LatIA,
Journal Year:
2023,
Volume and Issue:
1, P. 17 - 17
Published: Nov. 30, 2023
This
article
explores
how
the
artificial
intelligence
(IA)
it
is
transforming
education
in
natural
sciences
by
means
of
strategies
pedagogic
innovators.
The
IA
allows
learning
personalization,
adjusting
content
and
rhythm
to
individual
necessities
students,
what
improves
understanding
retention
complex
concepts
significantly.
Also,
use
simulations
virtual
models
believe
interactive
visual
environments,
enriching
educational
experience.
These
tools
also
foment
development
critical
creative
skills,
promoting
a
more
active
collaborative
approach
resolution
scientific
problems.
On
whole,
these
not
only
improve
effectiveness
learning,
but
rather
they
prepare
students
face
challenges
XXI
century
with
solid
base
science
technology.
Nature Communications,
Journal Year:
2024,
Volume and Issue:
15(1)
Published: Sept. 8, 2024
Abstract
Pompe
disease
(PD)
is
a
rare
autosomal
recessive
glycogen
storage
disorder
that
causes
proximal
muscle
weakness
and
loss
of
respiratory
function.
While
enzyme
replacement
therapy
(ERT)
the
only
effective
treatment,
biomarkers
for
monitoring
are
scarce.
Following
ex
vivo
biomarker
validation
in
phantom
studies,
we
apply
multispectral
optoacoustic
tomography
(MSOT),
laser-
ultrasound-based
non-invasive
imaging
approach,
clinical
trial
(NCT05083806)
to
image
biceps
muscles
10
late-onset
PD
(LOPD)
patients
matched
healthy
controls.
MSOT
compared
with
magnetic
resonance
(MRI),
ultrasound,
spirometry,
testing
quality
life
scores.
Next,
results
validated
an
independent
LOPD
patient
cohort
from
second
site.
Our
study
demonstrates
enables
subcellular
pathology
increases
glycogen/water,
collagen
lipid
signals,
providing
higher
sensitivity
detecting
degeneration
than
current
methods.
This
translational
approach
suggests
implementation
complex
care
these
patients.
IEEE Journal of Translational Engineering in Health and Medicine,
Journal Year:
2024,
Volume and Issue:
12, P. 468 - 479
Published: Jan. 1, 2024
Objective:
Blood
circulation
is
an
important
indicator
of
wound
healing.
In
this
study,
a
tissue
oxygen
saturation
detecting
(TOSD)
system
that
based
on
multispectral
imaging
(MSI)
proposed
to
quantify
the
degree
(StO
2
)
in
cutaneous
tissues.
Methods:
A
segmentation
algorithm
used
segment
automatically
and
skin
areas,
eliminating
need
for
manual
labeling
applying
adaptive
optics.
Animal
experiments
were
conducted
six
mice
which
they
observed
seven
times,
once
every
two
days.
The
TOSD
illuminated
tissues
with
wavelengths
light
-
red
(λ
=
660
nm)
near-infrared
880
nm),
StO
levels
calculated
using
images
captured
monochrome
camera.
ResNet34-based
U-Net
was
integrated
computer
vision
techniques
improve
its
performance.
Results:
revealed
achieved
Dice
score
93.49%.
determined
varied
significantly
among
phases
Changes
detected
before
laser
speckle
contrast
(LSCI)
changes
blood
flux.
Moreover,
statistical
features
extracted
from
LSCI
utilized
principal
component
analysis
(PCA)
visualize
different
healing
phases.
average
silhouette
coefficients
(ResNet34-based
U-Net)
0.2890
0.0194,
respectively.
Conclusion:
By
segmentation,
accurately
distinguished.
This
method
can
support
medical
personnel
conducting
precise
assessments.
Photoacoustics,
Journal Year:
2023,
Volume and Issue:
34, P. 100575 - 100575
Published: Nov. 29, 2023
Accurate
needle
guidance
is
crucial
for
safe
and
effective
clinical
diagnosis
treatment
procedures.
Conventional
ultrasound
(US)-guided
insertion
often
encounters
challenges
in
consistency
precisely
visualizing
the
needle,
necessitating
development
of
reliable
methods
to
track
needle.
As
a
powerful
tool
image
processing,
deep
learning
has
shown
promise
enhancing
visibility
US
images,
although
its
dependence
on
manual
annotation
or
simulated
data
as
ground
truth
can
lead
potential
bias
difficulties
generalizing
real
images.
Photoacoustic
(PA)
imaging
demonstrated
capability
high-contrast
visualization.
In
this
study,
we
explore
PA
network
training
without
need
expert
annotation.
Our
(UIU-Net),
trained
ex
vivo
tissue
datasets,
remarkable
precision
localizing
needles
within
The
evaluation
segmentation
performance
extends
across
previously
unseen
human
(collected
from
an
open-source
repository).
Specifically,
data,
Modified
Hausdorff
Distance
(MHD)
value
stands
at
approximately
3.73,
targeting
error
around
2.03,
indicating
strong
similarity
small
orientation
deviation
between
predicted
actual
location.
A
key
advantage
our
method
applicability
beyond
images
captured
specific
systems,
extending
other
systems.
LatIA,
Journal Year:
2023,
Volume and Issue:
1, P. 17 - 17
Published: Nov. 30, 2023
This
article
explores
how
the
artificial
intelligence
(IA)
it
is
transforming
education
in
natural
sciences
by
means
of
strategies
pedagogic
innovators.
The
IA
allows
learning
personalization,
adjusting
content
and
rhythm
to
individual
necessities
students,
what
improves
understanding
retention
complex
concepts
significantly.
Also,
use
simulations
virtual
models
believe
interactive
visual
environments,
enriching
educational
experience.
These
tools
also
foment
development
critical
creative
skills,
promoting
a
more
active
collaborative
approach
resolution
scientific
problems.
On
whole,
these
not
only
improve
effectiveness
learning,
but
rather
they
prepare
students
face
challenges
XXI
century
with
solid
base
science
technology.
