Advances in computational intelligence and robotics book series,
Год журнала:
2025,
Номер
unknown, С. 67 - 92
Опубликована: Фев. 28, 2025
AI
biases
can
induce
existing
imbalances
and
affect
the
most
affected
populations
more
severely.
The
study
underlines
need
to
introduce
imperative
of
transparency
explainability
systems.
fact
that
many
algorithmic
systems
are
correspondence
opaque
raises
questions
about
how
such
decisions
made
who
is
accountable
when
using
artificial
intelligence,
which
leads
wrongful
arrest
or
unfair
sentencing.
research
calls
for
effective
legislative
frameworks
would
protect
constitutional
entitlement
due
widespread
use
criminal
justice
system
effectively
embrace
avoid
risk
infringing
individual
rights
make
technology
serve
rather
than
inimical
detrimental
basic
human
rights.
Exploratory Research in Clinical and Social Pharmacy,
Год журнала:
2023,
Номер
12, С. 100346 - 100346
Опубликована: Окт. 21, 2023
Artificial
intelligence
(AI)
is
a
transformative
technology
used
in
various
industrial
sectors
including
healthcare.
In
pharmacy
practice,
AI
has
the
potential
to
significantly
improve
medication
management
and
patient
care.
This
review
explores
applications
field
of
practice.
The
incorporation
technologies
provides
pharmacists
with
tools
systems
that
help
them
make
accurate
evidence-based
clinical
decisions.
By
using
algorithms
Machine
Learning,
can
analyze
large
volume
data,
medical
records,
laboratory
results,
profiles,
aiding
identifying
drug-drug
interactions,
assessing
safety
efficacy
medicines,
making
informed
recommendations
tailored
individual
requirements.
Various
models
have
been
developed
predict
detect
adverse
drug
events,
assist
decision
support
medication-related
decisions,
automate
dispensing
processes
community
pharmacies,
optimize
dosages,
adherence
through
smart
technologies,
prevent
errors,
provide
therapy
services,
telemedicine
initiatives.
incorporating
into
health
care
professionals
augment
their
decision-making
patients
personalized
allows
for
greater
collaboration
between
different
healthcare
services
provided
single
patient.
For
patients,
may
be
useful
tool
providing
guidance
on
how
when
take
medication,
education,
promoting
know
where
obtain
most
cost-effective
best
communicate
professionals,
monitoring
wearables
devices,
everyday
lifestyle
guidance,
integrate
diet
exercise.
Journal of the American Medical Informatics Association,
Год журнала:
2024,
Номер
31(5), С. 1172 - 1183
Опубликована: Март 23, 2024
Abstract
Objectives
Leveraging
artificial
intelligence
(AI)
in
conjunction
with
electronic
health
records
(EHRs)
holds
transformative
potential
to
improve
healthcare.
However,
addressing
bias
AI,
which
risks
worsening
healthcare
disparities,
cannot
be
overlooked.
This
study
reviews
methods
handle
various
biases
AI
models
developed
using
EHR
data.
Materials
and
Methods
We
conducted
a
systematic
review
following
the
Preferred
Reporting
Items
for
Systematic
Reviews
Meta-analyses
guidelines,
analyzing
articles
from
PubMed,
Web
of
Science,
IEEE
published
between
January
01,
2010
December
17,
2023.
The
identified
key
biases,
outlined
strategies
detecting
mitigating
throughout
model
development,
analyzed
metrics
assessment.
Results
Of
450
retrieved,
20
met
our
criteria,
revealing
6
major
types:
algorithmic,
confounding,
implicit,
measurement,
selection,
temporal.
were
primarily
predictive
tasks,
yet
none
have
been
deployed
real-world
settings.
Five
studies
concentrated
on
detection
implicit
algorithmic
employing
fairness
like
statistical
parity,
equal
opportunity,
equity.
Fifteen
proposed
especially
targeting
selection
biases.
These
strategies,
evaluated
through
both
performance
metrics,
predominantly
involved
data
collection
preprocessing
techniques
resampling
reweighting.
Discussion
highlights
evolving
mitigate
EHR-based
models,
emphasizing
urgent
need
standardized
detailed
reporting
methodologies
testing
evaluation.
Such
measures
are
essential
gauging
models’
practical
impact
fostering
ethical
that
ensures
equity
Artificial
intelligence
(AI)
algorithms
are
prone
to
bias
at
multiple
stages
of
model
development,
with
potential
for
exacerbating
health
disparities.
However,
in
imaging
AI
is
a
complex
topic
that
encompasses
coexisting
definitions.
Advances in human and social aspects of technology book series,
Год журнала:
2024,
Номер
unknown, С. 127 - 156
Опубликована: Окт. 17, 2024
In
an
era
where
AI
advancements
permeate
various
facets
of
daily
life,
ranging
from
healthcare
decision-making
to
personalized
content
delivery,
the
potential
for
biases
exacerbate
societal
inequalities
has
become
a
pressing
concern.
The
chapter
commences
by
defining
and
scrutinizing
forms
bias
in
artificial
intelligence,
elucidating
their
tangible
effects
through
compelling
case
studies.
Subsequently,
it
explores
theoretical
foundations
fairness
AI,
considering
conceptual
frameworks
such
as
distributive
justice
procedural
while
addressing
challenges
operationalizing
these
principles.
section
delves
into
methods
tools
identifying
measuring
datasets
algorithms,
introducing
metrics
benchmarks
assess
outcomes.
Strategies
best
practices
mitigating
are
examined,
encompassing
approaches
data
preprocessing,
algorithmic
adjustments,
post-hoc
corrections.
BMC Medical Ethics,
Год журнала:
2024,
Номер
25(1)
Опубликована: Май 16, 2024
Abstract
Background
Integrating
artificial
intelligence
(AI)
into
healthcare
has
raised
significant
ethical
concerns.
In
pharmacy
practice,
AI
offers
promising
advances
but
also
poses
challenges.
Methods
A
cross-sectional
study
was
conducted
in
countries
from
the
Middle
East
and
North
Africa
(MENA)
region
on
501
professionals.
12-item
online
questionnaire
assessed
concerns
related
to
adoption
of
practice.
Demographic
factors
associated
with
were
analyzed
via
SPSS
v.27
software
using
appropriate
statistical
tests.
Results
Participants
expressed
about
patient
data
privacy
(58.9%),
cybersecurity
threats
potential
job
displacement
(62.9%),
lack
legal
regulation
(67.0%).
Tech-savviness
basic
understanding
correlated
higher
concern
scores
(
p
<
0.001).
Ethical
implications
include
need
for
informed
consent,
beneficence,
justice,
transparency
use
AI.
Conclusion
The
findings
emphasize
importance
guidelines,
education,
autonomy
adopting
Collaboration,
privacy,
equitable
access
are
crucial
responsible
Cancers,
Год журнала:
2022,
Номер
14(12), С. 2897 - 2897
Опубликована: Июнь 12, 2022
Recent
technological
developments
have
led
to
an
increase
in
the
size
and
types
of
data
medical
field
derived
from
multiple
platforms
such
as
proteomic,
genomic,
imaging,
clinical
data.
Many
machine
learning
models
been
developed
support
precision/personalized
medicine
initiatives
computer-aided
detection,
diagnosis,
prognosis,
treatment
planning
by
using
large-scale
Bias
class
imbalance
represent
two
most
pressing
challenges
for
learning-based
problems,
particularly
(e.g.,
oncologic)
sets,
due
limitations
patient
numbers,
cost,
privacy,
security
sharing,
complexity
generated
Depending
on
set
research
question,
methods
applied
address
problems
can
provide
more
effective,
successful,
meaningful
results.
This
review
discusses
essential
strategies
addressing
mitigating
different
oncologic
domain.
