Integrating material properties from magnetic resonance elastography into subject-specific computational models for...

Area of Science:

• Neuroscience
• Biomechanical Engineering
• Medical Imaging

Background:

• Computational brain models aid in understanding brain trauma.
• Magnetic Resonance Elastography (MRE) non-invasively estimates brain material properties.
• Integrating MRE data into computational models remains an open challenge.

Purpose of the Study:

• To generate subject-specific computational brain models using novel MRI data.
• To incorporate MRE-derived material properties into these models.
• To compare simulated brain deformations with experimental data under non-injurious loading.

Main Methods:

• Developed subject-specific computational brain models for five volunteers.
• Estimated linear viscoelastic (LVE) material properties directly from MRE data.
• Compared model predictions to tagged MRI-acquired brain deformation data.

Main Results:

• Model predictions closely matched the spatial distribution and magnitude of measured peak strains.
• Accuracy was within 0.005 mm/mm for 95th percentile in-plane peak strains and 0.012 mm/mm for maximum principal strain.
• Simulations showed minimal sensitivity to material property heterogeneity.

Conclusions:

• Incorporating MRE-derived material properties into biofidelic, subject-specific models is feasible and accurate.
• This approach is a significant step towards simulating more severe loading conditions.
• Future research can explore higher-order model features and advanced simulations.