The Prediction Model of High-Frequency Ultrasound Combined with Artificial Intelligence-Assisted Scoring System Improved the Diagnosis of Sclerosing Adenosis and Early Breast Cancer
Breast Cancer Targets and Therapy,
Год журнала:
2025,
Номер
Volume 17, С. 145 - 155
Опубликована: Фев. 1, 2025
Objective:
The
study
aimed
to
apply
an
artificial
intelligence
(AI)-assisted
scoring
system,
and
improve
the
diagnostic
efficiency
of
Sclerosing
adenosis
early
breast
cancer.
Methods:
This
retrospectively
collected
adenopathy
patients
(156
cases)
cancer
(150
in
Henan
Provincial
People's
Hospital
from
August
2020
April
2023.
Results:
area
under
curve
model
constructed
by
clinical
ultrasound
features
combined
AI
predict
identify
two
training
group
was
0.89
0.94,
respectively.
with
best
performance
(training
AUC,
95%
CI,
0.91–
0.97
validation
0.95,
0.90–
0.99)
superior
feature
model,
decision
also
showed
that
Nomogram
had
good
practicability.
In
group,
AUC
sonographer
differential
diagnosis
0.67(95%
0.62–
0.71)
0.89(95%
0.84–
0.93),
respectively,
sonographer's
assessment
better
sensitivity
(1.00
VS
0.73),
but
a
higher
accuracy
rate
(0.66
0.80).
Conclusion:
Age,
lesion
size,
burr,
blood
flow,
risk
score
are
independent
predictors
sclerosing
correlated
score,
US
routine
features,
data,
BI-RADS
grading,
have
performance,
which
can
provide
clinicians
more
effective
tool.
Keywords:
adenosis,
tumor,
ultrasound,
AI,
computer-aided
Язык: Английский
Improving Ultrasound Diagnostic Precision for Breast Cancer and Adenosis with Modality-Specific Enhancement (MSE) - Breast Net
Cancer Letters,
Год журнала:
2024,
Номер
596, С. 216977 - 216977
Опубликована: Май 23, 2024
Язык: Английский
Comparison of ultrasound features and establishment of a predictive nomogram for triple-negative and non-triple-negative breast cancer
Research Square (Research Square),
Год журнала:
2024,
Номер
unknown
Опубликована: Фев. 9, 2024
Abstract
Objective:
The
objective
of
this
study
was
to
compare
ultrasound
features
and
establish
a
predictive
nomogram
for
distinguishing
between
triple-negative
breast
cancer
(TNBC)
non-triple-negative
(non-TNBC).
Materials
Methods:
included
total
205
patients
with
confirmed
TNBC
574
non-TNBC,
randomly
divided
into
training
set
validation
at
ratio
7:3.
All
underwent
examination
received
confirmatory
pathological
diagnosis.
Nodules
were
classified
according
the
Breast
Imaging-Reporting
Data
System
(BI-RADS)
standard.
Subsequently,
conducted
comparative
analysis
clinical
characteristics
ultrasonic
features.
Results:
A
statistically
significant
difference
observed
in
multiple
non-TNBC.
Specifically,
logistic
regression
on
set,
indicators
such
as
posterior
echo,
lesion
size,
presence
symptoms,
margin
characteristics,
internal
blood
flow
signals,
halo,
microcalcification
found
be
(
P
<0.05).
These
then
effectively
incorporated
static
dynamic
model,
demonstrating
high
performance
from
Conclusion:
results
our
demonstrated
that
can
valuable
margin,
flow,
halo
identified
factors
differentiation.
Microcalcification,
hyperechoic
symptoms
emerged
strongest
factors,
indicating
their
potential
reliable
identifying
Язык: Английский
Ultrasound-based comparative analysis and nomogram development for predicting triple-negative and non-triple-negative breast cancer: a 4-year institutional study in Quanzhou First Hospital
BMJ Open,
Год журнала:
2024,
Номер
14(6), С. e085340 - e085340
Опубликована: Июнь 1, 2024
Objective
The
objective
of
this
study
was
to
compare
ultrasound
features
and
establish
a
predictive
nomogram
for
distinguishing
between
triple-negative
breast
cancer
(TNBC)
non-TNBC.
Design
A
retrospective
cohort
study.
Setting
This
conducted
at
Quanzhou
First
Hospital,
grade
tertiary
hospital
in
Quanzhou,
China,
with
the
research
data
set
covering
period
from
September
2019
August
2023.
Participants
included
total
205
female
patients
confirmed
TNBC
574
non-TNBC,
who
were
randomly
divided
into
training
validation
ratio
7:3.
Main
outcome
measures
All
underwent
examination
received
confirmatory
pathological
diagnosis.
Nodules
classified
according
Breast
Imaging-Reporting
Data
System
standard.
Subsequently,
comparative
analysis
clinical
characteristics
ultrasonic
features.
Results
statistically
significant
difference
observed
multiple
Specifically,
logistic
regression
on
set,
indicators
such
as
posterior
echo,
lesion
size,
presence
symptoms,
margin
characteristics,
internal
blood
flow
signals,
halo
microcalcification
found
be
(p<0.05).
These
then
effectively
incorporated
static
dynamic
model,
demonstrating
high
performance
Conclusion
results
our
demonstrated
that
can
valuable
margin,
flow,
identified
factors
differentiation.
Microcalcification,
hyperechoic
halo,
symptoms
emerged
strongest
factors,
indicating
their
potential
reliable
identifying
Язык: Английский