Research Square (Research Square),
Год журнала:
2023,
Номер
unknown
Опубликована: Ноя. 20, 2023
Abstract
Background:
Artificial
intelligence
(AI)-based
medical
devices
and
digital
health
technologies
such
as
sensors,
wearable
trackers,
telemedicine,
mobile
(m)
Health,
large
language
models
(LLMs),
care
twins
(DCTs)
have
a
substantial
influence
on
the
process
of
clinical
decision
support
systems
(CDSS)
in
healthcare
medicine
application.
However,
given
complexity
decisions,
it
is
crucial
that
results
generated
by
AI
tools
not
only
deliver
accurate
but
are
also
evaluated
carefully,
will
be
understandable
explainable
to
end-users,
especially
clinicians.
The
absence
interpretability
realm
communicating
decisions
can
result
mistrust
decision-makers
fear
using
these
technologies.
Objective:
This
paper
presents
systematic
review
processes
challenges
related
interpretable
machine
learning
(IML)
artificial
(XAI)
within
domains.
main
objectives
this
IML
XAI,
methods,
applications,
their
implantation
context
DHIs,
particularly
with
quality
control
perspective
for
easy
understand
better
communicate
between
classified
into
three
parts:
pre-processing
interpretability,
modeling,
post-processing
interpretability.
intends
establish
comprehensive
understanding
importance
robust
approach
reviewing
experimental
results.
ultimate
aim
provide
future
researches
insights
creating
clinician-AI
more
communicable
systems,
well
offer
deeper
they
might
face.
Methods:
Our
research
questions,
eligibility
criteria
primary
goals
were
identified
Preferred
Reporting
Items
Systematic
reviews
Meta-Analyses
(PRISMA)
guideline
PICO
(population,
intervention,
control,
outcomes)
method,
PubMed,
Scopus
Web
Science
databases
systematically
searched
sensitive
specific
search
strings.
In
next
steps,
duplicate
papers
removed
EndNote
Covidence,
then
two-phase
selection
was
conducted
Covidence
via
title
abstract,
followed
full-text
appraisal.
Meta
appraisal
tool
(MetaQAT
tool)
used
risk
bias
assessment.
end,
standardized
data
extraction
reliable
mining.
Results:
searches
retrieved
2241
records;
555
removed.
At
abstract
screening
step,
958
excluded,
step
excluded
482
studies.
Then
assessment,
172
74
publications
selected
which
included
10
exciting
64
Conclusion:
offers
general
definitions
XAI
domain,
proposes
three-levels
discusses
XAI-related
applications
at
each
level
proposed
framework,
supported
Additionally,
provides
discussion
assessment
evaluating
intelligent
systems.
Moreover,
survey
introduces
step-by-step
roadmap
implementing
applications.
To
guide
addressing
existing
gaps,
delves
significance
from
various
perspectives
acknowledges
limitations.
Clinical and Translational Science,
Год журнала:
2025,
Номер
18(1)
Опубликована: Янв. 1, 2025
ABSTRACT
The
pharmaceutical
industry
constantly
strives
to
improve
drug
development
processes
reduce
costs,
increase
efficiencies,
and
enhance
therapeutic
outcomes
for
patients.
Model‐Informed
Drug
Development
(MIDD)
uses
mathematical
models
simulate
intricate
involved
in
absorption,
distribution,
metabolism,
excretion,
as
well
pharmacokinetics
pharmacodynamics.
Artificial
intelligence
(AI),
encompassing
techniques
such
machine
learning,
deep
Generative
AI,
offers
powerful
tools
algorithms
efficiently
identify
meaningful
patterns,
correlations,
drug–target
interactions
from
big
data,
enabling
more
accurate
predictions
novel
hypothesis
generation.
union
of
MIDD
with
AI
enables
researchers
optimize
candidate
selection,
dosage
regimens,
treatment
strategies
through
virtual
trials
help
derisk
candidates.
However,
several
challenges,
including
the
availability
relevant,
labeled,
high‐quality
datasets,
data
privacy
concerns,
model
interpretability,
algorithmic
bias,
must
be
carefully
managed.
Standardization
architectures,
formats,
validation
is
imperative
ensure
reliable
reproducible
results.
Moreover,
regulatory
agencies
have
recognized
need
adapt
their
guidelines
evaluate
recommendations
AI‐enhanced
methods.
In
conclusion,
integrating
model‐driven
a
transformative
paradigm
innovation.
By
predictive
power
computational
data‐driven
insights
synergy
between
these
approaches
has
potential
accelerate
discovery,
strategies,
usher
new
era
personalized
medicine,
benefiting
patients,
researchers,
whole.
Pharmaceuticals,
Год журнала:
2025,
Номер
18(3), С. 282 - 282
Опубликована: Фев. 20, 2025
Artificial
intelligence
(AI)
has
emerged
as
a
powerful
tool
in
medical
sciences
that
is
revolutionizing
various
fields
of
drug
research.
AI
algorithms
can
analyze
large-scale
biological
data
and
identify
molecular
targets
pathways
advancing
pharmacological
knowledge.
An
especially
promising
area
the
assessment
interactions.
The
analysis
large
datasets,
such
drugs’
chemical
structure,
properties,
pathways,
known
interaction
patterns,
provide
mechanistic
insights
potential
associations
by
integrating
all
this
complex
information
returning
risks
associated
with
these
In
context,
an
where
may
prove
valuable
underlying
mechanisms
interactions
natural
products
(i.e.,
herbs)
are
used
dietary
supplements.
These
pose
challenging
problem
since
they
mixtures
constituents
diverse
limited
regarding
their
pharmacokinetic
data.
As
use
herbal
supplements
continues
to
grow,
it
becomes
increasingly
important
understand
between
them
conventional
drugs
adverse
reactions.
This
review
will
discuss
approaches
how
be
exploited
providing
prediction
herbs,
exploitation
experimental
validation
or
clinical
utilization.
Computational Intelligence,
Год журнала:
2024,
Номер
40(3)
Опубликована: Июнь 1, 2024
Abstract
Personalized
health
monitoring
and
prediction
are
indispensable
in
advancing
healthcare
delivery,
particularly
amidst
the
escalating
prevalence
of
chronic
illnesses
aging
population.
Deep
learning
(DL)
stands
out
as
a
promising
avenue
for
crafting
personalized
systems
adept
at
forecasting
outcomes
with
precision
efficiency.
As
personal
data
becomes
increasingly
accessible,
DL‐based
methodologies
offer
compelling
strategy
enhancing
provision
through
accurate
timely
prognostications
conditions.
This
article
offers
comprehensive
examination
recent
advancements
employing
DL
prediction.
It
summarizes
diverse
range
architectures
their
practical
implementations
across
various
realms,
such
wearable
technologies,
electronic
records
(EHRs),
accumulated
from
social
media
platforms.
Moreover,
it
elucidates
obstacles
encountered
outlines
future
directions
leveraging
monitoring,
thereby
furnishing
invaluable
insights
into
immense
potential
this
domain.
Hepatitis
C
virus
(HCV)
infection
affects
over
71
million
people
worldwide,
leading
to
severe
liver
diseases,
including
cirrhosis
and
hepatocellular
carcinoma.
The
virus’s
high
mutation
rate
complicates
current
antiviral
therapies
by
promoting
drug
resistance,
emphasizing
the
need
for
novel
therapeutics.
Traditional
high-throughput
screening
(HTS)
methods
are
costly,
time-consuming,
prone
false
positives,
underscoring
necessity
more
efficient
alternatives.
Machine
learning
(ML),
particularly
quantitative
structure–activity
relationship
(QSAR)
modeling,
offers
a
promising
solution
predicting
compounds’
biological
activity
based
on
chemical
structures.
However,
“black-box”
nature
of
many
ML
models
raises
concerns
about
interpretability,
which
is
critical
understanding
action
mechanisms.
To
address
this,
we
propose
an
explainable
multi-model
stacked
classifier
(MMSC)
hepatitis
candidates.
Our
approach
combines
random
forests
(RF),
support
vector
machines
(SVM),
gradient
boosting
(GBM),
k-nearest
neighbors
(KNN)
using
logistic
regression
meta-learner.
Trained
tested
dataset
495
compounds
targeting
HCV
NS3
protease,
model
achieved
94.95%
accuracy,
97.40%
precision,
96.77%
F1-score.
Using
SHAP
values,
provided
interpretability
identifying
key
molecular
descriptors
influencing
model’s
predictions.
This
MMSC
improves
discovery,
bridging
gap
between
predictive
performance
while
offering
actionable
insights
researchers.
Drug Discovery Today,
Год журнала:
2025,
Номер
unknown, С. 104362 - 104362
Опубликована: Апрель 1, 2025
Artificial
intelligence
(AI)
and
machine
learning
(ML)
have
revolutionized
pharmaceutical
research,
particularly
in
protein
nucleic
acid
studies.
This
review
summarizes
the
current
status
of
AI
ML
applications
sector,
focusing
on
innovative
tools,
web
servers,
databases.
paper
highlights
how
these
technologies
address
key
challenges
drug
development
including
high
costs,
lengthy
timelines,
complexity
biological
systems.
Furthermore,
potential
personalized
medicine,
cancer
response
prediction,
biomarker
identification
is
discussed.
The
integration
research
promises
to
accelerate
discovery,
reduce
ultimately
lead
more
effective
therapeutic
strategies.
