The aim of the study was to evaluate feasibility of whole brain adiabatic T1r and Relaxation Along a Fictitious Field (RAFF) (1, 2) imaging in healthy volunteers and explore age related change in gray (GM) and white matter (WM).

Twenty-eight healthy volunteers (age 24-69) provided written informed consent and underwent MRI examination performed using a clinical 3 T MR scanner. Four healthy volunteers had second MR examination performed within 4 weeks after the initial scan to evaluate short term repeatability. The adiabatic T_1ρ and RAFF measurements were acquired using a 3D T1-FFE sequence with the following parameters: TR/TE 4.1/2.3 ms, acquisition voxel size with whole brain coverage 1.5x1.5x1.5 mm³, FOV 240x240 mm², TFE factor 20, SENSE factor 2.5 in right-left and 2.0 in feet-head direction, centric k-space coding, acquisition time 3 minutes 19 seconds per frame. The adiabatic T_1ρ data sets were obtained using hyperbolic secant (HS) pulses with the RF peak amplitude 575 Hz (corresponding to 13.50 mT, B₁), pulse duration 12 ms, pulse train duration of 72ms and 144 ms. A second-order rotating frame (RAFF2) and fourth-order rotating frame (RAFF4) were performed using RF peak amplitude 500 Hz (corresponding to 11.74 mT , B₁), 260 Hz (corresponding to 6.13 mT, B₁), respectively. The pulse train duration for both RAFF2 and RAFF4 were of 68 ms and 135 ms. In addition, T1-weighted images were acquired using a 3D T1-FFE sequence with the following parameters: TR/TE 8.0/3.7 ms, acquisition voxel size with whole brain coverage 1.0x1.0x1.0 mm³, FOV 240x240 mm², SENSE factor 2.2 in right-left direction, acquisition time 5 minutes 29 seconds. All MR examinations were performed within specific absorption rate limit. The relaxation values of adiabatic T_1ρ (T_{1ρ_adia}), RAFF2 (T_RAFF2) and RAFF4 (T_RAFF4) were calculated using two parameter monoexponential model. Rigid co-registration was performed using FSL (3) 5.0.4 while SPM8 (4) with VBM8 toolbox (5) was used in re-slicing and segmentation of T1-weighted images. T1-weighted images were segmented to GM, WM, Cerebrospinal Fluid using VBM8 toolbox (5) and subcortical segments was excluded from GM and WM masks using FIRST (6) tool. Cerebellum obtained from MNI atlas was excluded from both GM and WM. In addition, voxels located between left and right caudatus and thalamus were excluded from GM masks. Absolute differences, differences, and relative differences in the relaxation values on voxel level of volunteers undergoing repeated MR examination (4 in total) were calculated following co-registration of the imaging data sets. Differences in the median values of T1r_adia, TRAFF2 and TRAFF4 between the age groups (20-29, 30-39, 40-49, 50-59, 60-69) were compared using Mann–Whitney test. The correlation of the median relaxation values with age was explored using Pearson correlation coefficient. P-values below 0.05 were considered statistically significant. All statistics were performed using SPSS 22 (7).

The MR examination was successful in all healthy volunteers (Figure 1). The number of healthy volunteers in 20-29, 30-39, 40-49, 50-59, 60-69 age groups was 7, 5, 6, 6, 4, respectively. The relative differences on the voxel level were below 5% in GM and WM for all of the relaxation parameters except of T_RAFF4 in GM which was 7.3% (Figure 2). Representative differences on the voxel level between the repeated measures are shown in Figure 3. The differences in the median relaxation values between the groups did not reach the level of statistical significance (Figure 4). Similarly, Pearson correlation coefficients were not statistically significant for any of the relaxation parameters.We have demonstrated feasibility of whole brain adiabatic T_1ρ, RAFF2 and RAFF4 imaging using a clinical 3T MR scanner. All of the relaxation measurements shown high repeatability on the voxel level. In our study population no statistically significant age related changes in WM and GM were present as evaluated by adiabatic T_1ρ, RAFF2 and RAFF4 imaging.