We tested a novel type of analysis based on the distribution of the principal diffusion direction (PDD) orientations. In particular, we aimed to assess the power of this method in detecting changes induced in the brain by physiological aging. In this work, we focused on the human thalamus, since it was identified as the subcortical structure in which age-related changes of diffusion tensor imaging (DTI) scalar indexes were most prominent¹.One hundred-fifteen healthy volunteers (58 female, age [mean ± SD] 38.8 ± 11.1 years) were involved in this study. All subjects underwent the same 3 Tesla MRI protocol including whole-brain, three-dimensional, T1-weighted (BRAVO), spoiled gradient recall echo (TE/TR = 3.7/ 9.2 ms, flip angle 12°, voxel size= 1×1×1 mm³) and DTI with the following parameters: b=1000 s/mm2; diffusion-weighting along 27 non-collinear gradient directions; matrix size 128 × 128; 80 axial slice; number of b0 images = 4; NEX = 2; voxel size = 2 × 2 × 2 mm³). All structural images were examined to exclude scans with poor signal-to-noise ratio and/or brain abnormalities. The study was approved by the local ethics committee and all subjects signed their informed consent. Image processing of DTI data was performed as extensively described elsewhere¹ by using FSL². Furthermore, the FSL tool for Bayesian Estimation of Diffusion Parameters Obtained using Sampling Techniques (bedpostX)³ was employed to build up a distribution on diffusion parameters at each voxel, with the possibility of modeling crossing fibers within each voxel in the brain. After applying bedpostX to DTI data, we obtained samples from the distributions of theta and phi angles, which together represent the PDD in spherical polar coordinates. In order to identify samples belonging to the region of interest, binary masks for the left and right thalamus were selected from the Automated Anatomical Labeling (AAL) atlas and nonlinearly warped into each subject’s diffusion space. To assess age-related differences, subjects were divided in two subgroups based on their age: young adults (age range 18-40 years) and elder adults (age range 40-60 years). Statistical analysis and plot generation were performed using in-house Matlab code. In particular, Fisher statistics was used to determine differences between young and elder adults in the distributions of the PDD orientations. The isolines of the spherical polar coordinates distribution were displayed by using contour plots.As expected, widespread age-induced changes were found in bilateral thalami with a symmetrical pattern. As shown in Figure 1, contour plots of theta-phi distributions highlighted a region where frequency of the samples significantly differed between groups (peak at the right side of the plot, red-purple corresponds to young adults, green-blue to elder adults). In the elder adults group, specific PDD orientations were lost, which corresponded to theta values between π/6 and π/5 radians and phi values around -π/5 radians. Figure 2 shows the thalamic voxels in which significantly altered PDD orientations occurred, overlaid onto the MNI template. The Oxford Thalamic Connectivity Probability Atlas, used as post-hoc guide to identify a plausible location for significant voxels, confirmed that differences were located in the regions of the thalamus connected to frontal and parietal regions, in accordance with the frontal aging hypothesis⁴.In this study, we analyzed the distribution of PDD orientations in the thalamus of healthy subjects over a wide age range. The structure was widely altered in the elder adults group, especially in regions connected with frontal and prefrontal regions. These results are in line the well-known frontal aging hypothesis⁴ and further corroborate the datum that thalamo-frontal projections are most affected by age⁵. The coherence of these results with existing knowledge about the aging of the thalamus suggests that exploitation of PDD orientation analysis might become an innovative powerful tool to investigate brain integrity at the microscale.