The increased NMR sensitivity inherent at higher magnetic fields can produce richly-detailed anatomic, functional, and metabolic information that advances our understanding of physiology and disease. Yet, high magnetic fields present a formidable engineering challenge given two major issues that must be addressed: radiofrequency (RF) field inhomogeneity and power deposition in tissue. MRBEL addresses these challenges by focusing on novel hardware and methodology for high-field MR, particularly RF coil design, in vivo spectroscopy of multiple nuclei, and computational modeling of RF fields in the human body.
The Purdue Neurotrauma Group (PNG) is a collaborative research team that combines expertise in clinical diagnosis and care, biomechanics, and neuroimaging. Our mission is to identify the mechanisms that lead to traumatic brain injury (TBI) in athletes, soldiers, and victims of blunt force impacts, and to develop early detection methods, protective systems, and directed therapies to provide the fastest recovery possible.
The PNG is at the forefront of recent discoveries in detection and characterization of head trauma among adolescent athletes in contact sports. Notably, functional MRI (fMRI) scans elucidated measurable neurocognitive and neurophysiological impairments in high school football athletes without clinically-diagnosed concussion. This discovery suggests more players are subject to neurological injury than are currently being detected using traditional concussion-assessment tools.
The increased sensitivity inherent at ultra-high magnetic fields (>3T) can produce richly-detailed anatomic, functional, and metabolic information that advances our understanding of physiology and disease. However, high field MR presents a formidable engineering challenge owing to radiofrequency (RF) effects that intensify at higher frequencies. Two major issues must be addressed: RF field (B1) inhomogeneity and high power deposition, quantified as specific absorption rate (SAR). Unlocking the potential of high field MR requires a broad knowledge base incorporating hardware design, electromagnetic modeling, software development, and NMR physics.
B1 homogeneity hinges on the abilityto define transmission from each coil element. With a single transmit channel, the RF coil hardware itself can perform this task; we demonstrated this capability in a four-element, single-channel, quadrature volume coil for prone breast imaging (diagram shown above right). A novel technique now dubbed forced current excitation (FCE, diagram shown above left) exploited the properties of quarter wavelength transmission line segments to drive equal currents to multiple coil elements.
Magnetic resonance spectroscopy (MRS) has the ability to correlate between lesion anatomy and underlying biochemistry. Performing the technique at higher static magnetic fields such as 7T opens the doors to detecting in vivo spectra from non-hydrogen nuclei, e.g., carbon-13 (13C). Multiple coils tuned to different RF frequencies must employ new designs to overcome B1 and SAR challenges. For 13C spectroscopy, these challenges are compounded by j-coupling, necessitating simultaneous high-power proton-decoupling. We utilized the FCE breast coil to obtain the first in vivo broadband proton-decoupled carbon-13 spectrum from a human at 7T (at right).
A major challenge for high field studies is to limit power deposition in tissue, quantified as specific absorption rate (SAR). Resulting tissue heating is the primary patient safety concern for high field scans. It is instrumental to model the radiofrequency (RF) coil and its interactions with the lossy dielectric loading presented by the body. Our research focuses on the value of modeling as a design tool and its role in SAR safety characterization. To simulate coil performance, the MRE Lab employs our in-house MATLAB library in tandem with models utilizing the finite-difference time-domain (FDTD) method and the method of moments (MoM). SAR characterization is typically performed with FDTD and can be simulated with a growing number of available human body voxel models.
In certain cases such as prone breast imaging, custom tissue models may be necessary to accurately simulate coil loading conditions. Accordingly, we continue to research implications on projected SAR at 7T by simulating a whole-body voxel model seamlessly fused with high-resolution, anatomically-correct breast phantoms spanning the four breast tissue density classifications, differentiated by the proportions of lipid and fibroglandular tissues. The guidance presented in this work will facilitate more patient-specific scanning protocols incorporating the subject’s identified breast density classification.
Parallel transmit (pTx) is a technology in development to address some of the challenges in high field MR, namely non-uniform excitations due to susceptibility artifacts and the shortened RF wavelength. pTx is of particular interest for neuroimaging, yet its application is limited owing to SAR uncertainty. The MRE Lab is collaborating with investigators at the University of Pittsburgh to better understand the more complex SAR interactions posed by pTx MRI and MR spectroscopy (MRS) of the brain. The development of an adaptable and coherent SAR algorithm will be a key to 7T’s future approval for widespread clinical use.