International Journal of Ophthalmology,
Journal Year:
2023,
Volume and Issue:
17(1)
Published: Dec. 26, 2023
AIM:
To
develop
an
artificial
intelligence
(AI)
diagnosis
model
based
on
deep
learning
(DL)
algorithm
to
diagnose
different
types
of
retinal
vein
occlusion
(RVO)
by
recognizing
color
fundus
photographs
(CFPs).•
METHODS:
Totally
914
CFPs
healthy
people
and
patients
with
RVO
were
collected
as
experimental
data
sets,
used
train,
verify
test
the
diagnostic
RVO.All
images
divided
into
four
categories
[normal,
central
(CRVO),
branch
(BRVO),
macular
(MRVO)]
three
disease
experts.Swin
Transformer
was
build
model,
experiments
conducted.The
model's
performance
compared
that
experts.•
RESULTS:
The
accuracy
in
normal,
CRVO,
BRVO,
MRVO
reached
1.000,
0.978,
0.957,
0.978;
specificity
0.986,
0.982,
0.976;
sensitivity
0.955,
0.917,
1.000;
F1-Sore
0.955
0.943,
0.887
respectively.In
addition,
area
under
curve
diagnosed
0.900,
0.959
0.970,
respectively.The
results
highly
consistent
those
experts,
superior.•
CONCLUSION:
developed
this
study
can
well
RVO,
effectively
relieve
work
pressure
clinicians,
provide
help
for
follow-up
clinical
treatment
patients.
Medicine,
Journal Year:
2024,
Volume and Issue:
103(29), P. e38998 - e38998
Published: July 19, 2024
This
research
aimed
to
examine
the
effectiveness
of
artificial
intelligence
applications
in
asthma
and
chronic
obstructive
pulmonary
disease
(COPD)
outpatient
treatment
support
terms
patient
health
public
costs.
The
data
obtained
using
semiotic
analysis,
content
analysis
trend
methods
were
analyzed
with
strengths,
weakness,
opportunities,
threats
(SWOT)
analysis.
In
this
context,
18
studies
related
asthma,
COPD
evaluated.
strengths
stand
out
as
early
diagnosis,
access
more
patients
reduced
points
that
among
weaknesses
are
acceptance
use
technology
vulnerabilities
intelligence.
Opportunities
arise
developing
differential
diagnoses
examining
prognoses
for
diseases
effectively.
Malicious
use,
commercial
leaks
security
issues
threats.
Although
provide
great
convenience
process
diseases,
precautions
must
be
taken
on
a
global
scale
participation
international
organizations
against
addition,
there
is
an
urgent
need
accreditation
practices
carried
regard.
International Journal of Ophthalmology,
Journal Year:
2024,
Volume and Issue:
17(6), P. 1138 - 1143
Published: May 24, 2024
With
the
advancement
of
retinal
imaging,
hyperreflective
foci
(HRF)
on
optical
coherence
tomography
(OCT)
images
have
gained
significant
attention
as
potential
biological
biomarkers
for
neuroinflammation.
However,
these
biomarkers,
represented
by
HRF,
present
pose
challenges
in
terms
localization,
quantification,
and
require
substantial
time
resources.
In
recent
years,
progress
utilization
artificial
intelligence
(AI)
provided
powerful
tools
analysis
markers.
AI
technology
enables
use
machine
learning
(ML),
deep
(DL)
other
technologies
to
precise
characterization
changes
during
disease
progression
facilitates
quantitative
assessments.
Based
ophthalmic
images,
has
implications
early
screening,
diagnostic
grading,
treatment
efficacy
evaluation,
recommendations,
prognosis
development
common
diseases.
Moreover,
it
will
help
reduce
reliance
healthcare
system
human
labor,
which
simplify
expedite
clinical
trials,
enhance
reliability
professionalism
management,
improve
prediction
adverse
events.
This
article
offers
a
comprehensive
review
application
combination
with
HRF
OCT
diseases
including
age-related
macular
degeneration
(AMD),
diabetic
edema
(DME),
vein
occlusion
(RVO)
presents
prospects
their
utilization.
PRILOZI,
Journal Year:
2024,
Volume and Issue:
45(2), P. 5 - 13
Published: June 1, 2024
Over
the
past
period
different
reports
related
to
artificial
intelligence
(AI)
and
machine
learning
used
in
everyday
life
have
been
growing
intensely.
However,
AI
our
country
is
still
very
limited,
especially
field
of
medicine.
The
aim
this
article
give
some
review
about
medicine
fields
based
on
published
articles
PubMed
Psych
Net.
A
research
showed
more
than
9
thousand
available
at
mentioned
databases.
After
providing
historical
data,
applications
are
discussed.
Finally,
limitations
ethical
implications
IntechOpen eBooks,
Journal Year:
2024,
Volume and Issue:
unknown
Published: Nov. 27, 2024
This
chapter
provides
an
overview
of
various
retinal
abnormalities,
pathophysiologies,
structural
and
vascular
findings,
therapeutic
modalities
used
to
address
vein
obstruction
(RVO)
its
associated
consequences,
which
includes
vision
loss
due
macular
edema,
bleeding,
neovascular
glaucoma
(NVG).
RVO
encompasses
central
occlusion
(CRVO)
branch
(BRVO).
Recent
research
has
highlighted
the
significance
optical
coherence
tomographic
angiography
(OCT-A)
imaging
in
managing
complications
stemming
from
venous
occlusion.
Among
primary
causes
impairment
are
perfused
nonperfused
with
latter
being
most
prevalent.
OCT-A
been
instrumental
identifying
alterations
blood
perfusion
vessel
density.
Treatment
options
for
edema
resulting
include
laser
photocoagulation
therapy,
shown
inconsistent
results.
Additionally,
can
be
addressed
implant
that
releases
corticosteroids
directly
into
eye.
Current
treatments
involve
antivascular
endothelial
growth
factor
(anti-VEGF)
drugs,
such
as
ranibizumab
aflibercept,
well
recently
approved
dual-acting
faricimab.
Furthermore,
port
delivery
system
(PDS)
enhance
outcomes
compliance
management.
treatment
plays
a
critical
role
preventing
sight-threatening
complications.
Ear Nose & Throat Journal,
Journal Year:
2023,
Volume and Issue:
unknown
Published: Dec. 11, 2023
Objective:
We
introduce
a
novel
endoscopic
middle
ear
surgery
dataset
specifically
designed
for
evaluating
deep
learning
(DL)-based
semantic
segmentation
of
chorda
tympani.
Methods:
curated
comprising
8240
images
from
25
patients,
divided
into
training
set
(20%,
1648
images),
validation
(5%,
412
and
test
(75%,
6180
images).
employed
data
enhancement
techniques
to
expand
the
picture
size
sets
by
5
times
(training
set:
images,
verification
2060
Subsequently,
we
multistage
transfer
method
establish,
train,
validate
various
convolutional
neural
networks.
Results:
On
labeled
our
proposed
network
achieved
good
results,
with
U-net
exhibiting
highest
effectiveness
(mIOU
=
0.8737,
mPA
0.9263).
Furthermore,
when
applied
raw
contrasted
prediction
otologists,
overall
performance
was
excellent
(accuracy
0.911,
precision
0.9823,
sensitivity
0.8777,
specificity
0.9714).
Conclusions:
Our
findings
demonstrate
that
DL
can
be
successfully
automatic
tympani
in
surgery,
yielding
high-performance
results.
This
study
validates
potential
feasibility
future
intelligent
navigation
technologies
assist
surgery.
International Journal of Ophthalmology,
Journal Year:
2023,
Volume and Issue:
17(1)
Published: Dec. 26, 2023
AIM:
To
develop
an
artificial
intelligence
(AI)
diagnosis
model
based
on
deep
learning
(DL)
algorithm
to
diagnose
different
types
of
retinal
vein
occlusion
(RVO)
by
recognizing
color
fundus
photographs
(CFPs).•
METHODS:
Totally
914
CFPs
healthy
people
and
patients
with
RVO
were
collected
as
experimental
data
sets,
used
train,
verify
test
the
diagnostic
RVO.All
images
divided
into
four
categories
[normal,
central
(CRVO),
branch
(BRVO),
macular
(MRVO)]
three
disease
experts.Swin
Transformer
was
build
model,
experiments
conducted.The
model's
performance
compared
that
experts.•
RESULTS:
The
accuracy
in
normal,
CRVO,
BRVO,
MRVO
reached
1.000,
0.978,
0.957,
0.978;
specificity
0.986,
0.982,
0.976;
sensitivity
0.955,
0.917,
1.000;
F1-Sore
0.955
0.943,
0.887
respectively.In
addition,
area
under
curve
diagnosed
0.900,
0.959
0.970,
respectively.The
results
highly
consistent
those
experts,
superior.•
CONCLUSION:
developed
this
study
can
well
RVO,
effectively
relieve
work
pressure
clinicians,
provide
help
for
follow-up
clinical
treatment
patients.
