Cancers,
Год журнала:
2025,
Номер
17(2), С. 307 - 307
Опубликована: Янв. 18, 2025
Lung
cancer,
the
second
most
common
malignancy
in
both
men
and
women,
poses
a
significant
health
burden.
Early
diagnosis
remains
pivotal
reducing
lung
cancer
mortality.
Given
escalating
number
of
computed
tomography
(CT)
examinations
outpatient
inpatient
settings,
radiologists
play
crucial
role
identifying
early-stage
pulmonary
cancers,
particularly
non-nodular
cancers.
Screening
programs
have
been
instituted
to
achieve
this
goal,
they
raised
attention
within
scientific
community
cancers
associated
with
cystic
airspaces.
These
although
known
for
at
least
decade,
remain
understudied.
Limited
investigations
small
sample
sizes
estimated
their
prevalence
explored
radiological
pathological
features.
airspaces
exhibit
varying
complexities
components
demonstrate
suspicious
changes
over
time.
Adenocarcinoma
is
predominant
histological
type,
often
peripheral
location.
Differential
on
CT
scans
includes
inflammatory
processes
or
emphysema-related
changes.
Unfortunately,
prospective
studies
specifically
analyzing
airspace-associated
are
lacking.
However,
it
that
constitute
approximately
one-fourth
delayed
diagnoses.
Increased
awareness
among
could
lead
more
timely
identification
potentially
reduce
mortality
cost-effective
manner.
Scientific Reports,
Год журнала:
2024,
Номер
14(1)
Опубликована: Март 8, 2024
Abstract
Parasitic
organisms
pose
a
major
global
health
threat,
mainly
in
regions
that
lack
advanced
medical
facilities.
Early
and
accurate
detection
of
parasitic
is
vital
to
saving
lives.
Deep
learning
models
have
uplifted
the
sector
by
providing
promising
results
diagnosing,
detecting,
classifying
diseases.
This
paper
explores
role
deep
techniques
detecting
various
organisms.
The
research
works
on
dataset
consisting
34,298
samples
parasites
such
as
Toxoplasma
Gondii,
Trypanosome,
Plasmodium,
Leishmania,
Babesia,
Trichomonad
along
with
host
cells
like
red
blood
white
cells.
These
images
are
initially
converted
from
RGB
grayscale
followed
computation
morphological
features
perimeter,
height,
area,
width.
Later,
Otsu
thresholding
watershed
applied
differentiate
foreground
background
create
markers
for
identification
interest.
transfer
VGG19,
InceptionV3,
ResNet50V2,
ResNet152V2,
EfficientNetB3,
EfficientNetB0,
MobileNetV2,
Xception,
DenseNet169,
hybrid
model,
InceptionResNetV2,
employed.
parameters
these
fine-tuned
using
three
optimizers:
SGD,
RMSprop,
Adam.
Experimental
reveal
when
RMSprop
applied,
EfficientNetB0
achieve
highest
accuracy
99.1%
loss
0.09.
Similarly,
SGD
optimizer,
InceptionV3
performs
exceptionally
well,
achieving
99.91%
0.98.
Finally,
applying
Adam
InceptionResNetV2
excels,
99.96%
0.13,
outperforming
other
optimizers.
findings
this
signify
coupled
image
processing
methods
generates
highly
efficient
way
detect
classify
Journal of drug targeting,
Год журнала:
2025,
Номер
unknown, С. 1 - 85
Опубликована: Янв. 2, 2025
A
significant
area
of
computer
science
called
artificial
intelligence
(AI)
is
successfully
applied
to
the
analysis
intricate
biological
data
and
extraction
substantial
associations
from
datasets
for
a
variety
biomedical
uses.
AI
has
attracted
interest
in
research
due
its
features:
(i)
better
patient
care
through
early
diagnosis
detection;
(ii)
enhanced
workflow;
(iii)
lowering
medical
errors;
(v)
costs;
(vi)
reducing
morbidity
mortality;
(vii)
enhancing
performance;
(viii)
precision;
(ix)
time
efficiency.
Quantitative
metrics
are
crucial
evaluating
implementations,
providing
insights,
enabling
informed
decisions,
measuring
impact
AI-driven
initiatives,
thereby
transparency,
accountability,
overall
impact.
The
implementation
fields
faces
challenges
such
as
ethical
privacy
concerns,
lack
awareness,
technology
unreliability,
professional
liability.
brief
discussion
given
techniques,
which
include
Virtual
screening
(VS),
DL,
ML,
Hidden
Markov
models
(HMMs),
Neural
networks
(NNs),
Generative
(GMs),
Molecular
dynamics
(MD),
Structure-activity
relationship
(SAR)
models.
study
explores
application
fields,
highlighting
predictive
accuracy,
treatment
efficacy,
diagnostic
efficiency,
faster
decision-making,
personalized
strategies,
precise
interventions.
