Machine learning reveals correlations between brain age and mechanics

Our brain undergoes significant micro- and macroscopic changes throughout its life cycle. It is therefore crucial to understand the effect of aging on the mechanical properties of the brain in order to develop accurate personalized simulations and diagnostic tools. Here we systematically probed the mechanical behavior of n=439 brain tissue samples in tension and compression, in different anatomical regions, for different axon orientations, across five age groups. We used Bayesian statistics to characterize the relation between brain age and mechanical properties and quantify uncertainties. Our results, based on our experimental data and material parameters for the isotropic Ogden and the anisotropic Gasser-Ogden-Holzapfel models, reveal a non-linear relationship between age and mechanics across the life cycle of the porcine brain. Both tensile and compressive shear moduli reached peak values ranging from 0.4–1.0 kPa in tension to 0.16–0.32 kPa in compression at three years of age. Anisotropy was most pronounced at six months, and then decreased. These results represent an important step in understanding age-dependent changes in the mechanical properties of brain tissue and provide the scientific basis for more accurate and realistic computational brain simulations. Statement of significance: In this paper, we investigate the age-dependent mechanical properties of brain tissue based on different deformation modes, anatomical regions, and axon orientations. Hierarchical Bayesian modeling was used to identify isotropic and anisotropic material parameters. The study reveals a nonlinear relationship between shear modulus, degree of anisotropy, and tension-compression asymmetry over the life cycle of the brain. By demonstrating the non-linearity of these relationships, the study fills a significant knowledge gap in current research. This work is a fundamental step in accurately characterizing the complex relationship between brain aging and mechanical properties.

Subjects

DDC Class

610: Medicine, Health

620.1: Engineering Mechanics and Materials Science

Publication version

publishedVersion

Lizenz

https://creativecommons.org/licenses/by-nc-nd/4.0/

Name

1-s2.0-S1742706124005865-main.pdf

Type

Main Article

Size

4.14 MB

Format

Adobe PDF

Options

Machine learning reveals correlations between brain age and mechanics