Malaria,
a
dangerous
disease
transmitted
through
mosquito
bites
and
caused
by
Plasmodium
parasites,
presents
substantial
threat
to
human
health.
The
primary
aim
is
streamline
the
process,
rendering
it
quicker,
more
straightforward,
highly
efficient.
foremost
objective
create
robust
computer
model
capable
of
swiftly
distinguishing
cells
in
thin
blood
samples
obtained
from
standard
microscope
slides.
These
will
be
categorized
as
either
infected
or
uninfected,
employing
advanced
image
processing
techniques
facilitate
prompt
effective
testing.
Additionally,
authors
intend
harness
capabilities
machine
learning
for
classifying
cell
images.
purpose
firmly
rooted
desire
enhance
accuracy
speed
malaria
diagnosis,
ultimately
contributing
early
identification
management
this
life-threatening
ailment.
Indian Journal of Science and Technology,
Год журнала:
2023,
Номер
16(34), С. 2730 - 2739
Опубликована: Сен. 15, 2023
Objectives:
This
study
explores
the
potential
of
deep
learning-based
techniques
to
improve
disease
management
and
intervention
by
focusing
on
their
use
in
infectious
prediction
prognosis.
Methods:
The
research
used
learning
models
EfficientNetB0,
NASNetLarge,
DenseNet169,
ResNet152V2,
InceptionResNetV2.
For
this
study,
a
dataset
comprising
29,252
images
different
diseases
such
as
COVID-19,
MERS,
Pneumonia,
SARS,
tuberculosis.
To
visualize
pixel
intensity,
exploratory
data
analysis
was
performed
pictures.
Preprocessing
eliminated
disruptive
signals
via
image
augmentation
contrast
enhancement.
After
that,
Otsu
thresholding
contour
feature
morphological
values
retrieved
relevant
features.
Findings:
best
successful
model
found
be
EfficientNetB0.
During
training,
it
obtained
90.22%
accuracy
rate,
loss
0.279,
having
an
RMSE
value
0.578.
However,
InceptionResNetV2
showed
accuracy,
loss,
throughout
testing.
precise
results
were
88%,
0.399,
0.631,
respectively.
Novelty:
novelty
resides
exploring
methods
based
for
predicting
prognosticating
diseases,
with
handling
strategies
intervention,
public
health
decisions.
Keywords:
Tuberculosis;
Pneumonia;
Infectious
diseases;
Deep
learning;
Malaria,
a
dangerous
disease
transmitted
through
mosquito
bites
and
caused
by
Plasmodium
parasites,
presents
substantial
threat
to
human
health.
The
primary
aim
is
streamline
the
process,
rendering
it
quicker,
more
straightforward,
highly
efficient.
foremost
objective
create
robust
computer
model
capable
of
swiftly
distinguishing
cells
in
thin
blood
samples
obtained
from
standard
microscope
slides.
These
will
be
categorized
as
either
infected
or
uninfected,
employing
advanced
image
processing
techniques
facilitate
prompt
effective
testing.
Additionally,
authors
intend
harness
capabilities
machine
learning
for
classifying
cell
images.
purpose
firmly
rooted
desire
enhance
accuracy
speed
malaria
diagnosis,
ultimately
contributing
early
identification
management
this
life-threatening
ailment.
