Transformative low-cost “purpose-built” MRI
systems for brain imaging
Magnetic
resonance imaging (MRI) is a powerful non-invasive technology that provides a
unique window to the structure and the function of the body, with high
resolution, speed, and biological contrast. For the brain, MRI provides
the undisputed standard of care for the diagnosis and monitoring of
neurological disorders and injuries including traumatic brain injury
(TBI). MRI is not widely deployable because high-strength magnetic fields
(of order 1 T) are necessary to obtain useful brain images. These systems
involve large, heavy, fragile, and expensive equipment (such as
superconducting magnets) that is incompatible with operation in field
hospitals and other forward sites, and as a result high-field MRI
instruments offer limited utility for imaging in these contexts. We contend
that non-cryogenic low-magnetic-field implementations of MRI can be
developed to allow robust, transportable imaging modalities well suited to
diagnose the types of injuries prevalent in TBI, and practical for field
operation.
We have been
leading a program to develop tools and techniques for low-magnetic-field
implementations of MRI focused on brain imaging. This program has
been rethinking conventional approaches to MRI scanner
construction, and thus far has shown potential to create transformative
non-cyrogenic low-field deployable MRI scanners with high diagnostic impact and
low power and siting requirements. These scanners, by their ubiquity, will
lead to entirely new ways of informing and practicing field/forward medicine. Our
research program develops tools and techniques to revolutionize
conventional MRI scanner construction and create transformative and widely
deployable low-field scanners with high diagnostic impact. The potential
ubiquity of these inexpensive instruments would democratize MRI, and lead to
entirely new ways of informing and practicing medicine. Toward this goal,
the LFI is now a “high-performance” low-field scanner and challenges the
presuppositions about what is indeed possible at low field. Our recent results at
6.5 mT have demonstrated heretofore unattainable speed and resolution, and
result from a program of advanced hardware and high efficiency pulse
sequence development combined with state of the art undersampling and
compressed sensing techniques. This work has established a proof of principle
of high performance, robust, and potentially clinically relevant
techniques and technologies, and opens up doors to transformational and
revolutionary robust applications of MRI.
Advanced compute-based methods for quantitative image
acquisition and reconstruction
The availability
of inexpensive GPU-based compute has opened the door to new quantitative
model-based strategies for the acquisition and the reconstruction of
highly-undersampled imaging data. We have been developing neural network deep
learning based approaches such as AUTOMAP and other DRNs to leverage
scalable-compute and signficiantly deburden the need for precision hardware.
Nanodiamond-enhanced MRI using in situ Overhauser
hyperpolarization
Nanodiamonds are
of interest as nontoxic substrates for targeted drug delivery and as highly
biostable fluorescent markers for cellular tracking. Beyond optical techniques,
however, options for noninvasive imaging of nanodiamonds in vivo are severely
limited. We have demonstrated that the Overhauser effect, a proton–electron
polarization transfer technique, can enable high-contrast magnetic resonance
imaging (MRI) of nanodiamonds in water at room temperature and ultra-low
magnetic field. The technique transfers spin polarization from paramagnetic
impurities at nanodiamond surfaces to 1H spins in the surrounding water
solution, creating MRI contrast on-demand. We examine the conditions required
for maximum enhancement as well as the ultimate sensitivity of the technique. The
ability to perform continuous in situ hyperpolarization via the Overhauser
mechanism, in combination with the excellent in vivo stability of nanodiamond,
raises the possibility of performing noninvasive in vivo tracking of
nanodiamond over indefinitely long periods of time.
Molecular imaging using continuous SABRE induced
hyperpolarization
Current
hyperpolarized imaging with dissolution DNP and superconducting MRI scanners is
very powerful because of its unique ability to track chemical
transformations in vivo. However, the experiments are relatively
cumbersome, slow and expensive. We have been working in collaboration with
Warren Warren and Thomas Theis at Duke to develop a milliTesla (mT) approach to
hyperpolarized MRI that is simple, fast, and attaiable at modest cost with high
sensitivity. In addition, the mT approach promises biomolecular imaging of more
biological processes by taking advantage of significantly extended relaxation
times attainable at ultra-low fields.
We have combined
low-cost, ULF MRI with parahydrogen based hyperpolarization. Specifically, we
combine our 6.5 mT ULF MRI scanner with SABRE (Signal Amplification By
Reversible Exchange) which hyperpolarizes substrates directly in room
temperature liquids. SABRE uses an Iridium based catalyst which reversibly
binds parahydrogen and substrate to enable hyperpolarization transfer. This
process allows for continuous re-hyperpolarization of substrates even after
detection.SABRE hyperpolarization is optimized at 6.5 mT because of nuclear
spin energy level mixing of parahydrogen derived hydrides and substrate protons
at exactly this field. Our initial experiemnts in this realm are compelling:
SABRE enhacements of 1,650,000-fold of a 100 mM pyridine solution compared to
thermal (at 6.5 mT) are easily attainable!
New MRI methods for “injury imaging”—MRI of
free radicals
We have recently
demonstrated a a high-resolution approach for imaging free radicals with
low field MRI that offers new perspectives for the non-invasive
measurement of free radicals in living organisms: free radical
sensitive Overhauser-enhanced MRI (OMRI). An in vivo implementation in
which endogenous free-radical injury markers provide an unambiguous
non-invasive marker for ischemia reperfusion injury has potential to
transform the field of hyperacute stroke and transient ischemic attack
diagnosis and management, and clarify the mechanisms involved in
reperfusion injury and the local effects of novel therapies.
Similarly, free
radical imaging may transform diagnosis and treatment of secondary injury
following TBI. The use of technology in vivo has the
potential to become a powerful tool to obtain pathophysiologic insight into the
spatiotemporal kinetics of free radical generation following brain injury,
monitor the impact of therapies directed to alleviate free radical-mediated
cell damage, and ultimately aid in clinical evaluation and treatment of acute
TBI. Indeed, timely and non-invasive tomographic detection of free radical
production and oxidative stress can guide therapeutic interventions to be
tailored for each patient.
MRI of root growth in the field
We are developing
a deployable low-field MRI system that can image intact soil-root systems.
This project is a collaboation between Texas A&M (College Station), NIST
(Boulder), ABQNMR, and our laboratory, and the system will measure root
biomass, architecture, 3D mass distribution, and growth rate, and could be used
for selection of ideal plant characteristics based on these root metrics. It
will also have the ability to three-dimensionally image soil water content, a
key property that drives root growth and exploration. Operating much like a MRI
used in a medical setting, the system can function in the field without
damaging plants, unlike traditional methods such as trenching, soil coring, and
root excavation. The team will test two different approaches: an in-ground system
shaped like a cylinder that can be inserted into the soil to surround the
roots; and a coil device that can be deployed on the soil surface around the
plant stem. If successful, these systems can help scientists better understand
the root-water-soil interactions that drive processes such as nutrient uptake
by crops, water use, and carbon management. This new information is crucial for
the development of plants optimized for carbon sequestration without
sacrificing economic yield. The project also aims to help develop ideal energy
sorghum possessing high root growth rates, roots with more vertical angles, and
roots that are more drought resistant and proliferate under water limiting
conditions.
Hyperpolarization and Spin-exchange physics
Ultra-low field
MRI is made practical by the process of noble gas
“hyperpolarization”. In this laser optical pumping process, the NMR
signal of noble gases such as 3He and 129Xe can be increased by four to five
orders of magnitude by increasing the atoms’ nuclear spin polarization. This
field of research continues to be of great importance as NMR and MRI
applications for spin-polarized noble gases are myriad. Research by
investigators outside the small community of spin-physicists has, however,
been critically hampered by an almost complete lack of access to polarized gas
by the community of end users. As a way to tackle this problem, our group
is part of a multi-institution collaboration (Brigham and Women’s
Hospital, Harvard, SIU Carbondale, University of Nottingham, and Vanderbilt)
to further improve the performance and to aid in the dissemination of noble gas
polarizers. Our Consortium has constructed a large-scale (>1 L/hr) 129Xe
polarizer for clinical, pre-clinical, and materials NMR/MRI applications and
attained near-unity nuclear polarization with it. Comprised of mostly
off-the-shelf components, our automated, modular polarizer is easy to use
and employs a simplified “open-source” design, readily implementable in other
laboratories. This first “consortium polarizer” is installed at BWH and
collaborators there are using it to acquire images on clinical scanners,
including measurements of human pulmonary septal thickness in vivo.
Of present
interest are experimental aspects of clinical-scale production of hyperpolarized
(HP) xenon by stopped-flow SEOP,3,4 specifically SEOP cell design (including
the use of alternative materials, coatings, and geometries), laser technology,
and ultra-low field versus high-field polarimetry of 129Xe and 131Xe.
Long lived nuclear singlet states as a new contrast
mechanism in NMR and MRI
In collaboration
with Ron Walsworth’s group at Harvard, have been investigating novel and
efficient ways to create and manipulate nuclear-spin singlet states and related
dressed states in naturally-occurring molecules as powerful tools for nuclear
NMR spectroscopy and MRI. Such novel spin states have wide applicability in
chemistry, physics, biology, and materials science, e.g., to improve the
detection of NMR signals in crowded spectra, to measure and image weak magnetic
couplings in molecules, and to perform high-resolution NMR spectroscopy at very
low magnetic fields. This work has led to the development of a new quantum
filter utilizing nuclear spin singlet states which allows undesired NMR
spectral background to be removed and target spectral peaks to be revealed.
This new NMR contrast mechanism is named SUCCESS, or “Suppression of
Undesired Chemicals using Contrast-Enhancing Singlet States,” and has been
demonstrated in vitro for three target molecules relevant to
neuroscience: aspartate, threonine, and glutamine. We are working to
develop SUCCESS into a robust, widely-applicable technique for NMR and MRI,
with particular emphasis on brain biomolecules (such as GABA) for
application to neurology and neuroscience.
Shot noise limited optical detection of NMR
High-field NMR
and MRI systems are noise-limited by the induced voltages in the sample, and
not by coil losses or amplifier noise—this is the classic “body noise
dominated” regime of high-frequency operation. Although no gains are
likely to be had from improving coil Q, or by reducing the coil or amplifier
noise temperature in this regime, improved B sensitivity could allow for
more flexibility in the placement of the receive sensor for a fixed
detection SNR. Compared to a resistive coil, optical shot-noise (not
Johnson noise) would fundamentally limit a magnetofluorescence-based
sensor. Similarly, consideration of the noise-scaling laws of
optical-based detection devices might allow the high filling-factor requirement
manifestly engineered into close-fitting MRI probes to be significantly
relaxed, culminating in more open geometry detectors, for example an
massive array (N>512) of small, high Q, optical sensors around
a sample. Potential great gains in SNR and acceleration could be obtained
from such a massively parallel detector system. Continued work on this
project is ideally suited for a graduate student interested in the overlap
of coherent electron spin dynamics, quantum optics, and NMR.
Molecular imaging with OMRI using targeted
free-radical probes
NMR
relaxation-based targeted probes that recognize important biomarkers of
atherosclerosis, apoptosis, necrosis, angiogenesis, thrombosis and inflammation
have been developed, and manifest themselves in changes in local relaxation.
Once injected and bound this targeted molecular-contrast may not be “turned
off”. A new idea that we are working on in collaboration with Peter Caravan’s
Molecular Imaging Contrast Agent Lab at the Martinos Center is to leverage the
high-resolution and high speed OMRI attainable in the LFI with TEMPOL towards a
new paradigm of molecular imaging, one based on hyperpolarized radical-based
injected targeted agents, where the hyperpolarization is done via in vivo DNP
at low field. Unlike relaxation-based MRI molecular imaging agents, this new
approach using the Overhauser effect not only can enhance the signal at the
site of the functional agent, but also proves a way to modulate this
enhancement at will via applied RF, allowing for a more clean background-free
detection and potentially lowering the detection sensitivity using narrow-band
lock-in techniques. This could lead to purpose-built portable very specific
molecular-imaging based high-resolution and low cost screening systems. This
would allow targeted- and anatomical imaging to be performed in the same
system, at the same time, as a new deployable tool for transforming diagnosis
based on molecular imaging.
Open-Access, Low Magnetic Field Polarized gas
Pulmonary MRI
We have developed
a novel biomedical imaging technology: magnetic resonance imaging (MRI) at low
magnetic fields (less than 10 mT) of hyperpolarized noble gas (3He)
inhaled into human lungs. This technology enables MRI studies of human
ventilation in a simple, low-cost system with an open geometry (i.e., with a
person standing, sitting, or lying down), without the problems of high magnetic
fields for people with implants such as pacemakers, and free of magnetic
succeptibility artifacts that arise from air-tissue interfaces.
We use this
system to study pulmonary physiology by mapping ventilation and pulmonary
oxygen concentration as a function of body orientation in the gravitational
field. This technology may enable small, portable, low-field MRI systems with
important uses such as imaging the underdeveloped lungs of premature infants,
who often suffer from pulmonary problems and cannot be moved from the neonatal
intensive care unit to a conventional MRI scanner.
Overhauser-enhanced magnetic resonance elastography
(OMRE)
MR elastography
(MRE) is a powerful technique to assess the mechanical properties of living
tissues but its sensitivity is reduced in areas exhibiting short T2, and longer
acquisition times are required to mitigate the lower signal. The longer T2s
inherent to very-low-field MRI can be exploited of to perform high-speed MRE in
combination with Overhauser dynamic nuclear polarization. Indeed, OMRE at
very low magnetic field can be used to detect mechanical waves over short
acquisition times, and this new modality can extend the usefulness of low cost,
portable MRI systems to detect elasticity changes in patients with implanted
devices or iron overload.
