Providing Optimal ST‐Segment–Elevation Myocardial Infarction Care: How Do We Overcome Barriers?
Journal of the American Heart Association,
Год журнала:
2025,
Номер
unknown
Опубликована: Фев. 19, 2025
Язык: Английский
Dynamic Predictive Models of Cardiogenic Shock in STEMI: Focus on Interventional and Critical Care Phases
Journal of Clinical Medicine,
Год журнала:
2025,
Номер
14(10), С. 3503 - 3503
Опубликована: Май 16, 2025
Background:
While
early
risk
stratification
in
STEMI
is
essential,
the
threat
of
cardiogenic
shock
(CS)
persists
after
revascularization
due
to
reperfusion
injury
and
evolving
instability.
However,
prediction
later
phases—after
revascularization—is
less
explored,
despite
its
importance
guiding
intensive
care
decisions.
This
study
evaluates
machine
learning
(ML)
models
for
dynamic
assessment
interventional
cardiology
cardiac
unit
(CICU)
phases,
where
timely
detection
deterioration
can
guide
treatment
escalation.
Methods:
We
retrospectively
analyzed
clinical
procedural
data
from
158
patients
diagnosed
with
complicated
by
shock,
treated
between
2019
2022
at
Cardiology
Department
University
Emergency
Hospital
Bucharest,
Romania.
Machine
models—Random
Forest
(RF),
Quadratic
Discriminant
Analysis
(QDA)—were
developed
tested
specifically
CICU
phases.
Model
performance
was
evaluated
using
area
under
receiver
operating
characteristic
curve
(ROC-AUC),
accuracy
(ACC),
sensitivity,
specificity,
F1-score.
Results:
In
phase,
RF
QDA
achieved
highest
accuracy,
both
reaching
87.50%.
CICU,
demonstrate
best
performance,
ACCs
0.843.
maintained
consistent
across
Relevant
predictors
included
strategy,
TIMI
flow
before
percutaneous
coronary
intervention
(PCI),
Killip
class,
creatinine,
Creatine
Kinase
Index
(CKI)—all
parameters
routinely
assessed
patients.
These
effectively
identified
post-reperfusion
complications
hemodynamic
decline,
supporting
decisions
regarding
extended
monitoring
ICU-level
care.
Conclusions:
Predictive
implemented
advanced
phases
contribute
dynamic,
phase-specific
reassessment
optimize
resource
allocation.
findings
support
integration
ML-based
tools
into
post-PCI
workflows,
enabling
earlier
decline
more
efficient
deployment
resources.
When
combined
earlier-stage
models,
inclusion
forms
a
end-to-end
framework.
With
further
refinement,
this
system
could
be
as
mobile
application
throughout
continuum.
Язык: Английский
AI-Based Predictive Models for Cardiogenic Shock in STEMI: Real-World Data for Early Risk Assessment and Prognostic Insights
Journal of Clinical Medicine,
Год журнала:
2025,
Номер
14(11), С. 3698 - 3698
Опубликована: Май 25, 2025
Background:
Cardiogenic
shock
(CS)
is
a
life-threatening
complication
of
ST-elevation
myocardial
infarction
(STEMI)
and
remains
the
leading
cause
in-hospital
mortality,
with
rates
ranging
from
5
to
10%
despite
advances
in
reperfusion
strategies.
Early
identification
timely
intervention
are
critical
for
improving
outcomes.
This
study
investigates
utility
machine
learning
(ML)
models
predicting
risk
CS
during
early
phases
care—prehospital,
emergency
department
(ED),
cardiology-on-call—with
focus
on
accurate
triage
prioritization
urgent
angiography.
Results:
In
prehospital
phase,
Extra
Trees
classifier
demonstrated
highest
overall
performance.
It
achieved
an
accuracy
(ACC)
0.9062,
precision
0.9078,
recall
F1-score
0.9061,
Matthews
correlation
coefficient
(MCC)
0.8140,
indicating
both
high
predictive
power
strong
generalization.
ED
support
vector
model
outperformed
others
ACC
78.12%.
During
cardiology-on-call
Random
Forest
showed
best
performance
81.25%
consistent
values
across
other
metrics.
Quadratic
discriminant
analysis
generalizable
all
care
stages.
Key
features
included
Killip
class,
ECG
rhythm,
creatinine,
potassium,
markers
renal
dysfunction—parameters
readily
available
routine
settings.
The
greatest
clinical
was
observed
phases,
where
ML
could
critically
ill
patients
prioritize
coronary
catheterization,
especially
important
centers
limited
capacity
Conclusions:
Machine
learning-based
offer
valuable
tool
stratification
STEMI
at
cardiogenic
shock.
These
findings
implementation
ML-driven
tools
pathways,
potentially
survival
through
faster
more
decision-making,
time-sensitive
environments.
Язык: Английский
The role of resting-state perfusion CMR in the evaluation of microvascular obstruction in patients with acute myocardial infarction: A clinical perspective
European Journal of Radiology Open,
Год журнала:
2025,
Номер
14, С. 100662 - 100662
Опубликована: Май 31, 2025
Язык: Английский