Synopsis
The quantitative properties of the cerebellum were assessed by acquiring
T1 and T2*contrasts at 7T and mapping these onto the cerebellar cortex. T1 maps
showed medial-lateral alternating stripes of different intensities while T2*
values were homogeneously distributed across the lobes. This study showed the
heterogeneity of the cerebellar cortex in terms of tissue content, which in
part parallels a well-established gene expression pattern.Target
audience
Neuroscientists and physicists interested in the cerebellum and its
quantitative properties
Purpose
The purpose of this study was to map T
1 and T
2*values
on the cerebellar surface.
The cerebellum is usually considered to have a relatively homogenous
cortex in contrast to the cytoarchitectonic heterogeneity
of the neocortex. Nevertheless, the cerebellar cortex is heterogeneous in terms
of gene expression (e.g. zebrin II pattern), function and microcircuitry
activity pattern
1,2. The regional cortical
properties of the cerebellum have been so far studied using microscale
techniques such as electrophysiology and immunochemistry
3,4. Whether microscale patterns
can be seen at the submillimeter scale, as provided by ultra-high field MRI, is
still unknown. Here, we performed surface analysis of the cerebellar cortex using
quantitative T1 and T2*maps acquired at 7TMRI.
Methods
Nine healthy participants (2 females, age=31±7), took part in the study.
Whole-brain T
1 and T
2*maps were acquired using
the MP2RAGE
5 (TR/TE/TI1/TI2
6000/2.84/750/2350ms, matrix:300x320x160, 0.75x0.75x0.9mm
3) and 3D
multi gradient echo (MGE; TR 45ms TE1/ΔTE/TE9 4.59/4.59/41.3ms, matrix:300x320x160, 0.75x0.75x0.9mm
3) sequences at 7T (Siemens, Germany)
using a 32 channel head coil (Nova USA). Three dielectric pads were placed
around the upper neck to improve the inversion efficiency over the cerebellum
and whole brain B
1 homogeneity. A mono-exponential fit to the MGE
data was used to obtain the T
2*maps. T
2*was registered to
the MP2RAGE using Elastix
6. A SA2RAGE B
1 map
7 was acquired (TR/TE
2400/0.72ms, matrix:116x128x64, 2.3x2.3x4mm, same transmit voltage as
MP2RAGE) to correct the MP2RAGE for B
1 homogeneity
8. The corrected T
1
map was used for all the following steps. The tools used for the
image processing were part of the CBS tools
9. Both T
1 and T
2*maps
were brought into the MNI space. The T
1 images were subsequently segmented
using the multi-geometric deformable model segmentation algorithm
9. The central cerebellar
surface (middle gray matter layer) was extracted with an adaptation of the CRUISE algorithm
10 which preserves the surface
topology of the cerebellum. A diffeomorphic image registration algorithm
11 was used to realign the T
1
volumes to a high-resolution template from the CBS Tools. Transformation
maps were obtained from the previous procedure and applied for the geometric
averaging of the surfaces and the mapping of the T
1 and T
2*values.
Results
The geometric average of the 9 cerebellar surfaces allows the labelling of
the different cerebellar (Figure 1) because of the precise segmentation and
surface alignment. In the average T
1 cortical surface, T
1
values were higher around the vermis as well as in the fissures, while the
folia showed lower values (Figure 2). Most interestingly, the T
1
values of the folia appeared to have a pattern of medial-lateral alternating stripes
of higher and lower T
1 values (Figure 2). More precisely, the most
medial and most lateral part of the hemispheres had similar values and the T
1
was lower in the longitudinal middle part of a single hemisphere. The T
2*surface
displayed different patterns (Figure 3). The superior lobe of the cerebellum had
higher T
2*values compared to the more inferior lobes. Whether
there is a media-lateral alternating pattern for T
2*values following
the folia of the middle lobe in a medio-lateral fashion is more difficult to
observe. However, an increase of T
2*can be seen when moving from the
posterior to the anterior side (Figure 3).
Moreover, some dots representing high T
2*values in the
superior lobe seem to be vaguely aligned with the blue pattern in Figure 2.
Discussion
While for most of the cerebellar cortical surface quantitative T
1 and
T
2* could be obtained with high confidence, the very low T
1 values
in the most inferior area are likely due to loss
of SNR and correspondingly poor T
1 estimates. Similarly, reduced T
2*values in the inferior lobules might indicate a non-perfect B
0 shim.
We identified two distinguishable patterns arising from the in vivo quantitative mapping. T
1 maps
showed longitudinal stripes while T
2*values appeared to be more homogeneously distributed
across the lobes. In both contrasts, there was an intensity difference between
the folia and the fissures. The alternating longitudinal pattern was very
similar to the well-known zebrin II (also called aldolase C) pattern in
rodent cerebelli
4. The latter has been shown to
have tight relationship with incoming climbing fibers
12. Although still hypothetical,
the observed contrast could be linked to the amount of input myelinated axons
as T
1 contrast is known to be sensitive to myelination
13.
Conclusion
This study showed for the first time in MRI various topographic changes
in tissue properties in the cerebellum, using high-resolution quantitative
acquisitions at 7T. A possible cause for sagittal T
1 zones might be found in
the presence of Zebrin regions, opening a window towards finer differentiation
of cerebellar architecture in vivo.
Acknowledgements
No acknowledgement found.References
1. Cerminara,
N. L., Lang, E. J., Sillitoe, R. V. & Apps, R. Redefining the cerebellar
cortex as an assembly of non-uniform Purkinje cell microcircuits. Nat. Rev.
Neurosci. 16, 79–93 (2015).
2. Apps, R. & Hawkes, R. Cerebellar
cortical organization: a one-map hypothesis. Nat. Rev. Neurosci. 10,
670–681 (2009).
3. Hawkes, R. Purkinje cell stripes and
long-term depression at the parallel fiber-Purkinje cell synapse. Front.
Syst. Neurosci. 8, 41 (2014).
4. Marzban, H. & Hawkes, R. On the
architecture of the posterior zone of the cerebellum. Cerebellum 10,
422–434 (2011).
5. Marques, J. P. et al. MP2RAGE,
a self bias-field corrected sequence for improved segmentation and T1-mapping
at high field. Neuroimage 49, 1271–81 (2010).
6. Klein, S., Staring, M., Murphy, K.,
Viergever, M. a. & Pluim, J. elastix: A Toolbox for Intensity-Based
Medical Image Registration. IEEE Trans. Med. Imaging 29, 196–205
(2010).
7. Eggenschwiler, F., Kober, T., Magill,
A. W., Gruetter, R. & Marques, J. P. SA2RAGE: a new sequence for fast B1+
-mapping. Magn. Reson. Med. 67, 1609–19 (2012).
8. Marques, J. P. & Gruetter, R. New
developments and applications of the MP2RAGE sequence--focusing the contrast
and high spatial resolution R1 mapping. PLoS One 8, e69294
(2013).
9. Bazin, P. L. et al. A
computational framework for ultra-high resolution cortical segmentation at 7
Tesla. Neuroimage 93, 201–209 (2013).
10. Han, X. et al. CRUISE: Cortical
reconstruction using implicit surface evolution. Neuroimage 23,
997–1012 (2004).
11. Avants, B. B., Epstein, C. L.,
Grossman, M. & Gee, J. C. Symmetric diffeomorphic image registration with
cross-correlation: Evaluating automated labeling of elderly and neurodegenerative
brain. Med. Image Anal. 12, 26–41 (2008).
12. Fujita, H. & Sugihara, I. Branching
patterns of olivocerebellar axons in relation to the compartmental organization
of the cerebellum. Front. Neural Circuits 7, 1–9 (2013).
13. Lutti, A., Dick, F., Sereno, M. I.
& Weiskopf, N. Using high-resolution quantitative mapping of R1 as an index
of cortical myelination. Neuroimage 93, 176–188 (2014).