Frontiers in Computer Science,
Год журнала:
2024,
Номер
6
Опубликована: Ноя. 5, 2024
Lung
cancer
is
the
leading
cause
of
deaths
worldwide.
It
a
type
that
commonly
remains
undetected
due
to
unpresented
symptoms
until
it
has
progressed
later
stages
which
motivates
requirement
for
accurate
methods
early
detection
lung
nodules.
Computer-aided
diagnosis
systems
have
adapted
aid
in
detecting
and
segmenting
cancer,
can
increase
patient's
chance
survival.
Automatic
segmentation
challenging
task
aspects
accuracy.
This
study
provides
comprehensive
review
current
popular
techniques
will
further
research
tumor
segmentation.
presents
implemented
solve
challenges
associated
with
compares
approaches
each
other.
The
used
evaluate
these
accuracy
rates
are
also
discussed
compared
give
insight
future
research.
Although
several
combination
been
proposed
over
past
decade,
an
effective
efficient
model
still
needs
be
improvised
routine
use.
2022 9th International Conference on Computing for Sustainable Global Development (INDIACom),
Год журнала:
2024,
Номер
unknown, С. 1767 - 1773
Опубликована: Фев. 28, 2024
Cancer
is
recognised
to
represent
an
extremely
high
risk
of
mortality,
despite
enormous
developments
having
been
made
in
science
and
medicine.
Characterized
by
widespread
metastases,
malignant
cells
spread
rapidly
evade
drugs,
making
it
a
fatal
disease
with
little
treatment
success.
have
heterogeneous
nature
that
makes
them
resistant
chemotherapy
other
forms
radiation.
Across
the
globe,
cancer
stands
be
second
most
leading
cause
death.
Among
many
types,
lung
colon
are
common
highest
mortality
rate.
Early
accurate
detection
tumor
patients
can
help
medical
industry
increase
patient
survival
statistics.
This
study
focuses
on
improving
current
state
technology
assisted
detection.
A
large
dataset
25,000
histopathological
photographs
tissues
analyzed
build
Deep-learning
model
using
Ensemble
Method
approach
for
reliable
To
efficiency,
photos
divided
into
total
five
different
classes.
The
methodology
underlying
aims
accuracy
building
which
learns
from
pre-existing
models
field;
thus
displaying
superiority
terms
predictive
power.
core
concept
transfer
learning
used
leverage
knowledge
pre-trained
create
better
improved
ensemble
models.
includes
comprehensive
data
preprocessing,
augmentation,
training,
validation
testing,
performance
evaluation.
With
0.96,
this
achieved
reliability
detecting
cells.
effort
holds
potential
improve
diagnosis
through
efficient
classification
images.
Using
effective
reduce
time
resources
required
develop
high-accuracy
Coloration Technology,
Год журнала:
2024,
Номер
unknown
Опубликована: Июнь 10, 2024
Abstract
There
is
a
lot
of
repetitive
work
involved
in
exploring
the
dyeing
performance
natural
dyes.
To
improve
experimental
efficiency,
save
material,
reduce
time
costs
and
shorten
research
cycle,
this
study
collects
analyses
literature
data
350
dye
experiments
to
construct
Natural
Dyes
Dataset,
achieves
rapid
prediction
optimal
reaction
conditions
effects
dyes
using
lightweight
integrated
learning
model.
The
size
trained
XGBoost
model
only
562
KB;
name
its
approximate
chemical
composition
need
be
input
predict
results
environment
pH,
colour
fastness
washing
(CFW)
rubbing
(CFR)
on
silk
fabrics
with
highest
K/S
very
short
52
ms.
accuracies
for
CFW
CFR
validation
set
are
as
high
94.12%,
93.75%
100%,
respectively,
77.78%,
91.67%
83.33%
real
test
set,
both
validity
transferability.
approach
provides
valuable
guidance
small
deployment
inference
time,
expanding
possibilities
cross‐application
disciplines
machine
textile
dyeing.
Frontiers in Computer Science,
Год журнала:
2024,
Номер
6
Опубликована: Ноя. 5, 2024
Lung
cancer
is
the
leading
cause
of
deaths
worldwide.
It
a
type
that
commonly
remains
undetected
due
to
unpresented
symptoms
until
it
has
progressed
later
stages
which
motivates
requirement
for
accurate
methods
early
detection
lung
nodules.
Computer-aided
diagnosis
systems
have
adapted
aid
in
detecting
and
segmenting
cancer,
can
increase
patient's
chance
survival.
Automatic
segmentation
challenging
task
aspects
accuracy.
This
study
provides
comprehensive
review
current
popular
techniques
will
further
research
tumor
segmentation.
presents
implemented
solve
challenges
associated
with
compares
approaches
each
other.
The
used
evaluate
these
accuracy
rates
are
also
discussed
compared
give
insight
future
research.
Although
several
combination
been
proposed
over
past
decade,
an
effective
efficient
model
still
needs
be
improvised
routine
use.
