Magnetic Resonance Elastography of the Brain

Magnetic resonance elastography (MRE) has emerged as an important technique to extract information about the mechanics of the brain. In MRE, magnetic resonance images (MRI) of waves propagating in soft tissue are acquired non-invasively and analyzed to estimate the tissue’s mechanical properties [1]. Because the brain it is well hidden and delicate, its mechanical behavior in vivo remains incompletely characterized. MRE measurement of mechanical properties and their variations can illuminate the processes of injury, development and disease in the human brain. Traumatic brain injury (TBI) is a leading cause of death and disability in children and young adults [2, 3], and TBI due to explosive blast is a common and debilitating injury among military personnel in combat [4]. In most TBI, the direct cause of injury is the large, fast deformation of brain tissue caused by acceleration of the skull. Brain deformation depends on tissue stiffness (shear modulus) as well as on the brain’s anatomical structure and connections to the skull. The magnitudes, locations, and directions of deformation remain largely unknown however, due to the difficulty of measuring or modeling the brain’s mechanical response. Reliable computer simulations of TBI would be valuable, but such models require (i) accurate estimates of material parameters, (ii) knowledge of mechanical interactions between the brain and skull, and perhaps most importantly, (iii) data with which to compare model predictions. MRE can address all three requirements in vivo. Changes in the brain’s material properties are thought to accompany normal aging [5, 6]. These changes may be amplified in degenerative disorders like Alzheimer’s disease [7, 8]. Brain tumors, like tumors in superficial tissue, also exhibit differences in stiffness relative to healthy brain [9, 10]. These changes in mechanical properties might be used for clinical purposes, for example as techniques for diagnosis or evaluation of therapy. Macroscopic mechanical changes, because they are related to cellular and microstructural properties, might also elucidate the underlying biology of disease and degeneration. Here, I review the experimental methods used in these studies and the techniques to estimate mechanical properties from MRE data. We discuss the assumptions, technical choices, and limitations specific to MRE of the mammalian brain. Particular attention is given to safe and effective actuation, the relationships between frequency, sampling, and length scale, and the choice of an appropriate material model of brain tissue, which in general is nonlinear, dissipative, anisotropic, and heterogeneous.