gm) of the DT-MRI to obtain the gmFA and gmMD with reducing the background gradient effects. Regional variations were assessed across the groups using one-way analysis of variance.

Results

Increased regional gmMD values in the AD subjects, as compared to the regional gmMD values in the MCI and CN subjects, were found primarily in the frontal, limbic, and temporal lobe regions. We also found increased nMD and pMD values in the AD subjects compared to those values in the MCI and CN subjects, including in the temporal lobe and the left limbic parahippocampal gyrus white matter. Results of comparisons among the three methods showed that the left limbic parahippocampal gyrus and right temporal gyrus were the increased MD in the AD patients for all three methods.

Conclusion

Background gradients affect the DT-MRI measurements in AD patients. Geometric average diffusion measures can be useful to minimize the intrinsic local magnetic susceptibility variations in brain tissue.