Predicting Mesoscopic Larmor Frequency Shifts in White Matter With Diffusion MRI—A Monte Carlo Study in Axonal Phantoms
NMR in Biomedicine,
Journal Year:
2025,
Volume and Issue:
38(3)
Published: Feb. 11, 2025
ABSTRACT
Magnetic
susceptibility
MRI
offers
potential
insights
into
the
chemical
composition
and
microstructural
organization
of
tissue.
However,
estimating
magnetic
in
white
matter
is
challenging
due
to
anisotropic
subvoxel
Larmor
frequency
shifts
caused
by
axonal
microstructure
relative
B0
field
orientation.
Recent
biophysical
models
have
analytically
described
how
influences
shifts,
relating
these
a
mesoscopically
averaged
that
depends
on
axons'
fiber
orientation
distribution
function
(fODF),
typically
estimated
using
diffusion
MRI.
This
study
aimed
at
validating
use
estimate
mesoscopic
fields
determining
whether
can
faithfully
dependence
shift
realistic
microstructure.
To
achieve
this,
we
developed
framework
for
performing
Monte
Carlo
simulations
signals
sized
axon
substrates
segmented
with
electron
microscopy.
Our
demonstrated
careful
experimental
design,
it
feasible
fields.
Additionally,
fODF
standard
model
could
predict
shift.
We
also
found
incorporating
intra‐axonal
axial
kurtosis
explain
significant
amount
signal
variance,
thereby
improving
estimation
factor
should
not
be
neglected
when
fitting
model.
Language: Английский
Quantitative susceptibility mapping in magnetically inhomogeneous tissues
Magnetic Resonance in Medicine,
Journal Year:
2025,
Volume and Issue:
unknown
Published: May 1, 2025
Abstract
Purpose
Conventional
quantitative
susceptibility
mapping
(QSM)
methods
rely
on
simplified
physical
models
that
assume
isotropic
and
homogeneous
tissue
properties,
leading
to
artifacts
inaccuracies
in
biological
tissues.
This
study
aims
develop
evaluate
DEEPOLE,
a
deep
learning–based
method
incorporates
macroscopically
nondipolar
Larmor
frequency
shifts
into
QSM
enhance
the
quality
accuracy
of
maps.
Methods
DEEPOLE
integrates
QUASAR
model
convolutional
neural
network
account
for
contributions
neglected
by
conventional
QSM.
We
trained
using
synthesized
data
reflecting
realistic
power
spectrum
distributions.
Its
performance
was
evaluated
against
traditional
algorithms—including
learning
QSM,
(quantitative
residual
mapping),
morphology‐enabled
dipole
inversion
(MEDI),
fast
nonlinear
(FANSI),
superfast
(SDI)—using
digital
brain
with
without
microstructure
effects,
as
well
vivo
human
data.
Quantitative
assessments
focused
estimation
accuracy,
artifact
reduction,
anatomical
consistency.
Results
In
models,
outperformed
producing
maps
fewer
greater
especially
regions
affected
effects.
vivo,
generated
more
anatomically
consistent
mitigated
such
inhomogeneities
streaking,
providing
improved
estimates
gray
matter
white
matter.
Conclusion
Incorporating
through
improves
methodological
advancement
enhances
reliability
measurements,
particularly
studies
neurodegenerative
demyelinating
conditions
where
are
substantial.
Language: Английский
In Vivo Characterization of Magnetic Inclusions in the Subcortex From Nonexponential Transverse Relaxation Decay
NMR in Biomedicine,
Journal Year:
2025,
Volume and Issue:
38(6)
Published: May 5, 2025
ABSTRACT
According
to
theoretical
studies,
MRI
signal
decay
due
transverse
relaxation
in
brain
tissue
with
magnetic
inclusions
(e.g.,
blood
vessels
and
iron‐rich
cells)
is
expected
follow
a
transition
from
Gaussian
behaviour
at
short
echo
times
exponential
longer
times.
The
parameters
carry
information
about
the
size
volume
fraction)
provide
unique
insights
into
microstructure.
However,
gradient‐echo
decays
typically
only
capture
long‐time
behaviour.
We
experimental
evidence
of
nonexponential
human
subcortical
grey
matter
vivo
data
acquired
3
T,
allowing
subsequent
characterization
inclusions.
Gradient‐echo
were
collected
interecho
spacings,
minimal
time
(1.25
ms)
novel
acquisition
strategies
mitigate
motion
cardiac‐induced
effects.
fitted
using
models
that
describe
impact
on
signal.
Nonexponential
provided
superior
fits.
strongest
deviations
detected
substantia
nigra
globus
pallidus.
Numerical
simulations
histological
maps
iron
concentration
replicated
data,
highlighting
non‐haem
can
be
source
decay.
To
investigate
potential
characterize
microstructure,
we
estimated
properties
underlying
two
analytical
models.
Under
static
dephasing
regime,
susceptibility
fractions
ranged
between
1.8–4
0.02–0.04
ppm,
respectively.
Alternatively,
under
diffusion
narrowing
typical
inclusion
was
~2.4
μm.
Both
suggest
an
intermediate
regime
non‐negligible
effect
water
diffusion.
allows
spatial
distribution
material
within
increased
specificity,
applications
for
Parkinson's
disease
other
pathologies.
Language: Английский
The magnetic properties of packings of cylinders
Journal of Magnetism and Magnetic Materials,
Journal Year:
2024,
Volume and Issue:
607, P. 172391 - 172391
Published: July 31, 2024
Powders
of
magnetic
particles
are
used
e.g.
in
additive
manufacturing
magnets,
necessitating
an
investigation
the
properties
such
powders.
In
this
work
we
consider
hard
modeled
as
infinitely
long
cylinders
2D
and
randomly
packed
a
square
container.
The
have
nonlinear
magnetization
curve
with
defined
remanence
coercivity
their
radii
follow
lognormal
distribution
standard
deviation
distinguishing
different
packings.
Using
finite
element
approach
calculate
average
individual
packings
from
these
subtract
value
corresponding
regions
solid
box
to
remove
shape
demagnetization
effect
overall
packing.
We
find
that
at
applied
fields
close
highest
probability
deviate
5%
box.
Away
near-identical
Considering
internally
each
particle,
near
has
10%
region
Thus
while
packing
appear
be
same
box,
there
is
larger
variation
both
between
within
particles,
compared
Language: Английский