Precision in Diagnosis of Liver Fibrosis in MASLD: Machine Learning‐Based Scores May Be More Accurate Than Conventional NITs
Liver International,
Год журнала:
2025,
Номер
45(4)
Опубликована: Март 3, 2025
Язык: Английский
Response to: Precision in Diagnosis of Liver Fibrosis in MASLD: Machine Learning Based Scores May Be More Accurate Than Conventional NITs
Liver International,
Год журнала:
2025,
Номер
45(5)
Опубликована: Март 29, 2025
We
thank
Soliman
et
al.
for
their
thoughtful
comments
on
our
recent
publication
[1].
appreciate
insights
and
the
opportunity
to
further
discuss
role
of
machine
learning
(ML)-based
non-invasive
tests
(NITs)
in
diagnosis
metabolic
dysfunction-associated
steatotic
liver
disease
(MASLD).
As
we
previously
demonstrated
LITMUS
study,
ML
models
incorporating
clinical
biomarker
data
can
enhance
detection
nonalcoholic
steatohepatitis
(NASH)
at-risk
NASH,
highlighting
potential
ML-based
approaches
diagnostics
[2].
Expanding
fibrosis
assessment,
highlight
scores,
particularly
FIB-6
index,
improving
accuracy
assessment
MASLD
[3].
Their
multicenter
study
that
score,
which
integrates
multiple
routine
laboratory
parameters,
could
offer
advantages
sensitivity
negative
predictive
value
(NPV)
compared
conventional
NITs
like
FIB-4
or
APRI
[4].
also
utility
resource-limited
settings,
where
advanced
imaging
techniques
transient
elastography
may
not
be
readily
available.
However,
generalisability
these
scores
across
diverse
populations
settings
remains
fully
validated.
While
has
been
evaluated
cohorts
patients
with
chronic
hepatitis
C,
B,
MASLD,
studies
are
needed
assess
its
performance
primary
care
comorbidities
beyond
those
studied.
In
addition,
while
such
as
hold
promise
refining
diagnostic
accuracy,
implementation
requires
careful
consideration.
Our
emphasised
importance
tailoring
NIT
thresholds
individual
patient
characteristics,
age,
body
mass
index
(BMI),
diabetes
status,
optimise
accuracy.
This
approach
aligns
principles
personalised
medicine
enhanced
by
models.
acknowledged
findings
primarily
reflect
secondary
tertiary
need
research
populations.
Moreover,
leveraged
rigorous
histological
assessments
from
centres,
variability
biopsy
interpretation
a
recognised
limitation.
The
upcoming
results
cohort,
using
centralised
AI-based
pathology,
will
provide
into
standardised
evaluation.
conclusion,
represent
promising
advancement
diagnosis,
integration
practice
should
guided
validation
consideration
patient-specific
factors.
look
forward
future
explore
other
diseases.
authors
declare
no
conflicts
interest.
Язык: Английский
Digital Pathology Tailored for Assessment of Liver Biopsies
Biomedicines,
Год журнала:
2025,
Номер
13(4), С. 846 - 846
Опубликована: Апрель 1, 2025
Improved
image
quality,
better
scanners,
innovative
software
technologies,
enhanced
computational
power,
superior
network
connectivity,
and
the
ease
of
virtual
reproduction
distribution
are
driving
potential
use
digital
pathology
for
diagnosis
education.
Although
relatively
common
in
clinical
oncology,
its
application
liver
is
under
development.
Digital
improving
subjective
histologic
scoring
systems
could
be
essential
managing
obesity-associated
steatotic
disease.
The
increasing
analyzing
specimens
particularly
intriguing
as
it
may
offer
a
more
detailed
view
biology
eliminate
incomplete
measurement
treatment
responses
trials.
objective
automated
quantification
histological
results
help
establish
standardized
diagnosis,
treatment,
assessment
protocols,
providing
foundation
personalized
patient
care.
Our
experience
with
artificial
intelligence
(AI)-based
enhances
reproducibility
accuracy,
enabling
continuous
detecting
subtle
changes
that
indicate
disease
progression
or
regression.
Ongoing
validation
highlights
need
collaboration
between
pathologists
AI
developers.
Concurrently,
analysis
can
address
issues
related
to
historical
failure
trials
stemming
from
challenges
assessment.
We
discuss
how
these
novel
tools
incorporated
into
research
complement
post-diagnosis
scenarios
where
necessary,
thus
clarifying
evolving
role
field.
Язык: Английский