Development of in vivo Human Brain DTI-MRE

Magnetic resonance elastography (MRE) and diffusion tensor imaging (DTI) are widely-used non-invasive MRI techniques that can characterize the mechanical and diffusive behaviors of biological tissue under various pathological conditions. Both MRE and DTI can provide potential biomarkers (stiffness and diffusivity) for neurological disorders, such as Alzheimer’s disease, amyotrophic lateral sclerosis, meningioma, traumatic brain injury and neuroinflammation. Besides applications in the diagnosis of neurological diseases, DTI and MRE can assess changes occurring in the brain during aging (atrophy, softening, demyelination). Such changes in the microstructural environment combine imaging with changes in tissue diffusion and stiffness that DTI and MRE measure in a complimentary manner by encoding vibration in the phase of the MRI signal while simultaneously modulating diffusion in the signal magnitude. Both techniques require specialized magnetic field gradients that are applied in three-dimensional space. Thus, simultaneous acquisitions of DTI and MRE can be achieved by inducing mechanical vibration and by choosing a motion encoding gradient (MEG) that is sensitive to both diffusion and vibration. The benefits of acquiring DTI and MRE together include obtaining multiple parameters without increasing scan time and requiring no co-registration between diffusive and mechanical property maps. Previous studies using a mouse brain demonstrated the feasibility of simultaneous acquisition of DTI and MRE. However, further improvement of the accuracy in property maps requires minimizing the interference between diffusion and vibration encoding (specific aim 1). Additionally, we propose simultaneous acquisition of DTI and multi-frequency MRE (mMRE), to increase the effective image resolution for brain stiffness (specific aim 2). Overall, both aims improve the sensitivity and specificity of DTI and MRE by more efficient diffusion and vibration encoding, and by minimization of signal loss. This dissertation seeks to develop and optimize the new dual-imaging tool that enables concurrent acquisitions of DTI and MRE (DTI-MRE) in the human brain. The development and optimization of the dual-imaging technique was conducted by simulating different MEG waveforms and by finding the optimal experimental parameters that minimize the signal loss due to intravoxel phase dispersion on the magnitude images. In order to test the simulation results, two sets of parameters, with different signal loss on the magnitude images were selected and tested on five healthy volunteers. Simultaneous acquisition of DTI and mMRE was performed at vibration frequencies of 40 Hz and 50 Hz with the same diffusion encoding on nineteen healthy volunteers. A correlation analysis was carried out on both voxel-wise values and ROI-wise averaged values between novel and conventional techniques. The ROI-wise correlation between multi-frequency DTI-MRE and conventional acquisitions was high on separate property maps, while taking account of the repeatability of each method. The multi-frequency DTI-MRE produced higher voxel-wise correlation of mechanical property maps than mono-frequency DTI-MRE, while maintained good correlation (greater than 0.75) of diffusive property maps. Furthermore, multifrequency DTI-MRE and 3D-multifrequency-MRE have close interindividual mean values of stiffness maps with acceptable error margin, in the meantime, the multifrequency DTI-MRE and conventional DTI measurements have nearly identical mean values of fractional anisotropy and mean diffusivity. In conclusion, multifrequency DTI-MRE saves imaging time and improves the diagnostic accuracy of DTI and MRE.