Sub-millimeter T2 weighted fMRI at 7 T: comparison of 3D-GRASE and 2D SE-EPI

Abstract

Functional magnetic resonance imaging (fMRI) allows studying human brain function non-invasively up to the spatial resolution of cortical columns and layers. Most fMRI acquisitions rely on the blood oxygenation level dependent (BOLD) contrast employing T*2 weighted 2D multi-slice echo-planar imaging (EPI). At ultra-high magnetic field (i.e., 7 T and above), it has been shown experimentally and by simulation, that T2 weighted acquisitions yield a signal that is spatially more specific to the site of neuronal activity at the cost of functional sensitivity. This study compared two T2 weighted imaging sequences, inner-volume 3D Gradient-and-Spin-Echo (3D-GRASE) and 2D Spin-Echo EPI (SE-EPI), with evaluation of their imaging point-spread function (PSF), functional specificity, and functional sensitivity at sub-millimeter resolution. Simulations and measurements of the imaging PSF revealed that the strongest anisotropic blurring in 3D-GRASE (along the second phase-encoding direction) was about 60% higher than the strongest anisotropic blurring in 2D SE-EPI (along the phase-encoding direction). In a visual paradigm, the BOLD sensitivity of 3D-GRASE was found to be superior due to its higher temporal signal-to-noise ratio (tSNR). High resolution cortical depth profiles suggested that the contrast mechanisms are similar between the two sequences, however, 2D SE-EPI had a higher surface bias owing to the higher T*2 contribution of the longer in-plane EPI echo-train for full field of view compared to the reduced field of view of zoomed 3D-GRASE.

Cortical depth profiles of luminance contrast responses in human V1 and V2 using 7 T fMRI

“…T2‐weighted sequences, such as GRASE (De Martino et al, ; Kemper, De Martino, Yacoub, & Goebel, ; Kemper et al, ), and sequences that are not based on the BOLD contrast, such as VASO (Huber et al, ; Huber, Uludağ, & Möller, ) and ASL (Huber et al, ; Pfeuffer et al, ; Uludağ et al, in preparation), are less affected by vascular biases and are therefore expected to yield a better estimation of local neuronal activity compared with GE sequences. However, this advantage comes at the price of a lower sensitivity and/or decreased coverage.…”

“…T2-weighted sequences, such as GRASE (De Martino et al, 2013;Kemper, De Martino, Yacoub, & Goebel, 2015;Kemper et al, 2016), and sequences that are not based on the BOLD contrast, such as VASO (Huber et al, 2014(Huber et al, , 2015Huber, Uluda g, & M€ oller, 2017b) and…”

Variable flip angle 3D‐GRASE for high resolution fMRI at 7 tesla

“…Compared with conventional refocusing schemes, in vivo VFA measurements exhibit significantly reduced blurring. With the VFA approach we were able to obtain a similar PSF to that of large FOV 2D SE‐EPI along the phase‐encoding direction . In addition, the coverage of VFA2 was almost doubled compared with REF.…”

“…To characterize the PSF, we computed the full width at half maximum (FWHM) by interpolating the magnitude of the PSF using a spline interpolation. This approach was chosen rather than a Gaussian fit because it characterized the different peak shapes of the VFA acquisitions better.…”

“…The refocusing RF pulses select a perpendicular slab achieving inner‐volume selection. We have recently shown that 3D‐GRASE is functionally more specific and more sensitive than 2D SE‐EPI with a large field of view (FOV) , however, blurring along the secondary phase‐encoding direction (partition direction) and the small FOV of 3D‐GRASE still limit its usability. In particular, the FOV of 3D‐GRASE cannot easily be extended as doing so introduces T 2 decay induced blurring (additionally blurring and functional contrast when increasing the in‐plane FOV).…”