Research Square (Research Square),
Journal Year:
2023,
Volume and Issue:
unknown
Published: Sept. 5, 2023
Abstract
Background
Cerebral
alveolar
echinococcosis
(CAE)
and
brain
metastases
(BM)
are
similar
in
locations
imaging
appearance.
While,
CAE
is
usually
treated
with
chemotherapy
surgical
treatment,
BM
often
radiotherapy
targeted
primary
malignancy
treatment.
Accurate
diagnosis
critical
due
to
the
vastly
different
treatment
approaches
for
these
conditions.
Purpose
This
study
aims
investigate
effectiveness
of
radiomics
machine
learning
on
magnetic
resonance
(MRI)
distinguishing
BM.
Methods
We
have
retrospectively
analyzed
MRI
images
130
patients
(30
CAE,
100
BM,
training
set
=
91,
testing
39)
who
confirmed
or
Xinjiang
medical
university's
first
affiliated
hospital
from
January
2014
December
2022.
Three
dimensional
tumors
were
segmented
by
radiologists
contrast-enhanced
T1WI
open
resources
software
3D
Slicer.
Features
extracted
Pyradiomics,
further
feature
reduction
was
carried
out
using
univariate
analysis,
correlation
least
absolute
shrinkage
selection
operator
(LASSO).
Finally,
we
built
five
models,
support
vector
machine,
logistic
regression,
linear
discrimination
KNeighbors
classifier,
Gaussian
NB
evaluated
their
performance
via
several
metrics
including
sensitivity
(recall),
specificity,
positive
predictive
value
(precision),
negative
value,
accuracy
area
under
curve
(AUC).
Results
The
(AUC)
SVC,
LR,
LDA,
KNN,
algorithms
(testing)
sets
0.99
(0.94),
1.00
(0.87),
0.98
(0.92),
0.97
(0.97),
(0.93)
respectively.
Nested
cross-validation
demonstrated
robustness
generalizability
models.
Additionally,
calibration
plot
decision
analysis
practical
usefulness
models
clinical
practice,
lower
bias
toward
subgroups
during
decision-making.
Conclusion
combination
approach
contrast
enhanced
could
well
distinguish
holds
promise
assisting
doctors
accurate
decision-making
Deleted Journal,
Journal Year:
2024,
Volume and Issue:
37(1), P. 81 - 91
Published: Jan. 18, 2024
Endometrial
carcinoma
(EC)
risk
stratification
prior
to
surgery
is
crucial
for
clinical
treatment.
In
this
study,
we
intend
evaluate
the
predictive
value
of
radiomics
models
based
on
magnetic
resonance
imaging
(MRI)
and
staging
early-stage
EC.
The
study
included
155
patients
who
underwent
MRI
examinations
were
pathologically
diagnosed
with
EC
between
January,
2020,
September,
2022.
Three-dimensional
features
extracted
from
segmented
tumor
images
captured
by
scans
(including
T2WI,
CE-T1WI
delayed
phase,
ADC),
1521
each
three
modalities.
Then,
using
five-fold
cross-validation
a
multilayer
perceptron
algorithm,
these
filtered
Pearson's
correlation
coefficient
develop
prediction
model
performance
was
assessed
analyzing
ROC
curves
calculating
AUC,
accuracy,
sensitivity,
specificity.
terms
stratification,
CE-T1
sequence
demonstrated
highest
accuracy
0.858
±
0.025
an
AUC
0.878
0.042
among
sequences.
However,
combining
all
sequences
resulted
in
enhanced
reaching
0.881
0.040,
corresponding
increase
0.862
0.069.
context
staging,
utilization
combination
involving
T2WI
led
notably
elevated
0.956
0.020,
surpassing
achieved
when
employing
any
singular
feature.
Correspondingly,
0.979
0.022.
When
incorporating
concurrently,
reached
0.000,
accompanied
0.986
0.007.
It
noteworthy
that
level
surpassed
radiologist,
which
stood
at
0.832.
has
potential
accurately
predict
early
Translational Cancer Research,
Journal Year:
2024,
Volume and Issue:
13(2), P. 1083 - 1090
Published: Feb. 1, 2024
Background:
Endometrial
cancer
(EC)
is
an
epithelial
malignancy
occurring
in
the
endometrium,
with
a
5-year
mortality
rate
of
above
10%.
However,
there
currently
lack
studies
exploring
potential
predictive
model
tumor-specific
death
after
surgery
these
patients.
Methods:
From
January
2015
to
December
2017,
data
related
482
patients
EC
admitted
Dushu
Lake
Hospital
Affiliated
Soochow
University
were
analyzed.
Patients
divided
into
(n=62)
and
survival
(n=420)
groups
according
whether
occurred
at
5
years
postoperatively
or
not.
The
clinical
characteristics
two
compared,
risk
factors
for
investigated
by
logistics
regression
analysis.
A
nomogram
prediction
was
established
relevant
factors.
Results:
Tumor
size,
Ki-67
positive
rate,
Federation
International
Gynecology
Obstetrics
(FIGO)
stage,
vascular
tumor
thrombus
between
(P<0.05)
found
be
statistically
significant
Positive
Ki-67,
size
>3.35
cm,
stage
III,
that
influenced
(P<0.05).
obtained
area
under
receiver
operating
characteristic
(ROC)
curves
training
verification
sets
0.847
[95%
confidence
interval
(CI):
0.779–0.916]
0.886
(95%
CI:
0.803–0.969),
respectively.
Conclusions:
model,
which
this
study,
proved
valuable
predicting
EC.
Heliyon,
Journal Year:
2024,
Volume and Issue:
10(12), P. e32940 - e32940
Published: June 1, 2024
This
study
aimed
to
develop
and
validate
a
radiomics
nomogram
based
on
multiparameter
MRI
for
preoperative
differentiation
of
type
II
I
endometrial
carcinoma
(EC).
Magnetic Resonance in Medical Sciences,
Journal Year:
2024,
Volume and Issue:
unknown
Published: Jan. 1, 2024
This
review
explores
the
significant
progress
and
applications
of
artificial
intelligence
(AI)
in
obstetrics
gynecological
MRI,
charting
its
development
from
foundational
algorithmic
techniques
to
deep
learning
strategies
advanced
radiomics.
features
research
published
over
last
few
years
that
has
used
AI
with
MRI
identify
specific
conditions
such
as
uterine
leiomyosarcoma,
endometrial
cancer,
cervical
ovarian
tumors,
placenta
accreta.
In
addition,
it
covers
studies
on
application
for
segmentation
quality
improvement
gynecology
MRI.
The
also
outlines
existing
challenges
envisions
future
directions
this
domain.
growing
accessibility
extensive
datasets
across
various
institutions
multiparametric
are
significantly
enhancing
accuracy
adaptability
AI.
potential
enable
more
accurate
efficient
diagnosis,
offering
opportunities
personalized
medicine
field
gynecology.
Oncology Letters,
Journal Year:
2023,
Volume and Issue:
25(3)
Published: Feb. 6, 2023
The
purpose
of
the
present
study
was
to
investigate
predictive
value
metabolic
syndrome
in
evaluating
myometrial
invasion
(MI)
patients
with
endometrial
cancer
(EC).
retrospectively
included
EC
who
were
diagnosed
between
January
2006
and
December
2020
at
Department
Gynecology
Nanjing
First
Hospital
(Nanjing,
China).
risk
score
(MRS)
calculated
using
multiple
indicators.
Univariate
multivariate
logistic
regression
analyses
performed
determine
significant
factors
for
MI.
A
nomogram
then
constructed
based
on
independent
identified.
calibration
curve,
a
receiver
operating
characteristic
(ROC)
curve
decision
analysis
(DCA)
used
evaluate
effectiveness
nomogram.
total
549
randomly
assigned
training
or
validation
cohort,
2:1
ratio.
Data
gathered
predictors
MI
including
MRS
[odds
ratio
(OR),
1.06;
95%
confidence
interval
(CI),
1.01-1.11;
P=0.023],
histological
type
(OR,
1.98;
CI,
1.11-3.53;
P=0.023),
lymph
node
metastasis
3.15;
1.61-6.15;
P<0.001)
tumor
grade
(grade
2:
OR,
1.71;
1.23-2.39;
P=0.002;
Grade
3:
2.10;
1.53-2.88;
P<0.001).
Multivariate
indicated
that
an
factor
both
cohorts.
generated
predict
patient's
probability
four
factors.
ROC
showed
that,
compared
clinical
model
(model
1),
combined
2)
significantly
improved
diagnostic
accuracy
(area
under
1
vs.
0.737
0.828
cohort
0.713
0.759
cohort).
Calibration
plots
cohorts
well
calibrated.
DCA
net
benefit
is
obtained
from
application
Overall,
developed
validated
MRS-based
predicting
preoperatively.
establishment
this
may
promote
use
precision
medicine
targeted
therapy
has
potential
improve
prognosis
affected
by
EC.
Journal of Cancer,
Journal Year:
2023,
Volume and Issue:
14(18), P. 3523 - 3531
Published: Jan. 1, 2023
Endometrial
cancer
(EC)
is
a
common
gynecologic
malignancy,
with
rising
trend
in
related
mortality
rates.The
assessment
based
on
imaging
examinations
contributes
to
the
preoperative
staging
and
surgical
management
of
EC.However,
conventional
diagnosis
has
limitations
such
as
low
accuracy
subjectivity.Radiomics,
utilizing
advanced
feature
analysis
from
medical
images,
extracts
more
information,
ultimately
establishing
associations
between
features
disease
phenotypes.In
recent
years,
radiomic
studies
EC
have
emerged,
employing
combined
clinical
characteristics
model
predict
histopathological
features,
protein
expression,
prognosis.This
article
elaborates
application
radiomics
research
discusses
its
implications.
Scientific Reports,
Journal Year:
2024,
Volume and Issue:
14(1)
Published: Nov. 6, 2024
It
aimed
to
analyze
the
value
of
deep
learning
algorithm
combined
with
magnetic
resonance
imaging
(MRI)
in
risk
diagnosis
and
prognosis
endometrial
cancer
(EC).
Based
on
convolutional
neural
network
(CNN)
architecture
residual
101
layers
(ResNet-101),
spatial
attention
channel
modules
were
introduced
optimize
model.
A
retrospective
collection
MRI
image
data
from
210
EC
patients
was
used
for
model
segmentation
reconstruction,
140
cases
as
test
set
70
validation
set.
The
performance
compared
traditional
ResNet-101
model,
based
mechanism
(SA-ResNet-101),
(CA-ResNet-101),
using
accuracy
(AC),
precision
(PR),
recall
(RE),
F1
score
evaluation
metrics.
Among
set,
there
45
low-risk
25
high-risk
EC.
Using
ROC
curve
analysis,
it
found
that
area
under
(AUC)
proposed
this
article
(0.918)
visibly
larger
against
(0.613),
SA-ResNet-101
(0.760),
CA-ResNet-101
models
(0.758).
AC,
PR,
RE,
values
higher
(P
<
0.05).
In
postoperative
recurrence
occurred
13
did
not
occur
57
cases.
AUC
prediction
by
(0.926)
(0.620),
(0.729),
(0.767).
article,
assisted
MRI,
presented
superior
diagnosing
patients,
sensitivity
(Sen)
specificity
(Spe),
also
demonstrated
excellent
predictive
AC
prediction.
Acta Radiologica,
Journal Year:
2024,
Volume and Issue:
unknown
Published: Nov. 21, 2024
Background
The
depth
of
myometrial
invasion
determines
whether
International
Federation
Gynecology
and
Obstetrics
stage
I
endometrioid
endometrial
carcinoma
(EEC)
patients
undergo
lymph
node
dissection.
However,
subjective
evaluation
results
relying
on
magnetic
resonance
imaging
(MRI)
are
not
always
satisfactory.
Purpose
To
develop
a
nomogram
based
whole-volume
tumor
MRI
histogram
parameters
to
preoperatively
predict
deep
(DMI)
in
with
EEC.
Material
Methods
This
retrospective
analysis
included
131
EEC
training/validation
cohort
92/39
at
7:3
ratio.
were
obtained
from
multiple
sequences
(ADC
mapping
T2-weighted
imaging)
within
volumes
interest.
Univariate
analysis,
least
absolute
shrinkage
selection
operator
(LASSO)
regression,
multivariate
logistic
regression
used
for
feature
selection.
performance
clinical
model,
was
evaluated
by
calculating
the
area
under
receiver
operating
characteristic
curve
(AUC).
Results
Age
two
morphological
features
(maximum
anteroposterior
diameter
sagittal
images
[APsag]
ratio
[TAR])
selected
construct
model.
Five
creation
nomogram,
which
combines
parameters,
age,
APsag,
TAR,
achieved
highest
AUCs
both
training
validation
cohorts
(nomogram
vs.
model:
0.973
0.871
0.934
[training]
0.972
0.870
0.928
[validation]).
Conclusion
MR
can
help
DMI
preoperatively,
assisting
physicians
development
personalized
treatment
strategies.
Research Square (Research Square),
Journal Year:
2023,
Volume and Issue:
unknown
Published: Sept. 5, 2023
Abstract
Background
Cerebral
alveolar
echinococcosis
(CAE)
and
brain
metastases
(BM)
are
similar
in
locations
imaging
appearance.
While,
CAE
is
usually
treated
with
chemotherapy
surgical
treatment,
BM
often
radiotherapy
targeted
primary
malignancy
treatment.
Accurate
diagnosis
critical
due
to
the
vastly
different
treatment
approaches
for
these
conditions.
Purpose
This
study
aims
investigate
effectiveness
of
radiomics
machine
learning
on
magnetic
resonance
(MRI)
distinguishing
BM.
Methods
We
have
retrospectively
analyzed
MRI
images
130
patients
(30
CAE,
100
BM,
training
set
=
91,
testing
39)
who
confirmed
or
Xinjiang
medical
university's
first
affiliated
hospital
from
January
2014
December
2022.
Three
dimensional
tumors
were
segmented
by
radiologists
contrast-enhanced
T1WI
open
resources
software
3D
Slicer.
Features
extracted
Pyradiomics,
further
feature
reduction
was
carried
out
using
univariate
analysis,
correlation
least
absolute
shrinkage
selection
operator
(LASSO).
Finally,
we
built
five
models,
support
vector
machine,
logistic
regression,
linear
discrimination
KNeighbors
classifier,
Gaussian
NB
evaluated
their
performance
via
several
metrics
including
sensitivity
(recall),
specificity,
positive
predictive
value
(precision),
negative
value,
accuracy
area
under
curve
(AUC).
Results
The
(AUC)
SVC,
LR,
LDA,
KNN,
algorithms
(testing)
sets
0.99
(0.94),
1.00
(0.87),
0.98
(0.92),
0.97
(0.97),
(0.93)
respectively.
Nested
cross-validation
demonstrated
robustness
generalizability
models.
Additionally,
calibration
plot
decision
analysis
practical
usefulness
models
clinical
practice,
lower
bias
toward
subgroups
during
decision-making.
Conclusion
combination
approach
contrast
enhanced
could
well
distinguish
holds
promise
assisting
doctors
accurate
decision-making
