Assessing the (ani)sotropic component of R2 as a mean of studying White Matter properties
Rita Gil1, Diana Khabipova1,2, Marcel Zwiers1, Tom Hilbert3,4,5, Tobias Kober3,4,5, and José P. Marques1

1Donders Institute, Radboud University, Nijmegen, Netherlands, 2Centre d'Imagerie BioMédicale, École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, 3Advanced Clinical Imaging Technology (HC CMEA SUI DI BM PI), Siemens Healthcare AG, Lausanne, Switzerland, 4Department of Radiology, University Hospital (CHUV), Lausanne, Switzerland, 5LTS5, École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland


In this study we investigate the orientation dependence of transverse relaxivity (R2) maps in white matter (WM) due to susceptibility effects of myelin microstructure. Subjects’ heads were rotated along different orientations with respect to B0 and R2 values (within different WM fibre populations) were decomposed into R2 isotropic and anisotropic components (orientation independent and dependent respectively). Differences found in isotropic values were associated with fibres different diameter whereas differences found in anisotropic values were associated with the susceptibility effects from myelin. It was showed that the orientation of WM fibres influences R2 contrast and coherence between hemispheres was also observed.


Researchers interested in characterization of White Matter (WM) and R2 mapping.


In this study we investigate the orientation dependence of transverse relaxivity (R2) maps in WM. Diffusion MRI has been widely used to map WM fibre orientations yet, it cannot provide information on myelin integrity or volume fraction. Apparent transverse relaxivity (R*2) has been shown to vary as a function of fibre orientation [1] due to the susceptibility effects generated by the structurally organized myelin. Such susceptibility effects of microstructure are expected to be present also on R2 maps in the fast diffusion regime. In this abstract we demonstrate this hypothesis in-vivo and study the level of anisotropy across brain fibre populations.


Data from six subjects was acquired on a 3T scanner (Magnetom Prisma, Siemens Healthcare, Germany) using a 32 channel head coil with a protocol approved by the local ethics committee. The following sequences were used:

1-R1 mapping MP2RAGE [2]: TR/TI1/TI2/TE=6000/700/2000/2.34msec; FA1/FA2=6o/5o; res:1mm isotropic; Taq=7min32sec;

2- DWI EPI: TR/TE=3490/74.6msec; res:1.5mm isotropic; matrix=150x150x90; b-value=1000s/mm2; diffusion encoding directions=137; MBfactor=3; Taq=8min47sec;

3- Model-based accelerated R2 mapping multi-echo-spin-echo (MESE) prototype sequence [3]: TR/TE1/TE10=4080/9.6/96msec; BW=363Hz/px; res:1.5mm isotropic; MARTINI [4] undersampling=3; GRAPPA=2; slices=80; Taq=2min53sec.

The R2 mapping protocol was repeated 6-7 times with the subject's head rotated along different orientations with respect to the main magnetic field (B0). Data was processed using custom-developed Matlab scripts and FSL tools [5]. R2 maps acquired with different head orientation were co-registed to the diffusion imaging space. The post-processed DWI was used to calculate angle maps (θ) between the main fibres relative to B0. The R2 co-registration matrices were used to calculate the θ maps of the different fibres in the different positions. Finally, five WM fibre masks were created using probabilistic tractography (probtrackX from FSL), see Fig.1c. The R2 values measured in each fibre in different head positions were decomposed into an orientation independent (isotropic component, R2iso) and an orientation dependent (anisotropic component, R2aniso) part by applying the model: $$$R_2=R_{2iso}+R_{2aniso}sin^4(\theta)$$$


Figure 1 shows the contrast seen in R2 maps and its visual correlation with diffusion main orientations. In Figure 2, the fitting model is plotted along with the 10th, 50th and 90th percentiles of the measured R2 of fibres (as shown on Fig.1c) for one subject as a function of θ. It can be seen that, for both the total WM inside the masks (in agreement with previous reports [6]) and for each fibre mask, the fitting of the R2 values within each mask closely matches the 50th percentile curve. Interestingly it can be seen that different fibres have different levels of orientation dependence.

The variations of the mean R1 over the 6 subjects (Fig.3c) are small (0.05 Hz) when compared to the R2 isotropic values (Fig. 3a), ruling out contributions of R1 contamination due to slice profile on the measured R2.

From the R2 isotropic bars it is seen that the corticospinal tracts (CST) and forceps major present inferior values than the rest of the fibres which could be attributed to water mobility and axonal size. On the other hand, Fig. 3b shows that the cingulum and forceps major and minor all present significant orientation-dependent R2 (of similar amplitude on left and right hemispheres), while the inferior long fasciculus (ILF) shows no significant anisotropic effects, suggesting reduced susceptibility effects. The CST have the largest variability of anisotropy which might be related with the smaller range of θ values (see Fig. 3d) observed even after the 6 head rotations. This is due to the combination of the restricted space for head movement inside the head coil (rotation <30o in all directions) and the fiber orientation with respect to B0 that lead to a further restriction on sin4(θ) values.


Regional WM variations of R2 were observed; for example, anterior WM had increased relaxivity in respect to posterior WM. It was shown that the orientation of WM fibres influences R2 contrast. For both the anisotropic and isotropic R2 of different fibres, coherence between hemispheres was observed. Isotropic values were very similar for most fibres with the exception of the CST and forceps major which could be associated with their different diameter [7, 8]. As for the anisotropic component of R2, this should be associated with the susceptibility effects from myelin. The lowest effects (0 Hz) were observed in the ILF and the largest effects ranging around 1Hz were found on the cingulum and the CST. Future work should be directed to ex-vivo samples where larger rotations can be achieved allowing a more robust assessment of the orientation dependence of R2.


No acknowledgement found.


[1] Bender, B. and Klose, U. (2010), ‘The in vivo influence of white matter fiber orientation towards B0 on T2* in the human brain’, NMR in biomedicine 23(9), 1071–1076;

[2] Marques, J. P., et al. (2010), ‘Mp2rage, a self bias-field corrected sequence for improved segmentation and t 1-mapping at high field’, Neuroimage 49(2);

[3] Hilbert, Tom, et al. "MARTINI and GRAPPA-When Speed is Taste." Proc. Intl. Soc. Mag. Reson. Med. No. EPFL-CONF-210638. 2014;

[4] Sumpf, Tilman J., et al. Model-based nonlinear inverse reconstruction for T2 mapping using highly undersampled spin-echo MRI. Journal of Magnetic Resonance Imaging 34.2 (2011): 420-428;


[6] Michael J. Knight et al Biomedical Spectroscopy and Imaging, "Anisotropy of spin-echo T2 Relaxation by magnetic resonance imaging in the human brain invite", in press;

[7] Stikov, Nikola, et al. "In vivo histology of the myelin g-ratio with magnetic resonance imaging." NeuroImage (2015);

[8] Truex, R. C. and Carpenter, M. B. (1969), Human Neuroanatomy, Pearson Education Inc.


Fig.1: Coronal (top) and transverse (bottom) slices of an R2 map of a representative subject. Overlaying of R2 map (left) with main diffusion directions (middle) and created WM fibre masks (right). Blue: Corticospinal tracts; Green: Cingulum; Yellow: Forceps Minor; Red: Forceps Major. Masks in dark colors come from probtrackX whereas masks in light colors are the final masks (where an FA mask was applied and areas of crossing fibres were excluded)

Fig. 2: Plots showing the fitting of the used model (green line) with the 10th (blue), 50th (red) and 90th (pink) percentile of the measured data for the left hemisphere of a representative subject. a) WM inside all fiber masks; b) Corticospinal Tract; c) Cingulum; d) Inferior Long Fasciculus; e) Forceps Major; f) Forceps Minor.

Fig. 3: a) R2 isotropic values; b) R2 anisotropic values; c) R1 values; d) range of sin4(θ). Plots show values for the different fibres of an averaged subject.

Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)