Ultrahigh Resolution fMRI at 7T Using Radial-Cartesian TURBINE Sampling

Abstract

Purpose: We investigate the use of TURBINE, a 3D radial-Cartesian acquisition scheme in which EPI planes are rotated about the phase-encoding axis to acquire a cylindrical k-space for high-fidelity ultrahigh isotropic resolution fMRI at 7 Tesla with minimal distortion and blurring.

Methods: An improved, completely self-navigated version of the TURBINE sampling scheme was designed for fMRI at 7 Telsa. To demonstrate the image quality and spatial specificity of the acquisition, thin-slab visual and motor BOLD fMRI at 0.67 mm isotropic resolution (16 mm slab, TRvol = 2.32 s), and 0.8 × 0.8 × 2.0 mm (whole-brain, TRvol = 2.4 s) data were acquired. To prioritize the high spatial fidelity, we employed a temporally regularized reconstruction to improve sensitivity without any spatial bias.

Results: TURBINE images provide high structural fidelity with almost no distortion, dropout, or T₂ * blurring for the thin-slab acquisitions compared to conventional 3D EPI owing to the radial sampling in-plane and the short echo train used. This results in activation that can be localized to pre- and postcentral gyri in a motor task, for example, with excellent correspondence to brain structure measured by a T₁ -MPRAGE. The benefits of TURBINE (low distortion, dropout, blurring) are reduced for the whole-brain acquisition due to the longer EPI train. We demonstrate robust BOLD activation at 0.67 mm isotropic resolution (thin-slab) and also anisotropic 0.8 × 0.8 × 2.0 mm (whole-brain) acquisitions.

Conclusion: TURBINE is a promising acquisition approach for high-resolution, minimally distorted fMRI at 7 Tesla and could be particularly useful for fMRI in areas of high B₀ inhomogeneity.

Keywords: 7T; TURBINE; fMRI; high-resolution fMRI; radial-Cartesian; ultrahigh-field MRI.

FIGURE 1

(A) TURBINE pulse sequence diagram: The gradient trajectory is the same as for standard EPI, but the angle of the readout direction changes from shot to shot, shown for 1 example (yellow) blade. Fat suppression (not shown in diagram) was achieved either via water‐selective binomial excitation or with a fat suppression module preceding excitation. (B) TURBINE trajectory: The k_xy‐k_z EPI “blades” are rotated in a golden ratio angle scheme about the (phase encoding) k_z‐axis. For acceleration within EPI blades, alternating lines are skipped for alternating blades. (C) Illustration of the flexible spatiotemporal resolution property of golden angle sampling, allowing postacquisition choice of undersampling and temporal resolution

FIGURE 2

(A) TURBINE image reconstruction pipeline (B) impact of correction steps: example images and temporal SD images for no correction, Nyquist ghost correction alone, and all correction steps (C) zero order phase variation (k₀) and filtered versions thereof to illustrate physiological contributions to k₀

FIGURE 3

Structural MPRAGE image, TURBINE image (mean functional image), and MPRAGE contours overlaid on TURBINE image

FIGURE 4

Impact of temporal regularization: 0.67 mm isotropic slab data set with motor (A) and visual (B) fMRI for 3 different regularization factors: very low, medium, and high. Right column shows same data reconstructed with 2 mm in‐plane resolution

FIGURE 5

Maps of tSNR for all 6 slab fMRI data sets. Example slices (same slices as for fMRI results shown in Figure 7) for motor (top) and visual (bottom) slab data tSNR, temporal SNR.

FIGURE 6

Qualitative comparison of TURBINE with 2 comparable 3D EPI protocols. Example functional image with z‐statistic maps overlaid in transversal, coronal, and sagittal view. Insets highlight resolution differences in the cerebellum (green box) and artifacts in the temporal lobes (white box)

FIGURE 7

Activation maps for all high‐resolution (0.67 mm isotropic) slab data sets overlaid on mean functional TURBINE images. Top row shows motor (finger tapping) experiment for all 3 subjects; bottom shows visual (flashing checkerboard) experiment

FIGURE 8

Motor activation map and time series of 0.67 mm isotropic fMRI. Insets with zoom on the primary motor cortex/central sulcus; bottom inset without activation maps shows row of 6 voxels (green) for which time‐course are shown on the right‐hand side. Demeaned time courses for the different voxels (separated by 1 au on the y‐axis) show progression from positive to negative z‐statistics and back to positive as we move across the central sulcus

FIGURE 9

Layer‐specific analysis of 0.67 mm isotropic motor slab data. Top left shows mean functional TURBINE image, with 3 ROIs used for layer‐specific analysis. Hand area in left/right motor cortices (green/blue) and a control ROI (red) in gray matter area not expected to be activated by the task. The mean z‐stat values for all voxels in a layer are plotted. The shaded area corresponds to +/− the SD over the 4 slices used in the layer‐specific analysis.

FIGURE 10

Whole‐brain task data (0.8 × 0.8 × 2 mm resolution) on 1 subject performing a simultaneous motor/visual task. Z‐statistic maps are overlaid on the mean functional TURBINE image. Top row shows inferior slices with visual activation, second row superior slices with motor activation. The bottom two rows display sagittal and coronal views respectively