Researchers interested in relaxivity mapping and white matter characterization.Apparent transverse relaxivity, $$$R_2^*$$$, in white matter has been shown to depend on the orientation of fibers in respect to the applied magnetic field. This effect originates in the anisotropic microstructure of the white matter (WM) and has been used in ex-vivo samples to create fiber orientation maps^[1,2]. The aim of this project is to combine the fiber orientation information retrieved from DTI acquisitions (which are the gold standard to study white matter orientation) with the sensitivity to microstructural information from $$$R_2^*$$$ acquired with different head positions. In this manner it is possible to decompose $$$R_2^*$$$ into orientation independent (isotropic component, $$$R_{2,ISO}^*$$$) and orientation dependent (anisotropic component, $$$R_{2,ANISO}^*$$$) components for several main WM fibres.

Data from six subjects were acquired on a 3T scanner (Magnetom Prisma, Siemens Healthcare, Germany) with a 32 channel head coil (Nova Medical)according to a protocol accepted by the local ethics committee using the following sequences:

1. $$$R_1$$$ mapping MP2RAGE: TR/TI1/TI2/TE= 6/700/2000/2.34s, α1/α2 = 6°/5°, res= 1mm, Tacq=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.$$$R^*_2$$$ mapping 3D GRE: TR/TE1-TE5=63/6.15-57.18ms, α=10°, res= 1.5 mm, iPAT=2x2, Tacq=2min39sec.The scan was repeated up to 8 times with the subject’s head oriented along different orientations relative to the main magnetic field.

The $$$R_2^{*}$$$ maps were created using the GRE data as in ^[3] and the relative head positions were co-registered using FSL-FLIRT pipeline as in ^[3]. Rotation matrices from the previous co-registration and DWI information were combined to calculate the angle, $$$\vartheta_i $$$, of fibres in respect to the main magnetic field for each head position $$$ i $$$. Probabilistic tractography (FSL-PROBTRACKX) was used to extract five major WM fibres masks (cst-corticospinal tract, cg-cingulum, ilf-inferior longitudinal fasciculus, fmj-forceps major, fmi-forceps minor). Although the complete model of orientation dependence is more complex ^[1], it has been shown that, $$ R_{2}^*= R_{2,ISO}^*+ R_{2,ANISO}^* \cdot \sin^4(\vartheta_i) $$ [Eq.1] Is sufficient for a good characterization on the orientation dependence of $$$ R_2^* $$$ ^[2] Even in ex-vivo with full rotation flexibility, the fitting of four parameters to a small set of orientations makes the problem sensitive to noise amplification ^[2]. In this in-vivo study, to improve the sensitivity to the $$$ R_2^*$$$ parameters, the used number of fitting parameters was reduced to two by using prior knowledge from the DWI on the orientation of fibres.

Figure 1 shows qualitatively the orientation dependence of $$$ R_{2}^* $$$ and the corresponding angle maps. It is possible to observe in the forceps major that $$$ R_{2}^* $$$ increases as the angle between the fibre and B0 increases. Due to the mobility restrictions inside the head coil (<30°), the whole angle range (0 to 90º) was not obtained for most studied WM fibres. Because of its initial orientation (head-foot) this was particularly low for the corticospinal tract (Fig.2). Although the variability of the $$$ R_{2}^* $$$ maps for cst and fmi was lower compared to the others, it was sufficient to obtain good fitting of the orientation independent and dependent components as can be seen on the top row of Figure 2 with the fit following the $$$50^{th}$$$ percentile curve. The maps of the isotropic component of $$$R_2^*$$$ maps for one subject (See Figure 3a,b) show relatively small variations of intensity throughout the brain, with only some variations in the corpus callosum/cingulum region Fig.3 (a). Orientation dependent $$$R_2^*$$$ of the same subject are shown in Figure 3 (d) and (e) demonstrating similar patterns to those observed by bound water pool fraction methods ^[4] and Vista ^[5]. These tend to show higher values of myelination of main fibre bundles in the middle of the brain, that decay towards the cortex. Despite the fibre differences, $$$R_{2,ISO}^*$$$ and $$$R_{2,ANISO}^*$$$ for all subjects are consistent between the left and right hemisphere and for both WM fibres components, Fig.3 (c,f; black and gray bars).Orientation dependent $$$R_{2,ANISO}^*$$$ and independent $$$R_{2,ISO}^*$$$ components of $$$ R_{2}^* $$$ were calculated using the fibre orientation prior knowledge from DTI. The $$$R_{2,ISO}^*$$$ shows a higher variability between the fibre bundles studied, with corticospinal tract having the highest value, and cingulum and inferior longitudinal fasciculus having the lowest values. The anisotropic component is a candidate to study myelination of fibre bundles. In order to increase the angle range and robustness of the fitting procedure and perform histology on the studied samples, future work should include ex-vivo measurements.