Voxel-based analysis of brain tissue microstructure and morphometry: effects of age, gender and physical activity
Variations in human brain anatomy are typically seen in morphometric features of the cortex and sub-cortical brain structures (cortical volume/surface, thickness and shape, gyrification and asymmetry) estimated from in-vivo magnetic resonance images (MRI). These morphometric indices correlate with behavioral traits (e.g. impulsivity, handedness) as well as neuropathological/-physiological processes (e.g. neurodegeneration). Measures of local morphometry derived from classical T1-weighted MR images (cortical thickness and regional grey matter volume) strongly depend on image contrast. The image contrast in T1-weighted MRI scans reflects macroscopic morphology of cortical and subcortical brain structures but also the local concentration of contrast-dominating tissue properties such as macromolecular (myelin) and iron content. More specifically, multiple tissue properties may interact on local image contrast complicating grey-white matter boundary estimation (Lorio et al., 2014). This is in particular the case for subcortical brain regions with large amounts of iron (Langkammer et al., 2010), potentiated by further iron accumulation during aging (Draganski et al., 2011). Consequently, the sensitivity of brain morphometric indices derived from classical T1-weighted images is a function of the differential microstructural tissue composition of brain regions of interest as well as the hypothesized effects of the phenomenon of interest (e.g. aging) on these particular tissue properties. One possibility to improve the sensitivity of image contrast and hence morphometric estimations is to use MR image sequences with a skewd weighting towards particular tissue properties, thereby reducing the number of potential biological composites interacting on local image contrast. Magnetization transfer (MT) imaging has been shown previously to provide improved image contrast in subcortical brain regions with measurable impact on the estimation of morphometric measures of grey matter volume (GMV) (Lorio et al., 2014). Indeed, MT-based estimations of local GMV, especially in basal ganglia, are more sensitive to age-related changes (Lorio et al., 2016).
Changes in local brain tissue properties across the lifespan were documented in the literature (Callaghan et al., 2014, Acosta-Carboneiro et al., 2015, Draganski et al., 2011, Lorio et al., 2016) but the heterogeneous age distributions across and within these studies (skewed or bimodal age distributions) along with the well-known non-linear trajectories of structural and functional changes across development/aging (Fjell et al., 2014) prohibit a comprehensive view on age-related morphometric and microstructural changes. Our project thereby investigates linear and non-linear age-related associations in MT-based GMV maps as well as in multi-parameter maps of local tissue properties in healthy elderly subjects.