Data-Driven Depiction of Aging Related Physiological Volume Shrinkage in Brain White Matter: An Image Processing Based Three-Dimensional Micromechanical Model

Abstract

Aging in the human brain, both in healthy and pathological conditions, leads to significant microstructural alterations, resulting in cognitive decline, with cerebral atrophy being a major contributing factor. This atrophy, characterized by the loss of neurons and glial cells, plays a crucial role in the reduction of brain function. While magnetic resonance imaging (MRI) and magnetic resonance elastography (MRE) provide noninvasive tools to measure brain morphology (volume changes) and regional mechanical properties (tissue stiffness) at the millimeter scale, they are unable to capture cellular-level or micron-scale changes in brain tissue. The challenge is in correlating the mechanical property changes observed at the millimeter scale with the underlying cellular-level micro-architectural alterations. To address this limitation, an ensemble of three-dimensional micromechanical finite element (FE) models was developed, utilizing MRI/MRE data to compute the mechanics of the aging brain with a higher level of detail. Using image processing techniques in Python's NIBABEL library, a mathematical model was constructed to quantify volume fraction (VF) shrinkage in brain white matter (BWM). These models incorporate uniaxial tensile loading and simulate the interactions between axons, myelin, and the glial matrix. Among the three finite element models compared, model type III, which includes both volume fraction changes and shear modulus degeneration, showed a high-order age-related atrophy and brain softening. This approach emphasizes the significant role of computational mechanics in linking macroscopic MRI measurements to cellular-scale changes, enhancing our understanding of brain tissue degeneration.