Journal of Artificial Intelligence and Soft Computing Research,
Journal Year:
2024,
Volume and Issue:
15(2), P. 167 - 195
Published: Dec. 1, 2024
Abstract
It
is
an
extremely
important
to
have
AI-based
system
that
can
assist
specialties
correctly
identify
and
diagnosis
diabetic
retinopathy
(DR).
In
this
study,
we
introduce
accurate
approach
for
DR
using
machine
learning
(ML)
techniques
a
modified
golf
optimization
algorithm
(mGOA).
The
mGOA
optimizes
ML
classifiers
through
finding
the
best
available
parameters
with
respect
objective
functions,
hence
decreases
number
of
features
increases
classifier’s
accuracy.
A
fitness
function
employed
minimize
feature
medical
dataset.
obtained
results
showed
superiority
higher
convergence
speeds
without
extra
processing
costs
across
datasets
compared
several
competitors.
Also,
attained
maximum
accuracy
optimally
reduced
in
binary
multi-class
achieving
CEC’2022
benchmark
other
metaheuristic
algorithms.
Based
on
findings,
three
optimized
called
mGOA-SVM,
mGOA-radial
SVM,and
mGOA-kNN
were
introduced
as
tools
classification
disease
their
performance
was
assessed
Messidor
EyePACS1
datasets.
Experimental
demonstrated
mGOA-SVM
SVM
achieved
remarkable
average
98.5%
precision
97.4%.
The
diabetic
retinopathy
is
a
dominant
stage
of
diabetes
mellitus
which
cause
vision
loss
on
the
retina
if
it
not
identified
early
stage.
Multi-Head
Self
Attention
based
Convolutional
Neural
Network
(MHSA-CNN)
proposed
for
detect
and
classify
retinopathy.
Messidor
STARE
dataset
used
in
this
research
Contrast
Limited
Adaptive
Histogram
Equalization
(CLAHE)
gaussian
filter
are
preprocessing
enhance
contrast
removing
noise.
Grey
Level
Cooccurrence
Matrix
(GLCM)
feature
extraction
Bald
Eagle
Search
Optimization
Algorithm
(BESOA)
selection.
MHSA-CNN
detection
classification
model
capability
at
every
layer
without
altering
parameters.
accuracy,
recall,
specificity,
f1score
precision
estimating
performance.
attains
accuracy
99.73%
98.67%
when
compared
to
Modified
ResNet,
Capsule
Network.
Journal of Chemometrics,
Journal Year:
2024,
Volume and Issue:
38(11)
Published: Aug. 19, 2024
ABSTRACT
Diabetes
is
a
common
and
serious
global
disease
that
damages
blood
vessels
in
the
eye,
leading
to
vision
loss.
Early
accurate
diagnosis
of
this
issue
crucial
reduce
risk
visual
impairment.
The
typical
deep
learning
(DL)
methods
for
diabetic
retinopathy
(DR)
grading
are
often
time‐consuming,
resulting
unsatisfactory
detection
performance
due
inadequate
representation
lesion
features.
To
overcome
these
challenges,
research
proposes
new
automated
mechanism
detecting
classifying
DR,
aiming
identify
DR
severities
different
stages.
figure
out
capture
feature
characteristics
from
samples,
conjugated
attention
transformer
utilized
within
collective
net
model,
which
automatically
generates
maps
diagnosing
DR.
These
extracted
then
fused
through
fusion
function
calculating
weights
produce
most
powerful
map.
Finally,
cases
identified
discriminated
using
kernel
extreme
machine
(KELM)
model.
For
evaluating
severity,
our
work
utilizes
four
benchmark
datasets:
APTOS
2019,
MESSIDOR‐2
dataset,
DiaRetDB1
V2.1,
DIARETDB0
datasets.
illuminate
data
noise
unwanted
variations,
two
preprocessing
steps
carried
out,
include
contrast
enhancement
illumination
correction.
experimental
results
evaluated
well‐known
indicators
demonstrate
suggested
method
achieves
higher
accuracy
99.63%
compared
other
baseline
methods.
This
contributes
development
screening
techniques
less
time‐consuming
capable
identifying
severity
levels
at
premature
level.
Microscopy Research and Technique,
Journal Year:
2024,
Volume and Issue:
unknown
Published: Nov. 2, 2024
ABSTRACT
In
the
worldwide
working‐age
population,
visual
disability
and
blindness
are
common
conditions
caused
by
diabetic
retinopathy
(DR)
macular
edema
(DME).
Nowadays,
due
to
diabetes,
many
people
affected
eye‐related
issues.
Among
these,
DR
DME
two
foremost
eye
diseases,
severity
of
which
may
lead
some
problems
blindness.
Early
detection
is
essential
preventing
vision
loss.
Therefore,
an
enhanced
capsule
generation
adversarial
network
(ECGAN)
optimized
with
rat
swarm
optimization
(RSO)
approach
proposed
in
this
article
coincide
grading
(DR‐DME‐ECGAN‐RSO‐ISBI
2018
IDRiD).
The
input
images
obtained
from
ISBI
unbalanced
data
set.
Then,
fundus
preprocessed
using
Savitzky–Golay
(SG)
filter
filtering
technique,
reduces
noise
image.
image
fed
discrete
shearlet
transform
(DST)
for
feature
extraction.
extracting
features
DR‐DME
given
ECGAN‐RSO
algorithm
categorize
disorders.
implemented
Python
achieves
better
accuracy
7.94%,
36.66%,
4.88%
compared
existing
models,
such
as
combined
cross‐disease
attention
(DR‐DME‐CANet‐ISBI
IDRiD),
category
block
(DR‐DME‐HDLCNN‐MGMO‐ISBI
classification
a
deep
learning‐convolutional
neural
network‐based
modified
gray‐wolf
optimizer
variable
weights
(DR‐DME‐ANN‐ISBI
