MRI-Based Mechanical Assessment of a Biofidelic Brain Phantom with Improved Skull-Brain Modeling

Introduction: Traumatic brain injury (TBI) can trigger a spectrum of short-term and long-term neurological issues, including cognitive deficits, memory loss, emotional and behavioral disturbances, and motor function impairments. Improving the understanding of brain mechanics during head impacts is therefore crucial to advance TBI research, particularly for the design of protective equipment. One approach to safely measure the effects of injurious forces on the living human brain is to use biofidelic brain phantoms, which offer a controlled, reproducible surrogate model for injury experiments, allowing the study of traumatic impacts without endangering lives. A challenge in the development of brain phantoms is to replicate key mechanical properties of the human brain such as stiffness and damping ratio. Our team has developed the Anthropomorphic Neurologic Gyrencephalic Unified Standard (ANGUS) phantom, which is an idealized gyrencephalic brain phantom composed of polyacrylamide hydrogel designed to model the response of the human brain. This study employs magnetic resonance elastography (MRE) and tagged magnetic resonance imaging (tMRI) to analyze and characterize the mechanical properties and deformation response of the ANGUS phantom, following refinements to its fabrication, and to compare its properties to those of the living human brain. MRE combines MRI with head vibration to induce harmonic motion of the brain and estimate material properties, while tMRI non-invasively applies spatial patterns within MRI images to track brain tissue deformation over time, enabling the analysis of strain and tissue displacement. These techniques provide complementary information for studying brain mechanics and for evaluating the biofidelity of brain phantoms.

Materials and

Methods: The ANGUS phantom consists of a 3D-printed “skull” encasing a polyacrylamide hydrogel (10 wt% polymer content) with anatomical features such as cortical gyri and ventricles, immersed in a cerebrospinal fluid simulant (Fig. 1a). The latest version featured inclusion of a falx and a 2% (w/v) polyacrylamide solution to simulate cerebrospinal fluid and arachnoid trabeculae. MRE was performed on a Siemens Aera 1.5T Scanner with a 16-channel receive-only head/neck coil, placing the phantom on a pillow driver connected to a pneumatic actuator for anterior-posterior excitations (Fig. 1b, 1c). MRE captured displacement data through single-shot echoplanar imaging at 50 and 70 Hz, achieving an isotropic resolution of 2.5 mm. Harmonic displacement from phase-contrast images was processed to create complex displacement fields for finite element-based optimization and mechanical property estimation (Fig. 1e, 1f, 2a). For tMRI, a custom device facilitated controlled head rotations of about 32 degrees along the inferior-superior axis, using a latch release and a padded stop for deceleration (Fig. 1d). tMRI acquisitions employed a 1-1 SPAMM tagging sequence, with motion tracked by harmonic phase (HARP) analysis to quantify maximum principal and shear strains within the hydrogel (Fig. 1g, 1h, 2b). Measurements made on the phantom were compared to in vivo measurements in ten healthy adults. Comparisons were also made with the previous version of the phantom, which utilized different polyacrylamide concentrations for the gray and white matter layers and an isotonic solution for cerebrospinal fluid simulation.

Results, Conclusions, and Discussions: The net weight of the entire skull-gel structure was 2.62 kg, and the volume of the polyacrylamide hydrogel was 1652 cm³. The ANGUS phantom exhibited median values for storage modulus, shear stiffness, and wave speed that were higher than those of corresponding in vivo human data at 50 Hz. However, the median values of shear stiffness and wave speed at 70 Hz were within the range of those properties for human data (Fig. 3). In contrast, the median values for loss modulus, damping ratio, and viscosity of the ANGUS phantom were lower than those of human data. The shear stiffness was slightly higher than in the previous version of the ANGUS phantom, yet both versions exhibited similar trends in shear stiffness observed in the human brain for 50Hz and 70Hz. The damping ratio remained consistent with the previous version, although both ANGUS phantoms displayed much lower damping ratios than human brain tissue, likely due to inherent differences in microstructure that affect tissue viscous behavior. In the tMRI experiments, the phantom demonstrated an average peak angular acceleration of 373.1±2.7 radians/s², significantly higher than the 183.0±47.2 radians/s² observed in in vivo human scans. This resulted in higher 50th and 95th percentile maximum principal strain (0.036 and 0.053) and maximum shear strain (0.070 and 0.101), respectively (Fig. 4). The current ANGUS phantom exhibited a markedly improved dynamic response in gel motion post-impact. The peak strain in the ANGUS phantom occurred within 100 ms and resolved within 150 ms, closely resembling in vivo responses. This likely reflects the effects of both the falx and the polyacrylamide solution, thereby better mimicking meningeal attachments at the skull-brain interface. Overall, the ANGUS phantom, which accurately replicates key brain properties and includes detailed anatomical features, demonstrated improved biofidelity in its response to impulsive rotational loading. Employing MRE and tMRI techniques enabled comparison of biomechanical properties to in vivo measurements. Further improvements to the biofidelity of the phantom may still be possible by adjusting the material properties of the phantom, or by incorporating materials that simulate the anisotropy of white matter tissue.

Acknowledgements (Optional): This work was partially supported by NINDS grant U01-NS112120, the Department of Defense in the Military Traumatic Brain Injury Initiative, the Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., and the intramural research program of the NIH. The authors have no conflicts of interest to disclose. The opinions and assertions expressed herein are those of the author(s) and do not reflect the official policy or position of the Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., the Uniformed Services University of the Health Sciences, or the Department of Defense.