The
brain
is
vulnerable
to
diseases,
including
infections,
injuries,
and
tumors,
that
can
substantially
influence
daily
life
health;
therefore,
early
diagnosis
treatment
are
necessary.
MRI,
because
of
its
ability
detect
abnormalities
without
interference,
crucial
for
evaluating
structure
function.
Generative
artificial
intelligence
(GAI)
model
disease
characteristics
in
MRI
images,
thereby
increasing
diagnostic
accuracy
by
comparing
healthy
diseased
brains.
This
review
examines
the
transformative
role
GAI
analyzing
images
diagnosing
diseases.
study
explores
five
foundational
models—generative
adversarial
networks,
diffusion
models,
transformers,
variational
autoencoders,
autoregressive
model—and
their
applications
imaging.
These
models
enhance
data
preprocessing,
image
segmentation,
feature
extraction,
supporting
detection.
highlights
GAI’s
superiority
addressing
scarcity
issues,
enhancing
quality,
providing
comprehensive
insights
into
pathology;
it
additionally
discusses
promising
directions
future
research.
Brain‐X,
Journal Year:
2024,
Volume and Issue:
2(2)
Published: April 26, 2024
Abstract
This
comprehensive
review
aims
to
clarify
the
growing
impact
of
Transformer‐based
models
in
fields
neuroscience,
neurology,
and
psychiatry.
Originally
developed
as
a
solution
for
analyzing
sequential
data,
Transformer
architecture
has
evolved
effectively
capture
complex
spatiotemporal
relationships
long‐range
dependencies
that
are
common
biomedical
data.
Its
adaptability
effectiveness
deciphering
intricate
patterns
within
medical
studies
have
established
it
key
tool
advancing
our
understanding
neural
functions
disorders,
representing
significant
departure
from
traditional
computational
methods.
The
begins
by
introducing
structure
principles
architectures.
It
then
explores
their
applicability,
ranging
disease
diagnosis
prognosis
evaluation
cognitive
processes
decoding.
specific
design
modifications
tailored
these
applications
subsequent
on
performance
also
discussed.
We
conclude
providing
assessment
recent
advancements,
prevailing
challenges,
future
directions,
highlighting
shift
neuroscientific
research
clinical
practice
towards
an
artificial
intelligence‐centric
paradigm,
particularly
given
prominence
most
successful
large
pre‐trained
models.
serves
informative
reference
researchers,
clinicians,
professionals
who
interested
harnessing
transformative
potential
IEEE Access,
Journal Year:
2024,
Volume and Issue:
12, P. 108876 - 108896
Published: Jan. 1, 2024
Magnetic
Resonance
Imaging
(MRI)
is
a
critical
imaging
technique
that
provides
detailed
visualization
of
internal
structures
without
harmful
radiation.
This
review
focuses
on
key
MRI
modalities,
including
T1-weighted
and
T2-weighted
imaging,
functional
(fMRI),
diffusion
(dMRI).
images
offer
precise
anatomical
details,
whereas
are
essential
for
highlighting
abnormalities
such
as
tumors
inflammation.
Functional
(fMRI)
captures
blood
flow
changes
related
to
neural
activity,
(dMRI)
tracks
the
movement
water
molecules
within
brain
tissues.
Our
synthesizes
insights
from
173
studies
across
major
databases,
PubMed,
ACM
Digital
Library,
IEEE
Xplore,
Google
Scholar.
We
emphasize
versatility
transformer
architectures
in
neuroimaging
applications,
segmentation,
detection,
reconstruction,
super-resolution,
with
particular
focus
tumor
segmentation
notable
achievement.
Despite
these
successes,
there
remains
significant
gap
research,
need
further
collaborative
efforts
fully
realize
potential
transformers
applications.
Following
PRISMA-ScR
guidelines,
this
analysis
explores
current
trends,
dataset
availability,
overall
research
landscape.
It
calls
scientific
community
investigate
underexplored
capabilities
transformers,
aiming
inspire
comprehensive
could
revolutionize
advance
fields
medical
neuroscience.
Physics in Medicine and Biology,
Journal Year:
2024,
Volume and Issue:
69(8), P. 085005 - 085005
Published: March 13, 2024
Abstract
Objective.
Multi-contrast
magnetic
resonance
imaging
(MC
MRI)
can
obtain
more
comprehensive
anatomical
information
of
the
same
scanning
object
but
requires
a
longer
acquisition
time
than
single-contrast
MRI.
To
accelerate
MC
MRI
speed,
recent
studies
only
collect
partial
k-space
data
one
modality
(target
contrast)
to
reconstruct
remaining
non-sampled
measurements
using
deep
learning-based
model
with
assistance
another
fully
sampled
(reference
contrast).
However,
reconstruction
mainly
performs
image
domain
conventional
CNN-based
structures
by
full
supervision.
It
ignores
prior
from
reference
contrast
images
in
other
sparse
domains
and
target
data.
In
addition,
because
limited
receptive
field,
networks
are
difficult
build
high-quality
non-local
dependency.
Approach.
paper,
we
propose
an
Image-Wavelet
ConvNeXt-based
network
(IWNeXt)
for
self-supervised
reconstruction.
Firstly,
INeXt
WNeXt
based
on
ConvNeXt
undersampled
refine
initial
reconstructed
result
wavelet
respectively.
generate
tissue
details
refinement
stage,
sub-bands
used
as
additional
supplementary
Then
design
novel
attention
block
feature
extraction,
which
capture
image.
Finally,
cross-domain
consistency
loss
is
designed
learning.
Especially,
frequency
deduces
data,
while
retain
high-frequency
final
Main
results.
Numerous
experiments
conducted
HCP
dataset
M4Raw
different
sampling
trajectories.
Compared
DuDoRNet,
our
improves
1.651
dB
peak
signal-to-noise
ratio.
Significance.
IWNeXt
potential
method
that
enhance
accuracy
reduce
reliance
images.
