Adnan Shah1,2, Guoxiang Liu1,2, and Takashi Ueguchi1,2
1Brain Function Analysis and Imaging Lab, CiNet, NICT, Osaka, Japan, 2Graduate School of Frontier Biosciences, Osaka University, Osaka, Japan
Synopsis
We demonstrate the
reconstruction of isotropic submillimeter-resolution distortion- and resolution-matched (DRM) T1w-Like anatomy from the
image inversion of T1-maps obtained as five slice-shifted anatomical volumes
covering the whole brain. The proposed anatomy avoids the distortion-mismatch
between function and anatomy resulting in a high alignment with the functional data, necessary for
activity localization in submillimeter-resolution fMRI. We use the same pulse
sequence for acquiring both functional and DRM anatomical images with aligned
slice acquisitions except differing in parameters unrelated to
distortions. Moreover, the proposed anatomy allows the generation
of cortical surface for investigating
depth-dependent activity in isotropic submillimeter-resolution fMRI analysis.
Introduction:
Distinguishing
cortical layers and investigating depth-dependent BOLD [1] responses in isotropic
submillimeter-resolution fMRI [2] require the use of a high-contrast
distortion- and resolution-matched (DRM) anatomy. Based on our earlier work of
DRM anatomical MR image acquisitions for T1w-Like image reconstruction [3,4],
we present here the results of isotropic submillimeter-resolution fMRI on a
T1w-Like whole-brain anatomical image reconstructed from the image inversion of
DRM T1-maps. These T1-maps are computed from slice varying multi-shot inversion
recovery-prepared echo-planar imaging (msIR-EPI) volumes [5, 6]. The same pulse sequence was used for acquiring both
functional and DRM anatomical images with aligned slice acquisitions except
differing in parameters unrelated to distortions. The reconstructed
T1w-Like anatomy in the native EPI space offers high
alignment necessary for activity localization providing benefits for studying
voxel-wise dynamics in isotropic submillimeter-resolution fMRI. Moreover, the
proposed T1w-Like anatomy allows the generation of cortical surface for investigating
depth-dependent responses [7].Materials & Methods:
An adult human brain was scanned using the BISEPI sequence [2] on a
Siemens MAGNETOM 7T scanner with a 32-channel phased-array head coil to obtain 144
fMRI volumes at a spatial resolution of voxel-wise 0.7 mm isotropic. The functional
scan (#Segments = 3, GRAPPA = 2, #Shots in block = 24, TR/TE/FA = 1000 ms /22.2
ms / 50°, #Slices
= 15, PF = 6/8) covering the region of interest in primary motor cortex was
acquired in a bilateral finger tapping block-design task 12s (ON) / 12s (OFF).
Functional scan was followed by five slice-shifted anatomical volumes at a
spatial resolution of voxel-wise 0.7 mm isotropic using 2D msIR-EPI (#Segments
= 6, TR/TE/FA = 3000 ms /22.2 ms/ 90°, Slices/Averages = 35x5/2, Slice thickness = 0.7 mm,
Spacing = 3.5 mm, PF = 6/8) with variable inversion recovery for each slice
generated by an IR-sweep with varying inversion time (TI) across slices (TI=100-2400ms)
[6]. The obtained anatomical volumes were used to generate T1-maps after
T1-fitting and combined and processed in Matlab to generate whole-brain
T1w-Like DRM anatomy based on image inversion [4]. The obtained native EPI
space T1w-Like anatomy was post-processed in BrainVoyager for generating
cortical surface after AC-PC transformation. Mid-GM surface was generated and
the functional data at four distinct depths moving from GM-WM interface to
GM-pial boundary were sampled and overlaid on this mid-GM surface. fMRI
pre-statistics processing includes mean intensity correction, motion
correction, and high-pass filtering with GLM-Fourier 2 sines/cosines and Gaussian
FWHM of 2 data points, with no spatial smoothing. The generalized linear model
analysis was applied with stimulation ON/OFF as a binary regression variable in
BrainVoyager 21.2. Results and Discussion:
Figure 1 shows fMRI BOLD activity overlaid on the obtained T1w-Like DRM
anatomy based on image inversion [4]. As shown here, the fMRI BOLD activity at 0.7
mm isotropic resolution follows the ribbon of the grey matter structures on the
reconstructed T1w-Like DRM whole-brain anatomy. Figure 2 shows the mid-GM cortical surface generated from this anatomy where
fMRI BOLD activity is sampled at distinct cortical depths moving from the GM-WM
interface (left-most) to the GM-pial boundary (right-most). Figure 3 shows overlaid fMRI BOLD activity as the
multiple-selection of all depth maps on the mid-GM cortical surface.Conclusion:
DRM T1w-Like anatomy in the native EPI space [3, 4] alleviates the
problem of distortion mismatch that exists between function and T1w anatomy. Moreover,
it improves the accuracy of activity localization and interpretation of the results
in submillimeter-resolution fMRI studies. In this study, we reconstructed
T1w-Like whole-brain DRM anatomical slices from the image inversion of T1-maps
computed from slice varying 2D msIR-EPI [5, 6] acquired at 0.7 mm isotropic
resolution. The reconstructed T1w-Like DRM anatomy demonstrates a strong
alignment with the acquired functional data, necessary for studying voxel-wise dynamics in submillimeter-resolution
fMRI. The proposed T1w-Like whole-brain
anatomy allows the generation of a cortical surface, useful for investigating
layer-specific functional responses in a region of interest. Acknowledgements
This
study was supported in part by Japan Society for the
Promotion of Science (JSPS) Grants-in-Aid for Scientific Research “KAKENHI” (Grant
Numbers JP26282223 and JP26350471 and JP19K08244).References
[1]
Ogawa S, Lee TM, Kay AR, Tank DW. Brain magnetic resonance imaging with
contrast dependent on blood
oxygenation.
Proc Natl Acad Sci USA. 1990;87(24):9868-9872.
[2] Liu G, Shah A, Ueguchi T. Block-Interleaved segmented EPI for voxel-wise
high-resolution fMRI studies at 7T. Proceedings of the International Society
for Magnetic Resonance in Medicine Joint Annual Meeting ISMRM-ESMRMB. 2018;5450.
[3] Shah A, Ueguchi T, Liu G. Distortion-Matched Anatomical Imaging using
Inversion Recovery-Prepared EPI for high-resolution fMRI. 24th Annual Meeting
of the Organization for Human Brain Mapping. 2018;1727. [4] Shah A, Liu G,
Ueguchi T. EPI based distortion- and resolution-matched T1-Like anatomy for
submillimeter-resolution fMRI at 7T. 25th Annual Meeting of the Organization
for Human Brain Mapping. 2019;2591. [5] Turner R, Panchuelo RS, Mougin O,
Francis S. Multi-shot inversion recovery EPI with SMS excitation for high
spatial resolution T1-mapping. 25th Annual Meeting of the Organization for
Human Brain Mapping. 2019;2627. [6] Panchuelo RS, Turner R, Mougin O, Francis
S. A 2D multi-shot inversion recovery EPI (MS-IR-EPI) sequence for high spatial
resolution T1-mapping at 7T. Proceedings of the International Society for
Magnetic Resonance in Medicine. 2018;60. [7] Shah A, Liu G, Ueguchi T.
Investigating Depth-Dependent Activity in Submillimeter-Resolution fMRI at 7T
with Native- Space Transformed T1w MPRAGE Anatomy. ICMRI. 2019;SS03-08.