Transcript From Scanner to Instant Camera in MRI: Parallel Imaging

Unsolved Problems and Unmet
Needs in Magnetic Resonance
Morning Categorical Course
ISMRM 14th Scientific Meeting & Exhibition
Seattle, WA, USA
May 9-12, 2005
Why a course on UNsolved problems?
• At the annual ISMRM meeting and in many
of our professional interactions, we tend to
focus on what we or others have recently
accomplished in our areas of interest, or else
we speculate together on current trends and
promising future directions in MR research
and practice.
Why a course on UNsolved problems?
• In the midst of all this lively and topical
activity, the less satisfying questions of what
we cannot but would very much like to
achieve with MR receive little concentrated,
collective attention.
Why a course on UNsolved problems?
• Discussions of unmet needs and research priorities
are often left to funding organizations, which
publish periodic “requests for proposals” and
“roadmaps” to which many of us as researchers are
encouraged to respond.
• The process of assessing needs and formulating
priorities, however, could very well benefit from
broader participation by our MR community at
large.
Overview
• The morning sessions on Unsolved Problems and
Unmet Needs in MR are intended to bring new
attention to such questions.
• Over the course of the next four days, the top ten
reviewed submissions from a prior call for abstracts
on key problems and needs will be presented.
• The sessions will also include substantial time for
open discussion, in order to promote interactions
and to foster innovation.
Educational objectives
Upon completion of this session, participants should be able to:
• Identify and assess a sampling of key unsolved problems
and unmet needs in the field of magnetic resonance;
• Establish criteria for successful solutions to such problems;
• Consider any large-scale coordination across our field or
with other fields that may be called for to address some
classes of research or clinical needs;
• Identify and share additional unsolved problems or unmet
needs in your areas of interest.
Broader objectives:
The ISMRM Strategic Plan
• Vision: The ISMRM aspires to be the
premier international society working to
promote innovation, development,
implementation, and communication of
magnetic resonance science in medicine and
other related fields.
Broader objectives
•
•
•
•
Foster conversation (Strategic plan, goal 1)
Foster collaboration (Strategic plan, goal 1)
Foster innovation (Strategic plan, goal 1)
Promote interactions between our clinciallyfocused and our research-focused members (goal 1)
• Educate new entrants into our field (goal 2)
• Develop a shared sense of collective targets for
research (goals 1,5,6)
• Play an active role in informing funding
organizations by highlighting important directions
(goals 1,5,6)
Precedents in other fields
• Mathematics
– Millennium Prize Problems, Clay Mathematics Institute
(http://www.claymath.org/millennium)
– Multiple individual lists of unsolved problems: Hilbert, Croft, et al
• Physics, Astronomy
– Decadal Survey of Physics
– Decadal Survey of Astronomy and Astrophysics
(http://www7.nationalacademies.org/bpa/)
• Medicine
– The NIH Roadmap
• Others
– The National Academies Keck Futures Initiative Conferences
(http://www.keckfutures.org)
– Intellectual Ventures, Nathan Myhrvold et al
(http://www.intellectualventures.com)
What do we mean by unsolved
problems and unmet needs?
• “If only I had X, then I could diagnose /
monitor / treat Y”
• “If only I could measure / build / control Z,
then I could accomplish X”
• “What is the true mechanism underlying Q?”
• …
What do you mean by unsolved
problems and unmet needs?
• Survey of ISMRM Study Groups returned
nearly 100 items representing 7 Study
Groups
MR Engineering
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Wireless or optical transmission of the MR signal out of the bore, to facilitate
the use of large receiver coil arrays.
What number of channels is optimal/required for both receive (Parallel
Imaging) and transmit (Transmit SENSE)
SAR: how to control SAR at high fields (> 3T), how to know where the hot
spots are in vivo, how to monitor the SAR in the human body in real time.
Length of magnet/bore: how short can it be? What are the limits in openness?
Actively shielded > 7 T whole body magnets.
Dynamic shimming: development of shim coils that can be used in dynamic
shimming (strong, low inductance, actively shielded) and quantification of the
advantages to be gained via dynamic shimming.
Reduction of acoustic noise to due switched magnetic field gradients, especially at
high field.
Methods for imaging without the use of pulsed field gradients.
Methods for Precise QA of the MRI scanner in terms of SNR/ stability/ distortion/
artifacts and differences between MR methods. Currently, the lack of such
methods limit the pooled use of MR data generated from different
scanners/vendors.
Practical methods for measuring the electrical properties of tissues using MRI.
Methods for increasing patient throughput - can we produce an order-of-magnitude
improvement in any way (parallel imaging, higher field, dockable tables, etc.) in
the time needed to image (leading to a reduction in scanning costs).
A discussion board to draw an active group of engineers together for
dissemination of knowledge and experience, rather than retainment in
individual labs.
High-Field
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Solutions for the transmit B1 inhomogeneity at high field due to
wavelength/dielectric effects.
Development of standardized method for SAR monitoring and setting
SAR limits for multiple local transmit coils.
Development of transmit SENSE at ultra-high field, along with
development of algorithms for determining the phase and amplitude of
the RF for each driven element in a coil, such as a TEM coil.
Acoustic noise reduction at high field.
Low noise, high fidelity, and powerful gradient amplifiers.
Development of an actively shielded 7 T whole-body sized magnet.
Fast CSI pulse sequences.
Pulse sequences for clinical applications of hetero-nucleus imaging (such as
3Li, 31P, 17O, etc.)
Detection of molecular probes that can pass through the blood-brain
barrier and selectively bind to specific cells or tissues.
A method for identifying areal boundaries via high resolution MRI (to
allow definition of functional areas on individual subjects).
Long-term health consequences of exposure to very high magnetic field.
Methods to remove magnetic susceptibility-induced image distortion.
Diffusion/Perfusion
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
“Definitively quantify the contributions of the intra- and extra-cellular components to the diffusion signal
“Measure brain perfusion in acute stroke patients accurately enough to be useful” [Score = 27; Number of votes = 7]
“Have realistic phantoms for diffusion imaging” [Score = 26; Number of votes = 11]
”Quantify connectivity between different regions, separated by different distances, reliably and consistently” [Score = 25; Number
of votes = 8]
“Have an accepted and practical gold standard for tract tracing in the human brain” [Score = 23; Number of votes = 7]
“Reliably measure the vascular territories of individual arteries” [Score = 21; Number of votes = 6]
“Better understand the biophysical nature of diffusion MR signal, in order to optimize diffusion experiments more effectively”
[Score = 20; Number of votes = 4]
“Have a definitive, reproducible and easy way of calibrating ASL experiments with respect to M0 (of tissue or blood, depending on
the model) in order to get quantitative CBF values” [Score = 18; Number of votes = 4]
“Use diffusion-derived measures, other than the mean diffusivity, in a clinical manner” [Score = 16; Number of votes = 5]
“Have standard post-processing software (including motion correction) for ASL integrated onto a clinical scanner” [Score = 16;
Number of votes = 5]
“Perform meaningful group comparisons on low dimensionality diffusion data (scalar invariants of the tensor), even if we don’t
understand the biophysical mechanisms underlying them” / “Perform meaningful voxel-based comparisons of DTI data” [Score =
16; Number of votes = 4]
“Reliably quantify the dependence of diffusion on diffusion time, to identify different tissue types or geometrical features” [Score
= 15; Number of votes = 5]
“Have realistic phantoms for perfusion imaging” [Score = 15; Number of votes = 5]
“Use DTI to reliably discriminate tumor infiltration from bland (tumorfree) edema” [Score = 14; Number of votes = 5]
“Use arterial spin labeling to measure perfusion in white matter” [Score = 14; Number of votes = 3]
"Establish the definitive biophysical mechanism underlying the dependence of ADC on the b-value in the brain". [Score = 12;
Number of votes = 4]
“Resolve the topological ambiguity in diffusion displacement profiles (cross, kiss, twist, bend)” [Score = 11; Number of votes = 4]
“Reach a common consensus, once and for all, on the number of directions needed for DTI, and the b-values needed (i.e., do we
need b < 600 s/mm2 or b> 3000 s/mm2)” [Score = 11; Number of votes = 4]
“Identify the cellular correlates of changes in diffusion anisotropy in white matter” [Score = 11; Number of votes = 5]
“Know whether FA abnormalities in specific fibers correlate with fMRI abnormalities in gray matter (i.e., use DTI to predict the
outcome of fMRI experiments)” / “Identify the functional basis and the functional meaning of DTI” [Score = 10; Number of votes
=5
MSK
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Applications of Parallel Imaging in MSK MRI
New coils for 3T and 7T: phased array, small coils
Kinematic and loaded imaging of joints and spine
Novel sequences to image cartilage including SSFP and uTE
Automated segmentation and parameter mapping software
(morphology, T1, T2, T1rho)
Spectroscopy for MSK MRI - muscle, spine, cartilage (1H, 31P, Na)
Fat saturation for low field - robust Dixon imaging
New hardware: gradient inserts and dedicated extremity 3T systems
Fast T1 and T1rho mapping sequences
DWI and DTI imaging for muscle and nerves
11.
Marrow Imaging
Flow and Motion
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
How to achieve robust, accurate, high-temporal-and-spatial-resolution flow
measurements in, for example, the right and circumflex coronary arteries,
which move a lot.
We need a well-defined, complex flow phantom for comparison/calibration
of scanners.
How to have first-time, every-time measurements of flow with likely errors
less than 10% in PATIENTS in vessels down to 5mm diameter?
What consistutes a clinical report of an MR flow study? an MR cardiac
motion study?
When will MR velocimetry match ultrasound?
Spiral, EPI, segmented gradient echo, FVI, etc, etc, etc, do we have the
right tools for every situation? and do we really need them all in clinic and
in research settings?
Limits of temporal, spatial and velocity resolution - how low and how high
can we go (in flow and tagging)?
How well does phase mapping work with accelerated imaging techniques?
How do I know if a flow study was done well?
4D velocity mapping - how fast? how good?
Dynamic MR
1.
2.
3.
4.
A hand-held MRI scanner. The NMR mouse exists and is capable of measuring spectra at very small depths.
However, for hand held medical imaging we would need to
a. Create a reasonably homogeneous B0 field penetrating the body to organ depth.
b. Overcome B0 field inhomogeneities.
c. Create reasonably linear B0 gradients
d. Compensate for the residual non-linearity of the gradients.
e. Have MRI sequences with only adiabatic pulses
f. Have enough B1 penetration for the adiabatic pulse without exceeding SAR limits.
Create “The mother of all sequences” (so named by Paul Bottomley), a truly high resolution 3D pulse sequence
that in a reasonably short scan time measures spin density, T1 and T2. Then use this information to generate any
slice at any oblique angle with any T1 or T2 weighted contrast. This would replace the series of scout images,
oblique scouts, T1 and T2 weighted images that usually make examinations long.
Measure real time changes in spin density at a time scale shorter than T1. Example measure real time changes in
phosphorous metabolites or in intracellular sodium in vivo in muscle or other excitable tissue.
Measure changes in T1 at a time scale shorter than T1.
White Matter
1.
2.
3.
4.
5.
6.
How to distinguish between demyelination, inflammation and axonal loss?
Early diagnosis of Alzheimers (before occurrence of atrophy)
Increased specificity for diagnosis of Multiple Sclerosis in patient with WM lesions
Possibility to establish glioma grading with high confidelity
Expand the MRI capabilities for brain white matter to high resolution imaging of the spine columns
Diagnose White Matter Pathology Definitively using MRI.
This year’s initiative:
Abstract review
• Each abstract reviewed by at least 3 referees
(AMPC members, with expertise chosen to
match the abstract topics)
• Abstracts graded separately from regular
ISMRM abstracts, but on a similar point
system
• The ten top-rated submissions selected for
oral presentation, and grouped into themes
This year’s initiative:
Topics over the week
• Tue May 9: What are we missing? Seeing
through metal and divining with RF coils
• Wed May 10: Can MRI provide a
noninvasive biopsy?
• Thu May 11: Will new contrast agents
revolutionize MRI?
• Fri May 12: Do we need a virtual scanner,
and do we understand real ones?
Tuesday
• What are we missing? Seeing through metal
and divining with RF coils
• Moderators: Stuart Crozier, Michael Smith
• 07:00:
Introduction
Daniel K. Sodickson, M.D., Ph.D.
• 07:15:
Techniques for MR Imaging Near
Metallic Implants
Garry E. Gold, M.D.
• 07:30:
Prospects of Absolute B1 Calibration
Florian Wiesinger, Ph.D.
• 07:45:
Open Discussion
Wednesday
• Can MRI provide a noninvasive biopsy?
• Moderators: Michael Moseley, John Detre
• 07:00:
Cytoarchitectonic MRI: Can MRI Be
Used to Quantify Neural Tissue?
Itamar Ronen, Ph.D.
• 07:15:
Tissue Structure through Diffusion and
Transverse Relaxation Measurements
Valerij G. Kiselev, Ph.D.
• 07:30:
Unresolved Issues in Diffusion and Perfusion
MRI: A Consensus from the Study Group
Derek K. Jones, Ph.D.
• 07:45:
Open Discussion
Thursday
• Will new contrast agents revolutionize MRI?
• Moderator: Thomas Grist, Martin Prince
• 07:00:
Exclusively MRI - Based Molecular Imaging: Can
Magnetic Labeling of Physiologically Important
Compounds via DNP or Parahydrogen-Induced
Hyperpolarization Provide a Potential Supplement or
Replacement of PET?
Joachim Bargon, M.D.
• 07:15:
Direct Detection of Neuromodulation
Rachel Katz-Brull, Ph.D.
• 07:30:
Development of Status Tracers for Myocardial Perfusion
Imaging by MRI
Timothy F. Christian, M.D.
• 07:45:
Open Discussion
Friday
• Do we need a virtual scanner, and do we understand real
ones?
• Moderators: David Hoult, Richard Bowtell
• 07:00:
Need for a Non-Commercial Open-Source MR Simulator
Ralf B. Loeffler, Ph.D.
• 07:15:
Does the Principle of Reciprocity Hold at High Field MR?
Jinghua Wang, Ph.D.
• 07:30:
Reciprocity at High Field: Counterpoint
David Hoult, Ph.D.
• 07:40:
• 07:55:
Open Discussion
Conclusion
Daniel K. Sodickson, M.D., Ph.D.
After the meeting…
• Results of this year’s call for abstracts and survey
of Study Groups will be published on the ISMRM
Web site
• A web list of unsolved problems and unmet needs
will be maintained and updated as a resources
• Future possibilities: workshop, challenge grants, etc
• Your suggestions?
• Gather your submissions for next year…