Journal of X-Ray Science and Technology,
Journal Year:
2022,
Volume and Issue:
31(1), P. 211 - 221
Published: Nov. 29, 2022
Among
malignant
tumors,
lung
cancer
has
the
highest
morbidity
and
fatality
rates
worldwide.
Screening
for
been
investigated
decades
in
order
to
reduce
mortality
of
patients,
treatment
options
have
improved
dramatically
recent
years.
Pathologists
utilize
various
techniques
determine
stage,
type,
subtype
cancers,
but
one
most
common
is
a
visual
assessment
histopathology
slides.
The
subtypes
are
adenocarcinoma
squamous
cell
carcinoma,
benign,
distinguishing
between
them
requires
inspection
by
skilled
pathologist.
purpose
this
article
was
develop
hybrid
network
categorization
images,
it
did
so
combining
AlexNet,
wavelet,
support
vector
machines.
In
study,
we
feed
integrated
discrete
wavelet
transform
(DWT)
coefficients
AlexNet
deep
features
into
linear
machines
(SVMs)
nodule
sample
classification.
LC25000
Lung
colon
image
dataset,
which
contains
5,000
digital
images
three
categories
benign
(normal
cells),
adenocarcinoma,
carcinoma
cells
(both
cancerous
cells)
used
study
train
test
SVM
classifiers.
results
using
10-fold
cross-validation
method
achieve
an
accuracy
99.3%
area
under
curve
(AUC)
0.99
classifying
these
samples.
Cancers,
Journal Year:
2022,
Volume and Issue:
14(20), P. 5144 - 5144
Published: Oct. 20, 2022
Lung
cancer
is
the
leading
cause
of
cancer-related
death
worldwide,
with
non-small-cell
lung
(NSCLC)
being
primary
type.
Unfortunately,
it
often
diagnosed
at
advanced
stages,
when
therapy
leaves
patients
a
dismal
prognosis.
Despite
advances
in
genomics
and
proteomics
past
decade,
to
progress
developing
tools
for
early
diagnosis,
targeted
therapies
have
shown
promising
results;
however,
5-year
survival
NSCLC
only
about
15%.
Low-dose
computed
tomography
or
chest
X-ray
are
main
types
screening
tools.
without
specific,
actionable
mutations
currently
treated
conventional
therapies,
such
as
platinum-based
chemotherapy;
resistances
relapses
occur
these
patients.
More
noninvasive,
inexpensive,
safer
diagnostic
methods
based
on
novel
biomarkers
paramount
importance.
In
current
review,
we
summarize
genomic
proteomic
utilized
detection
treatment
NSCLC.
We
further
discuss
future
opportunities
improve
effective
Journal of X-Ray Science and Technology,
Journal Year:
2022,
Volume and Issue:
31(1), P. 211 - 221
Published: Nov. 29, 2022
Among
malignant
tumors,
lung
cancer
has
the
highest
morbidity
and
fatality
rates
worldwide.
Screening
for
been
investigated
decades
in
order
to
reduce
mortality
of
patients,
treatment
options
have
improved
dramatically
recent
years.
Pathologists
utilize
various
techniques
determine
stage,
type,
subtype
cancers,
but
one
most
common
is
a
visual
assessment
histopathology
slides.
The
subtypes
are
adenocarcinoma
squamous
cell
carcinoma,
benign,
distinguishing
between
them
requires
inspection
by
skilled
pathologist.
purpose
this
article
was
develop
hybrid
network
categorization
images,
it
did
so
combining
AlexNet,
wavelet,
support
vector
machines.
In
study,
we
feed
integrated
discrete
wavelet
transform
(DWT)
coefficients
AlexNet
deep
features
into
linear
machines
(SVMs)
nodule
sample
classification.
LC25000
Lung
colon
image
dataset,
which
contains
5,000
digital
images
three
categories
benign
(normal
cells),
adenocarcinoma,
carcinoma
cells
(both
cancerous
cells)
used
study
train
test
SVM
classifiers.
results
using
10-fold
cross-validation
method
achieve
an
accuracy
99.3%
area
under
curve
(AUC)
0.99
classifying
these
samples.
