Journal of Clinical Hypertension,
Journal Year:
2024,
Volume and Issue:
26(2), P. 187 - 196
Published: Jan. 12, 2024
Abstract
There
are
limited
data
available
regarding
the
connection
between
hypertension
and
heavy
metal
exposure.
The
authors
intend
to
establish
an
interpretable
machine
learning
(ML)
model
with
high
efficiency
robustness
that
identifies
based
on
Our
datasets
were
obtained
from
US
National
Health
Nutrition
Examination
Survey
(NHANES,
2013–2020.3).
developed
5
ML
models
for
identification
by
exposure,
tested
them
10
discrimination
characteristics.
Further,
chose
optimally
performing
after
parameter
adjustment
Genetic
Algorithm
(GA)
identification.
Finally,
in
order
visualize
model's
ability
make
decisions,
used
SHapley
Additive
exPlanation
(SHAP)
Local
Interpretable
Model‐Agnostic
Explanations
(LIME)
algorithm
illustrate
features.
study
included
19
368
participants
total.
A
best‐performing
eXtreme
Gradient
Boosting
(XGB)
GA
16
metals
was
selected
(AUC:
0.774;
95%
CI:
0.772–0.776;
accuracy:
87.7%).
According
SHAP
values,
Barium
(0.02),
Cadmium
(0.017),
Lead
Antimony
(0.008),
Tin
(0.007),
Manganese
(0.006),
Thallium
(0.004),
Tungsten
(0.004)
urine,
(0.048),
Mercury
(0.035),
Selenium
(0.05),
(0.007)
blood
positively
influenced
model,
while
(−0.001)
urine
negatively
model.
Study
participants'
associated
exposure
identified
efficient,
robust,
GA‐XGB
LIME.
Barium,
Cadmium,
Lead,
Antimony,
Tin,
Manganese,
Thallium,
Mercury,
Selenium,
correlated
hypertension,
is
hypertension.
Foods,
Journal Year:
2024,
Volume and Issue:
13(7), P. 1116 - 1116
Published: April 6, 2024
Fruits
and
vegetables
are
an
important
part
of
our
daily
diet
contain
low-content
components
that
crucial
for
health.
Detecting
these
accurately
is
paramount
significance.
However,
traditional
detection
methods
face
challenges
such
as
complex
sample
processing,
slow
speed,
the
need
highly
skilled
operators.
These
limitations
fail
to
meet
growing
demand
intelligent
rapid
in
fruits
vegetables.
In
recent
years,
significant
progress
has
been
made
technology,
particularly
detecting
high-content
accurate
remains
a
challenge
gained
considerable
attention
current
research.
This
review
paper
aims
explore
analyze
several
techniques
have
extensively
studied
this
purpose.
include
near-infrared
spectroscopy,
Raman
laser-induced
breakdown
terahertz
among
others.
provides
detailed
reports
analyses
application
components.
Furthermore,
it
offers
prospective
exploration
their
future
development
field.
The
goal
contribute
enhancement
widespread
adoption
technology
It
expected
will
serve
valuable
reference
researchers
practitioners
area.
Toxics,
Journal Year:
2023,
Volume and Issue:
11(8), P. 659 - 659
Published: Aug. 1, 2023
Manganese
(Mn),
cadmium
(Cd)
and
lead
(Pb)
have
toxic
effects
on
the
immune
system.
However,
their
independent
combined
immune-inflammation
responses
are
unclear.
In
recent
years,
systemic
index
(SII)
has
been
developed
as
an
integrated
novel
inflammatory
indicator.
A
retrospective
cross-sectional
study
of
2174
adults
≥20
years
old
from
National
Health
Nutrition
Examination
Survey
(NHANES)
2015-2016
was
conducted.
Generalized
linear
models
were
used
to
evaluate
associations
SII
with
blood
Mn,
Cd
Pb
levels.
As
continuous
variables,
both
Mn
showed
dose-dependent
relationships
before
after
adjusting
for
all
potential
confounding
factors.
Metal
concentrations
then
converted
into
categorical
variables.
Compared
in
lowest
or
tertile,
those
highest
tertile
had
higher
risks
elevated
SII.
Furthermore,
co-exposure
also
a
positive
relationship
single
effect
exposure
joint
other
metal
exposures
not
observed.
This
provides
important
epidemiological
evidence
Cd,
Pb.
