High-field FMRI for human applications: an overview of spatial resolution and signal specificity

Abstract

In the last decade, dozens of 7 Tesla scanners have been purchased or installed around the world, while 3 Tesla systems have become a standard. This increased interest in higher field strengths is driven by a demonstrated advantage of high fields for available signal-to-noise ratio (SNR) in the magnetic resonance signal. Functional imaging studies have additional advantages of increases in both the contrast and the spatial specificity of the susceptibility based BOLD signal. One use of this resultant increase in the contrast to noise ratio (CNR) for functional MRI studies at high field is increased image resolution. However, there are many factors to consider in predicting exactly what kind of resolution gains might be made at high fields, and what the opportunity costs might be. The first part of this article discusses both hardware and image quality considerations for higher resolution functional imaging. The second part draws distinctions between image resolution, spatial specificity, and functional specificity of the fMRI signals that can be acquired at high fields, suggesting practical limitations for attainable resolutions of fMRI experiments at a given field, given the current state of the art in imaging techniques. Finally, practical resolution limitations and pulse sequence options for studies in human subjects are considered.

Keywords: 3T; 7T; BOLD; Functional imaging; High field; High resolution; Human; Spatial specificity; fMRI..

Fig. (1)

Examples of high-resolution single-shot EPI acquired at 7T over the entire brain with different isotropic resolutions. 2.0 mm: matrix: 96 x 96, 50 slices, TR = 2 s, R=2, readout time: 20 msec. 1.5 mm: matrix: 128 x 128, 60 slices, TR = 3 sec, R = 3, readout time: 21 msec. 1.0 mm: matrix: 192 x 192, 100 slices, TR = 3-4 sec, R=4, readout time: 28 msec. 0.75 mm: matrix: 256 x 256, 128 slices, TR = 4- 5 sec, R=4, readout time: 48 msec.

Fig. (2)

Example (single-shot) EPI images acquired with body gradients and head gradients at 1mm isotropic resolutions at 7T using a 16 channel head coil. Head gradient image parameters: R=4, matrix: 192 x 192, partial Fourier= 6/8, echo spacing=0.58 msec, TE=22 msec, readout time = 18.5 msec (32/48 lines). Body gradient parameters: R=4, matrix: 192 x 192, partial Fourier= 6/8, echo spacing=0.96 msec, TE=25 msec, readout time = 30.7 msec (32/48 lines).

Fig. (3)

ODC maps from a flat gray matter region along the calcarine sulcus from the same subject on several different days. On the left is the overlap of 2 different scan days using SE at 7T, while on the right is the overlap of 2 different scans using GE. Red or blue pixels indicate a voxel’s preference to right or left eye stimulation. The expected spatial pattern is an alternating preference (1 mm width, 2 mm cycle) to right or left eye running along the sulcus. The SE map retains this pattern throughout the region, while the GE map is interrupted by extravascular changes around large vessels somewhere near the region of interest. (taken from Fig. (5) in Yacoub et al 2007 [56] ).

Fig. (4)

Top panel: a 3D GRASE coronal slab with 0.7mm (isotropic) resolution and 8 slices covers 0.56 cm in the anterior/posterior direction and shows robust activation in response to the presentation of partially occluded objects. Bottom panel: functional responses are visualized as a color overlay on an axial view of the subject’s T₁-weighted anatomy (also with 0.7mm resolution). Example stimulus shown at right; green arrow indicates correspondence between a specific cluster of activated voxels and the corresponding image region.

Fig. (5)

Reduced FOV fMRI. Single Shot SE-EPI with 0.5 mm in plane resolution at 7T (top). Visual activation in humans using a 30 sec on/30 sec off paradigm using a single 8 minute scan and the resulting fMRI map (bottom). Data acquired using inner-volume excitation and a surface coil positioned on the visual cortex. Imaging parameters: matrix: 50 x 256, FOV: 2.5 x 12.8 cm², resolution: 0.5 x 0.5 x 2 mm³, 35 msec readout TR/TE: 2s/ 50 msec.