rus
Issue #6/2021
O.P.Posnansky
Shell-based analysis of magnetic resonance diffusion tensor imaging. Part I
Shell-based analysis of magnetic resonance diffusion tensor imaging. Part I
DOI: 10.22184/1993-8578.2021.14.6.328.333
Even with substantial sensitivity at high magnetic field, it is still challenging to investigate small-scale structures, such as brain cortex. A recent study has shown that very high resolution diffusion imaging is possible allowing investigation the cortical depth dependence of diffusion properties over the whole human brain in vivo. The results revealed that the main diffusion tensor orientation in the cortex is perpendicular to the cortical surface. At the same time the main diffusion direction is mostly tangential at banks and radial at the tips of the white matter – grey matter interface. Thus quantitative diffusion tensor measures can reveal structural organization of cells in the human cerebral cortex with the potential to characterize cortical cytoarchitecture in vivo and to investigate the pathophysiology of diseases associated with changes in cortical grey matter.
Even with substantial sensitivity at high magnetic field, it is still challenging to investigate small-scale structures, such as brain cortex. A recent study has shown that very high resolution diffusion imaging is possible allowing investigation the cortical depth dependence of diffusion properties over the whole human brain in vivo. The results revealed that the main diffusion tensor orientation in the cortex is perpendicular to the cortical surface. At the same time the main diffusion direction is mostly tangential at banks and radial at the tips of the white matter – grey matter interface. Thus quantitative diffusion tensor measures can reveal structural organization of cells in the human cerebral cortex with the potential to characterize cortical cytoarchitecture in vivo and to investigate the pathophysiology of diseases associated with changes in cortical grey matter.
Теги: cortical cytoarchitecture diffusion tensor grey matter human cerebral cortex magnetic resonance diffusion tensor imaging pathophysiology shell-based analysis кортикальная клеточная архитектура магнито-резонансные диффузионные тензорные изображения метод оболочек патофизиология тензор диффузии
SHELL-BASED ANALYSIS OF MAGNETIC RESONANCE DIFFUSION TENSOR IMAGING. PART I
Even with substantial sensitivity at high magnetic field, it is still challenging to investigate small-scale structures, such as brain cortex. A recent study has shown that very high resolution diffusion imaging is possible allowing investigation the cortical depth dependence of diffusion properties over the whole human brain in vivo. The results revealed that the main diffusion tensor orientation in the cortex is perpendicular to the cortical surface. At the same time the main diffusion direction is mostly tangential at banks and radial at the tips of the white matter – grey matter interface. Thus quantitative diffusion tensor measures can reveal structural organization of cells in the human cerebral cortex with the potential to characterize cortical cytoarchitecture in vivo and to investigate the pathophysiology of diseases associated with changes in cortical grey matter.
INTRODUCTION
Magnetic resonance diffusion tensor imaging (DTI) has proven its strength detecting the microscopic fibrous structures such as white matter (WM). Very high magnetic field (≥3Tesla) boosted it to even higher spatial resolution, consequently providing a powerful tool to delineate the cortical gray matter (GM) with reduced partial volume effect in acceptable scan time. The present study of GM uses diffusion-weighted (DW) echo-planar imaging (EPI) sequence. Our results revealed that the main diffusion tensor orientation in the cortex is perpendicular to the cortical surface. At the same time the main diffusion direction is mostly tangential at banks and radial at the tips of WM-GM interface. Quantitative diffusion tensor measures can disentangle structural organization of cells in the human cerebral cortex with the potential to characterize cortical cyto-architecture in vivo and to investigate the pathophysiology of diseases associated with changes in cortical GM.
DATA GENERATION
In vivo human MRI data were downloaded from open database db.humanconnectome.org. Images were acquired on a 3T Siemens scanner (Siemens, Erlangen, Germany) with a 32-channel RX head coil. Twelve slabs of brain DWIs were acquired in 12 noncollinear and noncoplanar directions (b = 1000 s/mm2) and interleaved three nonDWI (b = 0 s/mm2) using the single-refocused Stejskal-Tanner spin-echo sequence (FoV = 210 × 180 × 139 mm3, voxel size = 1.25 × 1.25 × 1.25 mm3, in plane resolution 168 × 144, GRAPPA2, TE/TR = 89.5/5520 ms, partial Fourier 6/8, bandwidth 1488 Hz/px, 20 slices (covering 2.5 cm)). For excitation of 3 slices simultaneously multi-band radio frequency (MB-RF) pulse was used. Additionally, a set of reversed phase-encoded gradient b = 0 s/mm2 images were acquired for correction of geometrical distortions. Diffusion gradient was characterized with diffusion time Δ = 22 ms, diffusion gradient duration δ = 6 ms. No cardiac gating was used and the total scan time for all measurements was about 55 minutes including MP-RAGE and 3D-GE acquisitions for anatomical imaging. This experiment was repeated four times for axially shifted by 2.5 cm slabs to cover the entire brain volume.
IMAGE PROCESSING
The whole procedure of image processing is given in Fig.1 [1, 2]. After radio-frequency (RF) intensity inhomogeneity correction of anatomical images (MP-RAGE was divided by aligned 3D-GE), brain segmentation of WM, GM and CSF (cerebrospinal fluid) was performed [3]. To generate a whole brain volume, four different volumes from separated scans covering different regions of the brain were co-registered to the MP-RAGE image. Series of shells (lamellae) from the WM-GM to the GM-CSF interface were created by the surface expansion method (Fig.2) [3].
Lamelle № 1 and 7 were created on the edges of WM and CSF correspondently; Lamelle № 2 and 6 were determined by WM-GM and GM-CSF interfaces, and Lamellae № 3–5 were built in cortical depth. The normalized absolute scalar product (AbsScalarProd) of the major DTI eigenvectors and normal vectors perpendicular to the lamellae was calculated [4]. AbsScalarProd was mapped onto the inflated cortex followed by statistical analysis separately in GM banks, and GM gyri and sulci. GM banks were identified by a WM-GM surface curvature within the interval [–0.15, 0.15] 1/mm2, and gyri and sulci were characterized by curvature values outside this interval (Fig.3). The cortical depth dependence of diffusion properties including AbsScalarProd were studied using histogram analysis.
RESULTS
Fig.4 displays the major diffusion eigenvector (e1) fields in a gyrus of primary somatosensory (a) and motor cortex (b). The major eigenvectors at the WM-GM interface in banks (blue arrow 1) are mostly parallel to the interface and orthogonal at the tip of the gyrus (blue arrow 2). The major diffusion direction changes in the medial lamella, where the major eigenvectors are orthogonal to the cortex in banks (yellow arrow 1) and gyri (yellow arrow 2).
This observation is confirmed quantitatively in histograms of AbsScalarProd of vector fields calculated for the whole brain cortex at different depth. At the WM-GM interface Fig.5a demonstrates a strong peak of AbsScalarProd at 0 (major eigenvector orthogonal to surface normal) for banks while Fig.5b shows concentration of values at 1 for sulci and gyri (major eigenvector parallel to the surface normal). Histograms of the AbsScalarProd at the medial cortex lamella at banks (Fig.5c) and sulci/gyri (Fig.5d) evidence the radial character of the major eigenvector at this cortical depth throughout the brain.
DISCUSSION AND CONCLUSION
We had shown that major DTI eigenvectors at high magnetic fields has the ability detecting a sharp fiber turning in the human cortex in vivo [5, 6]. The analysis of DTI is facilitated by the AbsScalarProd parameter which reflects the curvature of the cortex and shape of lamellae in the cortical depth. Thus DTI with isotropic resolution robustly probes structural differences along the cortex and in the cortical depth. Potentially AbsScalarProd as a metric to measure the fiber bending in different parts of the WM-GM interface, may serve as an index for detecting diseases and brain malformation. ▪
Declaration of Competing Interest. The author declares that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Even with substantial sensitivity at high magnetic field, it is still challenging to investigate small-scale structures, such as brain cortex. A recent study has shown that very high resolution diffusion imaging is possible allowing investigation the cortical depth dependence of diffusion properties over the whole human brain in vivo. The results revealed that the main diffusion tensor orientation in the cortex is perpendicular to the cortical surface. At the same time the main diffusion direction is mostly tangential at banks and radial at the tips of the white matter – grey matter interface. Thus quantitative diffusion tensor measures can reveal structural organization of cells in the human cerebral cortex with the potential to characterize cortical cytoarchitecture in vivo and to investigate the pathophysiology of diseases associated with changes in cortical grey matter.
INTRODUCTION
Magnetic resonance diffusion tensor imaging (DTI) has proven its strength detecting the microscopic fibrous structures such as white matter (WM). Very high magnetic field (≥3Tesla) boosted it to even higher spatial resolution, consequently providing a powerful tool to delineate the cortical gray matter (GM) with reduced partial volume effect in acceptable scan time. The present study of GM uses diffusion-weighted (DW) echo-planar imaging (EPI) sequence. Our results revealed that the main diffusion tensor orientation in the cortex is perpendicular to the cortical surface. At the same time the main diffusion direction is mostly tangential at banks and radial at the tips of WM-GM interface. Quantitative diffusion tensor measures can disentangle structural organization of cells in the human cerebral cortex with the potential to characterize cortical cyto-architecture in vivo and to investigate the pathophysiology of diseases associated with changes in cortical GM.
DATA GENERATION
In vivo human MRI data were downloaded from open database db.humanconnectome.org. Images were acquired on a 3T Siemens scanner (Siemens, Erlangen, Germany) with a 32-channel RX head coil. Twelve slabs of brain DWIs were acquired in 12 noncollinear and noncoplanar directions (b = 1000 s/mm2) and interleaved three nonDWI (b = 0 s/mm2) using the single-refocused Stejskal-Tanner spin-echo sequence (FoV = 210 × 180 × 139 mm3, voxel size = 1.25 × 1.25 × 1.25 mm3, in plane resolution 168 × 144, GRAPPA2, TE/TR = 89.5/5520 ms, partial Fourier 6/8, bandwidth 1488 Hz/px, 20 slices (covering 2.5 cm)). For excitation of 3 slices simultaneously multi-band radio frequency (MB-RF) pulse was used. Additionally, a set of reversed phase-encoded gradient b = 0 s/mm2 images were acquired for correction of geometrical distortions. Diffusion gradient was characterized with diffusion time Δ = 22 ms, diffusion gradient duration δ = 6 ms. No cardiac gating was used and the total scan time for all measurements was about 55 minutes including MP-RAGE and 3D-GE acquisitions for anatomical imaging. This experiment was repeated four times for axially shifted by 2.5 cm slabs to cover the entire brain volume.
IMAGE PROCESSING
The whole procedure of image processing is given in Fig.1 [1, 2]. After radio-frequency (RF) intensity inhomogeneity correction of anatomical images (MP-RAGE was divided by aligned 3D-GE), brain segmentation of WM, GM and CSF (cerebrospinal fluid) was performed [3]. To generate a whole brain volume, four different volumes from separated scans covering different regions of the brain were co-registered to the MP-RAGE image. Series of shells (lamellae) from the WM-GM to the GM-CSF interface were created by the surface expansion method (Fig.2) [3].
Lamelle № 1 and 7 were created on the edges of WM and CSF correspondently; Lamelle № 2 and 6 were determined by WM-GM and GM-CSF interfaces, and Lamellae № 3–5 were built in cortical depth. The normalized absolute scalar product (AbsScalarProd) of the major DTI eigenvectors and normal vectors perpendicular to the lamellae was calculated [4]. AbsScalarProd was mapped onto the inflated cortex followed by statistical analysis separately in GM banks, and GM gyri and sulci. GM banks were identified by a WM-GM surface curvature within the interval [–0.15, 0.15] 1/mm2, and gyri and sulci were characterized by curvature values outside this interval (Fig.3). The cortical depth dependence of diffusion properties including AbsScalarProd were studied using histogram analysis.
RESULTS
Fig.4 displays the major diffusion eigenvector (e1) fields in a gyrus of primary somatosensory (a) and motor cortex (b). The major eigenvectors at the WM-GM interface in banks (blue arrow 1) are mostly parallel to the interface and orthogonal at the tip of the gyrus (blue arrow 2). The major diffusion direction changes in the medial lamella, where the major eigenvectors are orthogonal to the cortex in banks (yellow arrow 1) and gyri (yellow arrow 2).
This observation is confirmed quantitatively in histograms of AbsScalarProd of vector fields calculated for the whole brain cortex at different depth. At the WM-GM interface Fig.5a demonstrates a strong peak of AbsScalarProd at 0 (major eigenvector orthogonal to surface normal) for banks while Fig.5b shows concentration of values at 1 for sulci and gyri (major eigenvector parallel to the surface normal). Histograms of the AbsScalarProd at the medial cortex lamella at banks (Fig.5c) and sulci/gyri (Fig.5d) evidence the radial character of the major eigenvector at this cortical depth throughout the brain.
DISCUSSION AND CONCLUSION
We had shown that major DTI eigenvectors at high magnetic fields has the ability detecting a sharp fiber turning in the human cortex in vivo [5, 6]. The analysis of DTI is facilitated by the AbsScalarProd parameter which reflects the curvature of the cortex and shape of lamellae in the cortical depth. Thus DTI with isotropic resolution robustly probes structural differences along the cortex and in the cortical depth. Potentially AbsScalarProd as a metric to measure the fiber bending in different parts of the WM-GM interface, may serve as an index for detecting diseases and brain malformation. ▪
Declaration of Competing Interest. The author declares that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Readers feedback
