Development and Validation of MRI Radiomics Model for Predicting Perineural Invasion in Rectal Cancer
Abstract
Background
This
study
aims
to
explore
the
application
of
multiparametric
MRI
(mp-MRI)
based
radiomics
in
evaluating
perineural
invasion
(PNI)
status
rectal
cancer.
Methods
A
retrospective
analysis
was
conducted
on
clinical
and
data
from
423
cancer
patients
confirmed
by
surgical
pathology
across
two
centers.
total
343
Center
1
were
split
into
a
training
set
an
internal
validation
(in-vad)
8:2
ratio,
while
80
2
served
as
independent
external
(ex-vad)
set.
Univariate
multivariate
analyses
performed
features
construct
model.
Radiomic
extracted
using
Pyradiomics
software,
selected
reduced
mRMR
LASSO
methods
combined
model
integrating
subsequently
built,
nomogram
developed.
Results
Among
all
patients,
131
cases
(31.0%)
PNI-positive.
Multivariate
identified
mrT
(OR
=
1.038,
P
<
0.001)
mrN
predictors
PNI,
forming
After
radiomic
feature
selection,
30
used
build
The
area
under
curve
(AUC)
values
for
training,
in-vad,
ex-vad
sets
0.719,
0.631,
0.760,
respectively.
AUC
0.841,
0.815,
0.916,
those
0.899,
0.826,
0.914.
Delong
test
demonstrated
that
both
models
outperformed
datasets,
with
no
statistically
significant
difference
between
models.
Conclusions
mp-MRI
effectively
predicts
PNI
cancer,
providing
non-invasive
accurate
method
preoperative
evaluation.
Research Square (Research Square), Год журнала: 2025, Номер unknown
Опубликована: Янв. 31, 2025
Язык: Английский