I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
AI
)
V
ol
.
15
, N
o.
1
,
F
e
br
ua
r
y
2026
, pp.
191
~
212
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
15
.i
1
.pp
191
-
212
191
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
E
xp
l
ai
n
ab
l
e
h
y
b
r
i
d
m
o
d
e
l
s f
or
c
ar
d
i
ovasc
u
l
ar
d
i
se
ase
d
e
t
e
c
t
i
on
an
d
m
or
t
al
i
t
y p
r
e
d
i
c
t
i
on
A
li
A
l
-
A
t
ab
y, H
u
s
s
ai
n
A
t
t
ia
D
e
pa
r
t
m
e
nt
of
E
l
e
c
t
r
i
c
a
l
a
nd E
l
e
c
t
r
oni
c
s
E
ngi
ne
e
r
i
ng, S
c
hool
o
f
E
ngi
ne
e
r
i
ng a
nd C
om
put
i
ng, A
m
e
r
i
c
a
n U
ni
ve
r
s
i
t
y of
R
a
s
A
l
K
ha
i
m
a
h,
R
a
s
A
l
K
ha
i
m
a
h, U
ni
t
e
d
A
r
a
b E
m
i
r
a
t
e
s
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
S
e
p
1
,
2025
R
e
vi
s
e
d
D
e
c
21
,
2025
A
c
c
e
pt
e
d
J
a
n
10
,
2026
The
impact
of
cardiovascular
diseases
(CVDs)
is
devastating,
with
20.5
million
deaths
annually.
Early
detection
and
prediction
tools
exi
st,
but
current
approaches
struggl
e
to
balance
predicti
ve
performance
with
c
linical
interpreta
bility.
In
this
work,
a
two
-
stage
machine
learning
(ML)
fram
ework
is
proposed
for
heart
disease
detection
and
mortality
pr
ediction
in
heart
failure
patients.
Logisti
c
r
egression
(LR),
random
forest
(RF),
and
g
radient
boosting
(GB)
models
were
trained
using
the
publicly
available
he
art
failure
datasets,
and
their
performance
was
compared,
then
a
stacked
en
semble
approach
was
employed
to
enhance
predicti
on
accuracy.
Model
interpreta
bility
was
achieve
d
through
Shapley
additive
expla
nations
(SHAP),
which
provides
global
feature
rankings
and
specific
patient
attribut
es,
support
ing
explainab
le
artificial
intell
igence
(XAI)
in
c
linical
practice.
The
GB
model
achieved
the
highest
performance
in
the
firs
t
stage
with
a
receiver
operating
characteristic
area
under
the
curve
(ROC
A
UC)
of
96%
and
an
accuracy
of
89%
on
internal
testing,
while
external
vali
dation
confirmed
strong
generalizat
ion
(ROC
AUC
of
94%).
In
the
second
stage,
stacked
ensemble
model
was
employed
and
achieved
m
arginal
improvements.
Two
interactiv
e
web
applications
were
developed
to
enable
real
-
time
predictions
with
SHAP
visualizations.
The
results
demonstra
te
that
combini
ng
high
-
performance
ML
models
with
interpretable
outputs
can
significantly improve tr
ust in real
-
world healthcare environment
s.
K
e
y
w
o
r
d
s
:
C
a
r
di
ova
s
c
ul
a
r
di
s
e
a
s
e
E
ns
e
m
bl
e
l
e
a
r
ni
ng
E
xp
l
a
i
n
a
b
l
e
a
r
ti
f
i
c
i
a
l
in
t
e
l
li
g
e
n
c
e
H
e
a
r
t
f
a
il
ur
e
M
a
c
hi
ne
l
e
a
r
ni
ng
M
or
ta
li
ty
pr
e
di
c
ti
on
S
ha
pl
e
y a
ddi
ti
ve
e
xpl
a
na
ti
ons
This is an
open
acce
ss artic
le unde
r the
CC BY
-
SA
license.
C
or
r
e
s
pon
di
n
g A
u
th
or
:
A
li
A
l
-
A
ta
by
D
e
pa
r
tm
e
nt
of
E
le
c
tr
ic
a
l
a
nd E
le
c
tr
oni
c
s
E
ngi
ne
e
r
in
g, S
c
hool
o
f
E
ngi
ne
e
r
in
g a
nd C
om
put
in
g
A
m
e
r
ic
a
n U
ni
ve
r
s
it
y of
R
a
s
A
l
K
ha
im
a
h
R
a
s
A
l
K
ha
im
a
h, U
ni
t
e
d A
r
a
b E
m
ir
a
te
s
E
m
a
il
:
a
li
.a
ta
by@
a
ur
a
k.a
c
.a
e
1.
I
N
T
R
O
D
U
C
T
I
O
N
W
it
h
a
n
e
s
ti
m
a
te
d
20.5
m
il
li
on
de
a
th
s
a
nnua
ll
y,
or
ne
a
r
ly
33%
o
f
a
ll
de
a
th
s
w
or
ld
w
id
e
,
c
a
r
di
ova
s
c
ul
a
r
di
s
e
a
s
e
s
(
C
V
D
s
)
c
ont
in
ue
to
b
e
th
e
le
a
di
ng
c
a
us
e
of
d
e
a
th
[
1]
.
D
e
s
pi
te
im
pr
ove
m
e
nt
s
in
m
e
di
c
a
l
tr
e
a
tm
e
nt
s
,
he
a
r
t
f
a
il
ur
e
is
of
te
n
m
is
s
e
d,
w
hi
c
h
l
e
a
d
s
to
a
hi
gh
r
is
k
of
pr
e
m
a
tu
r
e
de
a
th
.
T
im
e
ly
tr
e
a
tm
e
nt
pl
a
nni
ng
de
pe
nds
on
th
e
e
a
r
ly
de
te
c
ti
on
of
C
V
D
s
a
nd
th
e
pr
e
c
is
e
pr
e
di
c
ti
on
of
pa
ti
e
nt
out
c
om
e
s
.
H
ow
e
ve
r
,
it
is
s
ti
ll
di
f
f
ic
ul
t
to
a
c
hi
e
ve
bot
h
hi
gh
pr
e
di
c
ti
v
e
a
c
c
ur
a
c
y
a
nd
c
li
ni
c
a
l
in
te
r
pr
e
ta
bi
li
ty
[
2]
.
T
r
a
di
ti
ona
l
s
ta
ti
s
ti
c
a
l
m
ode
ls
s
uc
h
a
s
lo
gi
s
ti
c
r
e
gr
e
s
s
io
n
(
L
R
)
pr
ovi
de
e
a
s
y
in
te
r
pr
e
ta
bi
li
ty
but
la
c
k
th
e
pr
e
di
c
ti
ve
pow
e
r
of
a
dva
nc
e
d
e
ns
e
m
bl
e
-
ba
s
e
d
m
a
c
hi
ne
le
a
r
ni
ng
(
M
L
)
a
lg
or
it
h
m
s
[
3]
,
[
4
]
.
C
om
pl
e
x
m
ode
ls
li
ke
r
a
ndom
f
or
e
s
ts
(
R
F
)
a
nd
gr
a
di
e
nt
boos
ti
ng
(
G
B
)
te
nd
to
a
c
t
a
s
“
bl
a
c
k
box
e
s
”
,
w
hi
c
h
li
m
it
s
th
e
ir
c
li
ni
c
a
l
a
c
c
e
pt
a
nc
e
.
E
xpl
a
in
a
bi
li
ty
a
ppr
oa
c
he
s
,
s
u
c
h
a
s
S
ha
pl
e
y
a
d
di
ti
ve
e
xpl
a
na
ti
ons
(
S
H
A
P
)
,
w
hi
c
h
pr
ovi
de
tr
a
ns
pa
r
e
nt
,
pa
ti
e
nt
-
s
pe
c
if
ic
f
e
a
tu
r
e
a
tt
r
ib
ut
io
ns
,
h
a
ve
be
e
n
in
t
r
oduc
e
d
r
e
c
e
nt
ly
in
in
te
r
pr
e
ta
bl
e
M
L
to
c
lo
s
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
15
, N
o.
1
,
F
e
br
ua
r
y
20
26
:
191
-
212
192
th
is
ga
p. W
it
h t
he
s
e
a
ppr
oa
c
he
s
, pr
e
di
c
ti
ve
m
ode
ls
c
a
n pr
ovi
de
c
li
ni
c
ia
ns
w
it
h pr
e
c
is
e
i
nf
or
m
a
ti
on a
nd us
e
f
ul
in
s
ig
ht
s
, w
hi
c
h r
e
s
ul
t
in
i
m
pr
ovi
ng r
is
k a
s
s
e
s
s
m
e
nt
a
nd pe
r
s
ona
li
z
e
d c
a
r
e
[
4]
.
T
hi
s
s
tu
dy
pr
opos
e
s
a
two
-
s
ta
ge
M
L
f
r
a
m
e
w
or
k.
T
he
f
ir
s
t
s
ta
g
e
f
oc
us
e
s
on
de
te
c
ti
ng
he
a
r
t
di
s
e
a
s
e
,
w
hi
le
th
e
s
e
c
ond
pr
e
di
c
ts
m
or
ta
li
ty
in
he
a
r
t
f
a
il
ur
e
pa
ti
e
nt
s
.
A
va
r
ie
ty
of
m
ode
ls
,
in
c
lu
di
ng
L
R
,
R
F
,
a
nd
G
B
,
a
r
e
c
om
pa
r
e
d,
a
nd
a
s
t
a
c
ke
d
e
ns
e
m
bl
e
is
de
v
e
lo
pe
d
to
m
a
xi
m
iz
e
pe
r
f
or
m
a
nc
e
.
T
o
e
ns
ur
e
c
li
ni
c
a
l
tr
us
t,
th
e
f
r
a
m
e
w
or
k
in
te
gr
a
te
s
S
H
A
P
a
na
ly
s
is
f
or
bot
h
gl
oba
l
a
nd
in
di
v
id
ua
l
le
ve
l
in
te
r
pr
e
ta
bi
li
ty
.
M
or
e
ove
r
,
th
e
be
s
t
pe
r
f
or
m
in
g
m
ode
ls
w
il
l
be
de
pl
oye
d
in
to
a
us
e
r
-
f
r
ie
ndl
y
w
e
b
a
ppl
ic
a
ti
on
to
pr
ovi
de
c
li
ni
c
ia
ns
w
it
h
a
n
in
te
r
f
a
c
e
f
or
r
e
a
l
-
ti
m
e
r
is
k
a
s
s
e
s
s
m
e
nt
a
nd
de
c
i
s
io
n
s
uppor
t.
A
c
c
or
di
ngl
y,
th
e
a
im
of
th
is
w
or
k
is
to
de
li
ve
r
a
r
obus
t,
in
te
r
pr
e
ta
bl
e
,
a
nd
de
pl
oya
bl
e
s
ol
ut
io
n
th
a
t
s
uppor
ts
bot
h
th
e
e
a
r
ly
di
a
gnos
i
s
of
he
a
r
t
di
s
e
a
s
e
a
nd
th
e
pr
oa
c
ti
ve
m
a
na
ge
m
e
nt
of
hi
gh
-
r
is
k pa
ti
e
nt
s
.
T
he
m
a
in
c
ont
r
ib
ut
io
ns
of
t
hi
s
w
or
k a
r
e
:
‒
T
he
de
ve
lo
pm
e
nt
of
a
t
w
o
-
s
ta
g
e
M
L
pi
pe
li
ne
f
or
he
a
r
t
di
s
e
a
s
e
de
te
c
ti
on a
nd mor
ta
li
ty
pr
e
di
c
ti
on.
‒
T
he
us
e
a
s
ta
c
ke
d e
ns
e
m
bl
e
a
ppr
oa
c
h t
o i
m
pr
ove
m
ode
l
pr
e
di
c
t
io
n pe
r
f
or
m
a
nc
e
.
‒
T
he
us
e
of
S
H
A
P
i
nt
e
r
pr
e
ta
ti
on t
o e
nha
nc
e
d
e
c
is
io
n t
r
a
ns
pa
r
e
n
c
y.
‒
T
he
de
pl
oym
e
nt
of
t
he
be
s
t
p
e
r
f
or
m
in
g m
ode
ls
i
n i
nt
e
r
a
c
ti
ve
w
e
b a
ppl
ic
a
ti
ons
f
or
r
e
a
l
-
w
or
ld
us
a
bi
li
ty
.
T
he
r
e
s
t
of
t
hi
s
pa
pe
r
i
s
or
ga
ni
z
e
d a
s
f
ol
lo
w
s
. S
e
c
ti
on 2 pr
ovi
de
s
a
s
um
m
a
r
y of
l
it
e
r
a
tu
r
e
w
or
k a
bout
th
e
us
e
of
M
L
in
C
V
D
de
te
c
ti
on.
S
e
c
ti
on
3
pr
ovi
de
s
th
e
m
e
th
od
f
ol
lo
w
e
d
in
th
is
w
or
k,
in
c
lu
di
ng
da
ta
s
e
t
a
na
ly
s
is
a
nd
pr
e
pr
oc
e
s
s
in
g,
th
e
de
ve
lo
pe
d
m
ode
l
s
,
a
nd
th
e
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
.
S
e
c
ti
on
4
pr
ovi
de
s
de
ta
il
s
a
bout
th
e
im
pl
e
m
e
nt
a
ti
on
of
th
e
m
od
e
ls
.
S
e
c
ti
on
5
pr
ovi
de
s
th
e
obt
a
in
e
d
r
e
s
ul
ts
w
it
h
a
di
s
c
us
s
io
n
a
bout
pe
r
f
or
m
a
nc
e
.
F
in
a
ll
y,
s
e
c
ti
on
6
c
on
c
lu
de
s
th
e
p
a
pe
r
w
it
h
a
s
u
m
m
a
r
y
of
th
e
m
a
in
c
ont
r
ib
ut
io
ns
,
li
m
it
a
ti
ons
,
a
nd f
ut
ur
e
w
or
k.
2.
L
I
T
E
R
A
T
U
R
E
R
E
V
I
E
W
A
n
um
be
r
of
r
e
f
e
r
e
nc
e
s
w
e
r
e
u
s
e
d
in
th
e
r
e
vi
e
w
of
r
e
c
e
n
t
l
it
e
r
a
tu
r
e
r
e
l
a
t
e
d
to
t
he
u
s
e
of
M
L
in
h
e
a
r
t
di
s
e
a
s
e
de
te
c
t
io
n
a
n
d
m
or
t
a
li
t
y
pr
e
d
ic
ti
o
n.
T
he
s
um
m
a
r
y
of
th
i
s
r
e
vi
e
w
i
s
g
iv
e
n
i
n
T
a
bl
e
1
(
s
e
e
in
A
p
pe
ndi
x)
[
2]
,
[
3]
,
[
5]
–
[
2
2]
.
T
h
e
t
a
b
le
pr
ov
id
e
s
th
e
w
or
k
c
a
r
r
i
e
d
o
ut
i
n
e
a
c
h
r
e
f
e
r
e
n
c
e
a
l
on
g
w
i
th
t
he
ga
p.
R
e
c
e
nt
e
xp
la
in
a
bl
e
a
n
d
h
ybr
id
A
I
s
t
udi
e
s
ha
ve
e
xc
e
ll
e
d
a
t
C
V
D
p
r
e
d
ic
t
io
n
by
e
m
ph
a
s
i
z
i
ng
in
t
e
r
pr
e
t
a
b
le
c
li
n
ic
a
ll
y
r
e
li
a
b
le
m
od
e
l
s
t
h
a
t
ba
la
nc
e
pr
e
di
c
t
iv
e
a
c
c
ur
a
c
y
w
it
h
tr
a
n
s
pa
r
e
n
c
y.
S
our
o
v
e
t
a
l.
[
23]
i
nt
r
o
du
c
e
d a
n e
xp
la
in
a
bl
e
AI
-
e
nh
a
n
c
e
d
f
r
a
m
e
w
or
k
f
or
C
V
D
de
te
c
t
io
n
a
n
d
r
i
s
k
a
s
s
e
s
s
m
e
n
t,
d
e
m
o
n
s
tr
a
ti
ng
how
m
o
de
l
e
xpl
a
i
na
bi
li
ty
c
a
n
c
o
e
xi
s
t
w
i
th
hi
g
h
p
e
r
f
or
m
a
n
c
e
.
N
a
p
a
e
t
al
.
[
24]
c
ond
u
c
te
d
a
c
o
m
pa
r
a
t
iv
e
a
n
a
ly
s
i
s
of
e
xp
la
in
a
bl
e
m
od
e
l
s
us
in
g
S
H
A
P
a
nd
hi
g
hl
i
gh
te
d
h
ow
m
o
de
l
tr
a
n
s
p
a
r
e
n
c
y
a
i
d
s
i
n
c
a
r
di
ov
a
s
c
ul
a
r
r
i
s
k
d
e
t
e
r
m
in
a
ti
on
a
n
d
f
e
a
tu
r
e
in
t
e
r
pr
e
t
a
ti
on
.
S
im
i
la
r
ly
,
B
i
la
l
e
t
al
.
[
25]
d
e
v
e
l
op
e
d
a
n
e
xp
la
in
a
bl
e
A
I
s
y
s
t
e
m
f
or
a
c
c
ur
a
te
pr
e
di
c
ti
o
n
of
C
V
D
a
nd
e
m
p
ha
s
i
z
e
d
th
e
i
m
p
or
t
a
n
c
e
of
e
xpl
a
i
na
bi
l
it
y
f
or
c
l
in
i
c
a
l
a
do
pt
io
n
of
A
I
t
ool
s
i
n
h
e
a
lt
h
c
a
r
e
.
D
e
s
pi
te
th
e
a
v
a
il
a
bl
e
l
it
e
r
a
t
ur
e
a
bout
th
e
pot
e
nt
i
a
l
of
M
L
m
o
de
ls
(
e
.g.,
R
F
,
X
G
B
oo
s
t,
s
up
por
t
v
e
c
to
r
m
a
c
hi
n
e
(
S
V
M
)
,
a
nd
d
e
e
p
n
e
ur
a
l
n
e
twor
k
(
D
N
N
)
)
f
or
a
c
c
ur
a
t
e
pr
e
di
c
ti
on
of
C
V
D
s
,
m
a
ny
g
a
p
s
e
xi
s
t.
M
a
ny
s
tu
di
e
s
ha
ve
pr
io
r
it
i
z
e
d
a
c
c
ur
a
c
y
ov
e
r
m
o
de
l
in
t
e
r
pr
e
ta
bi
li
ty
a
n
d
c
li
ni
c
a
l
u
s
a
b
il
it
y,
w
hi
c
h
a
r
e
e
s
s
e
nt
i
a
l
f
or
r
e
a
l
-
w
or
ld
a
do
pt
io
n.
A
ls
o
,
c
om
p
a
r
a
ti
ve
s
tu
di
e
s
of
t
e
n
u
s
e
in
c
on
s
is
te
nt
da
ta
s
e
ts
,
l
a
c
k
e
xt
e
r
na
l
v
a
li
da
t
io
n,
a
nd
r
e
por
t
s
upe
r
i
or
pe
r
f
or
m
a
nc
e
on
s
m
a
ll
,
i
m
ba
l
a
nc
e
d
d
a
ta
s
e
ts
,
w
hi
c
h
l
im
it
s
g
e
ne
r
a
li
z
a
bi
li
t
y.
F
ur
th
e
r
m
or
e
,
a
lt
h
ough
e
xpl
a
i
na
bl
e
A
I
a
ppr
oa
c
h
e
s
h
a
ve
be
e
n
pr
opo
s
e
d
to
e
nha
n
c
e
tr
us
t
a
nd
tr
a
ns
p
a
r
e
n
c
y,
th
e
in
t
e
gr
a
ti
on
in
to
C
V
D
s
pr
e
di
c
ti
on
w
or
kf
lo
w
s
r
e
qui
r
e
s
m
or
e
in
v
e
s
ti
ga
ti
on. T
hi
s
s
tu
d
y
a
ddr
e
s
s
e
s
a
num
be
r
of
th
e
s
e
g
a
p
s
by
de
v
e
lo
pi
ng
a
r
obu
s
t
a
n
d
in
t
e
r
pr
e
ta
bl
e
h
ybr
id
M
L
m
od
e
l
f
or
he
a
r
t
di
s
e
a
s
e
a
nd
m
or
ta
li
t
y
pr
e
d
ic
ti
o
n
by
c
om
bi
ni
ng
c
l
a
s
s
i
c
a
l
a
nd e
n
s
e
m
bl
e
t
e
c
h
ni
qu
e
s
w
it
h S
H
A
P
e
xpl
a
in
a
bi
li
ty
.
T
hi
s
c
ont
r
i
but
e
s
t
o pr
e
d
ic
ti
v
e
a
c
c
ur
a
c
y
a
nd
a
ls
o t
o c
li
n
ic
a
l
tr
a
ns
p
a
r
e
n
c
y
a
nd t
r
u
s
t,
w
hi
c
h a
r
e
c
r
i
ti
c
a
l
f
a
c
to
r
s
f
or
e
th
ic
a
l
a
n
d
pr
a
c
ti
c
a
l
de
pl
o
ym
e
nt
i
n h
e
a
lt
hc
a
r
e
c
li
ni
c
s
.
3.
M
E
T
H
O
D
T
h
i
s
s
e
c
ti
on
pr
o
vi
d
e
s
m
e
th
od
f
o
ll
o
w
e
d
f
or
d
e
v
e
l
op
in
g
, t
r
a
in
in
g
,
e
v
a
l
ua
ti
n
g,
a
nd
i
nt
e
r
pr
e
t
i
ng
M
L
m
od
e
l
s
f
or
h
e
a
r
t
d
i
s
e
a
s
e
a
n
d
m
or
t
a
li
t
y
pr
e
di
c
t
i
on
.
T
h
e
w
or
k
c
o
n
s
i
s
t
s
o
f
d
a
ta
s
e
t
s
e
l
e
c
ti
o
n,
pr
e
p
r
o
c
e
s
s
i
n
g,
a
n
d
f
e
a
t
ur
e
e
n
g
in
e
e
r
in
g
.
I
t
a
l
s
o
i
n
c
l
u
de
s
m
o
d
e
l
d
e
v
e
l
op
m
e
n
t,
m
o
de
l
e
v
a
l
u
a
t
io
n
,
a
n
d
m
od
e
l
in
te
r
p
r
e
t
a
b
il
it
y
u
s
in
g
S
H
A
P
.
3.1. Dat
as
e
t
d
e
s
c
r
ip
t
io
n
an
d
d
at
a p
r
e
p
r
oc
e
s
s
in
g
3.1.1.
H
e
ar
t
d
is
e
as
e
p
r
e
d
ic
t
io
n
d
at
as
e
t
A
K
a
ggl
e
da
ta
s
e
t
w
it
h
12
a
tt
r
ib
ut
e
s
,
in
c
lu
di
ng
a
ge
,
s
e
x,
c
hol
e
s
te
r
ol
,
c
he
s
t
pa
in
,
a
nd
E
C
G
f
in
di
ngs
,
w
a
s
u
s
e
d
[
26]
.
T
hi
s
d
a
ta
s
e
t
c
ont
a
in
s
918
e
nt
r
ie
s
a
nd
12
c
ol
um
ns
.
T
h
e
f
ol
lo
w
in
g
poi
nt
s
s
um
m
a
r
iz
e
im
por
ta
nt
obs
e
r
va
ti
ons
f
r
om
t
he
da
ta
s
e
t:
‒
N
o m
is
s
in
g va
lu
e
s
w
e
r
e
f
ound.
‒
T
h
e
r
e
a
r
e
c
a
t
e
gor
i
c
a
l
f
e
a
tu
r
e
s
s
uc
h
a
s
S
e
x,
C
h
e
s
tP
a
in
T
yp
e
,
R
e
s
ti
ngE
C
G
,
E
x
e
r
c
i
s
e
A
n
gi
n
a
,
a
nd
S
T
_S
lo
p
e
.
‒
T
he
r
e
a
r
e
num
e
r
ic
a
l
a
nom
a
li
e
s
,
w
h
e
r
e
R
e
s
ti
ngB
P
a
nd
C
hol
e
s
te
r
ol
ha
ve
a
m
in
im
um
of
0,
w
hi
c
h
m
a
y
in
di
c
a
te
m
is
s
in
g or
e
r
r
one
ous
va
lu
e
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
E
x
pl
ai
nabl
e
hy
br
id
m
ode
ls
f
or
c
ar
di
ov
a
s
c
ul
ar
di
s
e
as
e
d
e
te
c
ti
o
n and mor
ta
li
ty
pr
e
di
c
ti
on
(
A
li
A
l
-
A
ta
by
)
193
‒
F
a
s
ti
ngB
S
c
a
n be
t
r
e
a
te
d a
s
c
a
t
e
gor
ic
a
l
(
0/
1)
.
‒
T
a
r
ge
t
c
la
s
s
(
H
e
a
r
tDi
s
e
a
s
e
)
i
s
f
a
ir
ly
ba
la
nc
e
d (
~
55%
pos
it
iv
e
c
a
s
e
s
)
.
T
a
b
l
e
2
s
ho
w
s
a
s
u
m
m
a
r
y
of
t
he
s
p
e
c
if
i
c
a
ti
on
s
of
t
hi
s
d
a
ta
s
e
t.
T
h
e
pr
o
p
o
s
e
d
pr
e
pr
oc
e
s
s
in
g
in
v
ol
v
e
s
h
a
n
dl
in
g
m
i
s
s
in
g
v
a
lu
e
s
,
ou
tl
i
e
r
r
e
m
ov
a
l,
e
n
c
od
in
g
c
a
te
g
or
i
c
a
l
v
a
r
i
a
b
l
e
s
,
a
n
d
f
e
a
tu
r
e
s
c
a
l
in
g.
R
e
c
or
d
s
w
it
h
C
h
o
le
s
t
e
r
ol
=
0
or
R
e
s
t
in
g
B
P
=
0
w
e
r
e
d
r
o
pp
e
d.
A
c
c
or
di
n
gl
y,
t
h
e
r
e
m
a
in
i
ng
r
e
c
o
r
d
s
a
r
e
7
46
s
a
m
pl
e
s
a
f
t
e
r
c
l
e
a
ni
ng
.
T
a
bl
e
2. H
e
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on da
ta
s
e
t
s
pe
c
if
ic
a
ti
ons
C
ol
um
n
D
e
s
c
r
i
pt
i
on
A
ge
A
ge
of
t
he
pa
t
i
e
nt
S
e
x
B
i
ol
ogi
c
a
l
s
e
x (
M
/
F
)
C
he
s
t
P
a
i
nT
ype
T
ype
of
c
he
s
t
pa
i
n (
e
.g., A
T
A
, A
S
Y
, N
A
P
, T
A
)
R
e
s
t
i
ngB
P
R
e
s
t
i
ng bl
ood pr
e
s
s
ur
e
(
0
–
200 m
m
H
g)
C
hol
e
s
t
e
r
ol
S
e
r
um
c
hol
e
s
t
e
r
ol
i
n m
g/
dl
(
0
-
603
)
F
a
s
t
i
ngB
S
F
a
s
t
i
ng bl
ood s
uga
r
>120 mg/
dl
(
1=
t
r
ue
, 0=
f
a
l
s
e
)
R
e
s
t
i
ngE
C
G
E
C
G
r
e
s
ul
t
s
(
N
or
m
a
l
, S
T
, L
V
H
)
M
a
xH
R
M
a
xi
m
um
he
a
r
t
r
a
t
e
E
xe
r
c
i
s
e
A
ngi
na
E
xe
r
c
i
s
e
-
i
nduc
e
d a
ngi
na
(
Y
/
N
)
O
l
dpe
a
k
S
T
de
pr
e
s
s
i
on i
nduc
e
d by e
xe
r
c
i
s
e
S
T
_S
l
ope
S
l
ope
of
t
he
pe
a
k e
xe
r
c
i
s
e
S
T
s
e
gm
e
nt
H
e
a
r
t
D
i
s
e
a
s
e
T
a
r
ge
t
va
r
i
a
bl
e
(
1=
ye
s
, 0=
no)
3.1.2.
M
or
t
al
it
y p
r
e
d
ic
t
io
n
d
at
as
e
t
T
he
he
a
r
t
f
a
il
ur
e
c
li
ni
c
a
l
r
e
c
or
ds
d
a
ta
s
e
t
f
r
om
[
27]
w
a
s
us
e
d.
I
t
c
ons
is
t
s
of
299
r
e
c
or
d
s
a
nd
13
f
e
a
tu
r
e
s
,
in
c
lu
di
ng
s
e
r
um
c
r
e
a
ti
ni
ne
, e
je
c
ti
on
f
r
a
c
ti
on,
bl
ood
pr
e
s
s
ur
e
,
a
nd
di
a
b
e
te
s
hi
s
to
r
y.
T
he
f
ol
lo
w
in
g
poi
nt
s
a
r
e
i
m
por
ta
nt
obs
e
r
va
ti
ons
f
r
om
t
hi
s
da
ta
s
e
t:
‒
T
he
r
e
a
r
e
no mi
s
s
in
g va
lu
e
s
.
‒
A
ll
f
e
a
tu
r
e
s
a
r
e
nume
r
ic
a
l
or
bi
na
r
y, s
o pr
e
pr
oc
e
s
s
in
g w
il
l
be
m
in
im
a
l.
‒
C
la
s
s
di
s
tr
ib
ut
io
n
of
D
E
A
T
H
_E
V
E
N
T
is
im
ba
la
n
c
e
d,
w
it
h
3
2%
de
c
e
a
s
e
d
(
1)
a
nd
68%
a
li
ve
(
0)
.
S
o,
th
e
r
e
i
s
a
c
la
s
s
i
m
ba
la
nc
e
. T
hi
s
m
u
s
t
be
ha
ndl
e
d i
n m
ode
l
e
v
a
lu
a
ti
on w
it
h s
tr
a
ti
f
ie
d s
pl
it
s
.
T
a
bl
e
3
s
how
s
a
s
um
m
a
r
y
of
th
e
s
pe
c
if
ic
a
ti
on
s
of
th
is
d
a
ta
s
e
t.
G
iv
e
n
th
e
c
la
s
s
im
ba
la
nc
e
of
th
e
D
E
A
T
H
_E
V
E
N
T
la
be
l
(
32%
pos
it
iv
e
,
68%
ne
ga
ti
ve
)
,
im
ba
l
a
nc
e
-
a
w
a
r
e
le
a
r
ni
ng
s
tr
a
te
gi
e
s
w
e
r
e
a
dopt
e
d.
S
pe
c
if
ic
a
ll
y,
f
or
L
R
,
R
F
,
a
nd
G
B
c
la
s
s
_w
e
ig
ht
=
'
ba
la
n
c
e
d'
opt
io
n
w
a
s
us
e
d
in
s
c
ik
it
-
le
a
r
n
to
up
-
w
e
ig
ht
th
e
m
in
or
it
y
c
la
s
s
dur
in
g
tr
a
in
in
g.
I
n
a
ddi
ti
ona
l
e
xpe
r
im
e
nt
s
,
s
ynt
he
ti
c
m
in
or
it
y
ove
r
s
a
m
pl
in
g
te
c
hni
que
(
S
M
O
T
E
)
w
a
s
e
va
lu
a
te
d
on
th
e
tr
a
in
in
g
s
e
t
to
ge
ne
r
a
t
e
s
ynt
h
e
ti
c
m
in
or
it
y
s
a
m
pl
e
s
,
a
nd
it
w
a
s
f
ound
th
a
t
pe
r
f
or
m
a
nc
e
tr
e
nds
w
e
r
e
c
ons
i
s
te
nt
,
s
o
th
e
c
la
s
s
-
w
e
ig
ht
r
e
s
ul
ts
w
e
r
e
r
e
por
te
d
f
or
s
im
pl
ic
it
y.
A
ll
tr
a
in
/t
e
s
t
s
pl
it
s
a
nd
c
r
os
s
-
va
li
da
ti
on
f
ol
ds
w
e
r
e
s
tr
a
ti
f
ie
d
to
pr
e
s
e
r
ve
th
e
c
la
s
s
pr
opor
ti
ons
.
E
xpl
or
a
to
r
y
da
ta
a
na
ly
s
is
(
E
D
A
)
w
il
l
be
c
a
r
r
ie
d
out
on
e
a
c
h
da
ta
s
e
t
be
f
or
e
m
ode
l
d
e
ve
lo
pm
e
nt
.
T
hi
s
in
c
lu
de
s
ope
r
a
ti
ons
s
uc
h
a
s
c
or
r
e
la
ti
ons
a
nd f
e
a
tu
r
e
i
m
por
ta
nc
e
.
T
a
bl
e
3. M
or
ta
li
ty
pr
e
di
c
ti
on da
ta
s
e
t
s
pe
c
if
ic
a
ti
ons
F
e
a
t
ur
e
D
e
s
c
r
i
pt
i
on
a
ge
A
ge
of
t
he
pa
t
i
e
nt
a
ne
m
i
a
1=
ye
s
, 0=
no
c
r
e
a
t
i
ni
ne
_phos
phoki
na
s
e
E
nz
ym
e
l
e
ve
l
di
a
be
t
e
s
1=
ye
s
, 0=
no
e
j
e
c
t
i
on_f
r
a
c
t
i
on
P
e
r
c
e
nt
a
ge
of
bl
ood l
e
a
vi
ng t
he
he
a
r
t
hi
gh_bl
ood_pr
e
s
s
ur
e
1=
ye
s
, 0=
no
pl
a
t
e
l
e
t
s
P
l
a
t
e
l
e
t
c
ount
s
e
r
um
_c
r
e
a
t
i
ni
ne
K
i
dne
y f
unc
t
i
on i
ndi
c
a
t
or
s
e
r
um
_s
odi
um
S
odi
um
l
e
ve
l
s
e
x
1=
M
a
l
e
, 0=
F
e
m
a
l
e
s
m
oki
ng
1=
ye
s
, 0=
no
time
D
ur
a
t
i
on of
f
ol
l
ow
-
up
(
da
ys
)
D
E
A
T
H
_E
V
E
N
T
T
a
r
ge
t
(
1=
de
a
t
h oc
c
ur
r
e
d, 0=
s
ur
vi
ve
d)
3.2. M
od
e
l
d
e
ve
lo
p
m
e
n
t
T
o
e
ns
ur
e
r
obus
tn
e
s
s
a
nd
ge
ne
r
a
li
z
a
ti
on
of
th
is
w
or
k,
a
num
be
r
of
M
L
m
ode
ls
w
e
r
e
de
ve
lo
pe
d
a
nd
c
om
pa
r
e
d,
in
c
lu
di
ng:
L
R
,
S
V
M
,
R
F
,
a
nd
G
B
.
A
ddi
ti
ona
ll
y,
a
s
ta
c
ke
d
e
ns
e
m
bl
e
m
ode
l
w
a
s
c
ons
tr
uc
te
d
by
c
om
bi
ni
ng
th
e
be
s
t
pe
r
f
or
m
in
g
c
la
s
s
if
ie
r
s
u
s
in
g
a
m
e
ta
-
c
la
s
s
if
ie
r
(
w
hi
c
h
is
L
R
)
tr
a
in
e
d
on
th
e
ir
out
put
pr
oba
bi
li
ti
e
s
.
H
ype
r
pa
r
a
m
e
te
r
tu
ni
ng
w
a
s
c
a
r
r
ie
d
out
us
in
g
G
r
id
S
e
a
r
c
hC
V
w
it
h
s
tr
a
ti
f
ie
d
5
-
f
ol
d
c
r
os
s
-
va
li
da
ti
on
to
ke
e
p
th
e
c
la
s
s
di
s
tr
ib
ut
io
n
in
e
a
c
h
f
ol
d.
F
or
e
a
c
h
m
ode
l,
a
num
be
r
of
hyp
e
r
pa
r
a
m
e
te
r
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
15
, N
o.
1
,
F
e
br
ua
r
y
20
26
:
191
-
212
194
(
e
.g.,
num
be
r
of
tr
e
e
s
,
m
a
xi
m
um
de
pt
h,
le
a
r
ni
ng
r
a
te
,
a
nd
r
e
gul
a
r
iz
a
ti
on
te
r
m
s
)
w
a
s
te
s
t
e
d,
a
nd
th
e
be
s
t
c
onf
ig
ur
a
ti
on
w
a
s
s
e
le
c
te
d
ba
s
e
d
on
m
e
a
n
r
e
c
e
iv
e
r
ope
r
a
ti
n
g
c
ha
r
a
c
te
r
is
ti
c
(
R
O
C
)
a
r
e
a
unde
r
th
e
c
ur
ve
(
A
U
C
)
a
c
r
os
s
f
ol
ds
.
F
or
r
e
pr
oduc
ib
il
it
y,
th
e
f
in
a
l
hype
r
pa
r
a
m
e
te
r
s
e
tt
in
gs
f
or
h
e
a
r
t
di
s
e
a
s
e
de
t
e
c
ti
on
a
nd
m
or
ta
li
ty
pr
e
di
c
ti
on
m
ode
ls
us
e
d
in
th
e
r
e
por
te
d
e
xpe
r
im
e
nt
s
a
r
e
s
um
m
a
r
iz
e
d
in
T
a
bl
e
4.
E
a
c
h
m
ode
l
i
s
th
e
n
e
v
a
lu
a
te
d
ba
s
e
d
on
m
e
tr
ic
s
a
nd
m
e
a
s
ur
e
m
e
nt
s
s
u
c
h
a
s
a
c
c
ur
a
c
y,
pr
e
c
is
i
on,
r
e
c
a
ll
,
a
nd
R
O
C
A
U
C
,
w
hi
c
h
a
r
e
us
e
d
f
or
bi
na
r
y
c
la
s
s
if
ic
a
ti
on
m
ode
ls
.
T
o
e
nha
nc
e
in
te
r
pr
e
ta
bi
li
ty
,
S
H
A
P
w
a
s
a
ppl
ie
d
to
th
e
be
s
t
-
pe
r
f
or
m
in
g
m
ode
ls
.
S
H
A
P
va
lu
e
s
w
e
r
e
c
om
put
e
d
to
id
e
nt
if
y
f
e
a
tu
r
e
im
por
ta
nc
e
a
nd
th
e
di
r
e
c
ti
on
of
f
e
a
tu
r
e
in
f
lu
e
nc
e
f
or
bot
h
in
di
vi
dua
l
pr
e
di
c
ti
ons
a
nd ove
r
a
ll
m
ode
l
be
ha
vi
or
.
T
a
bl
e
4. F
in
a
l
hype
r
pa
r
a
m
e
te
r
s
e
tt
in
g
s
f
or
he
a
r
t
di
s
e
a
s
e
d
e
te
c
ti
on a
nd mor
ta
li
ty
pr
e
di
c
ti
on mode
ls
M
ode
l
H
ype
r
pa
r
a
m
e
t
e
r
V
a
l
ue
D
e
s
c
r
i
pt
i
on
LR
pe
na
l
t
y
L2
R
e
gul
a
r
i
z
a
t
i
on t
ype
C
1.0
I
nve
r
s
e
of
r
e
gul
a
r
i
z
a
t
i
on s
t
r
e
ngt
h
s
ol
ve
r
l
bf
gs
O
pt
i
m
i
z
a
t
i
on a
l
gor
i
t
hm
m
a
x_i
t
e
r
5000
M
a
xi
m
um
i
t
e
r
a
t
i
ons
f
or
c
onve
r
ge
nc
e
r
a
ndom
_s
t
a
t
e
42
S
e
e
d f
or
r
e
pr
oduc
i
bi
l
i
t
y
RF
n_e
s
t
i
m
a
t
or
s
C
V
D
de
t
e
c
t
i
on:
200
M
or
t
a
l
i
t
y pr
e
di
c
t
i
on:
100
N
um
be
r
of
t
r
e
e
s
i
n t
he
f
or
e
s
t
m
a
x_de
pt
h
N
one
F
ul
l
y gr
ow
n t
r
e
e
s
(
no l
i
m
i
t
)
c
r
i
t
e
r
i
on
gi
ni
S
pl
i
t
qua
l
i
t
y c
r
i
t
e
r
i
on
m
i
n_s
a
m
pl
e
s
_s
pl
i
t
2
M
i
ni
m
um
s
a
m
pl
e
s
r
e
qui
r
e
d t
o s
pl
i
t
m
i
n_s
a
m
pl
e
s
_l
e
a
f
1
M
i
ni
m
um
s
a
m
pl
e
s
pe
r
l
e
a
f
m
a
x_f
e
a
t
ur
e
s
s
qr
t
N
um
be
r
of
f
e
a
t
ur
e
s
t
o c
ons
i
de
r
pe
r
s
pl
i
t
boot
s
t
r
a
p
T
r
ue
S
a
m
pl
i
ng w
i
t
h r
e
pl
a
c
e
m
e
nt
r
a
ndom
_s
t
a
t
e
42
S
e
e
d f
or
r
e
pr
oduc
i
bi
l
i
t
y
GB
n_e
s
t
i
m
a
t
or
s
C
V
D
de
t
e
c
t
i
on:
200
M
or
t
a
l
i
t
y pr
e
di
c
t
i
on:
100
N
um
be
r
of
boos
t
i
ng s
t
a
ge
s
l
e
a
r
ni
ng_r
a
t
e
0.05
S
hr
i
nka
ge
f
a
c
t
or
f
or
e
a
c
h s
t
a
ge
m
a
x_de
pt
h
3
D
e
pt
h of
i
ndi
vi
dua
l
w
e
a
k l
e
a
r
ne
r
s
s
ubs
a
m
pl
e
1.0
F
r
a
c
t
i
on of
s
a
m
pl
e
s
us
e
d pe
r
i
t
e
r
a
t
i
on
l
os
s
l
og_l
os
s
L
os
s
f
unc
t
i
on f
or
bi
na
r
y c
l
a
s
s
i
f
i
c
a
t
i
on
r
a
ndom
_s
t
a
t
e
42
S
e
e
d f
or
r
e
pr
oduc
i
bi
l
i
t
y
3.3. T
e
m
p
or
al
f
e
at
u
r
e
ab
la
t
io
n
T
he
“
ti
m
e
”
va
r
ia
bl
e
in
th
e
he
a
r
t
f
a
il
ur
e
da
ta
s
e
t,
w
hi
c
h
r
e
pr
e
s
e
nt
s
th
e
f
ol
lo
w
-
up
dur
a
ti
on
in
da
y
s
,
is
hi
ghl
y
c
or
r
e
la
te
d
w
it
h
th
e
D
E
A
T
H
_E
V
E
N
T
be
c
a
us
e
de
c
e
a
s
e
d
pa
ti
e
nt
s
of
te
n
ha
v
e
s
hor
te
r
ob
s
e
r
va
ti
on
pe
r
io
ds
.
F
r
om
one
pe
r
s
p
e
c
ti
ve
,
if
th
e
goa
l
i
s
to
pr
e
di
c
t
de
a
t
h
or
s
ur
vi
va
l
of
a
pa
ti
e
nt
,
th
e
n
ti
m
e
s
houl
d
pr
oba
bl
y
not
be
u
s
e
d a
s
a
n
in
put
to
th
e
m
ode
l.
O
n
th
e
ot
he
r
ha
n
d,
th
e
“
ti
m
e
”
va
r
ia
bl
e
c
a
n e
nc
ode
qui
te
us
e
f
ul
in
f
or
m
a
ti
on by e
xt
r
a
c
ti
ng f
e
a
tu
r
e
s
f
r
om
i
t.
T
o
a
s
s
e
s
s
th
e
e
f
f
e
c
t
of
te
m
por
a
l
in
f
or
m
a
ti
on
on
m
or
ta
li
ty
p
r
e
di
c
ti
on,
a
nd
to
m
in
im
iz
e
pot
e
nt
ia
l
in
f
or
m
a
ti
on l
e
a
ka
ge
, a
n a
bl
a
ti
on s
tu
dy
w
a
s
c
onduc
te
d us
in
g t
he
“
ti
m
e
”
va
r
ia
bl
e
f
r
o
m
t
he
he
a
r
t
f
a
il
ur
e
c
li
ni
c
a
l
r
e
c
or
ds
da
ta
s
e
t
(
pa
ti
e
nt
s
w
ho
s
ur
vi
ve
lo
nge
r
na
tu
r
a
ll
y
ha
ve
hi
ghe
r
“
ti
m
e
”
va
lu
e
s
)
.
T
hr
e
e
opt
io
ns
w
e
r
e
e
xa
m
in
e
d t
o qua
nt
if
y t
he
i
nf
lu
e
nc
e
of
t
e
m
por
a
l
f
e
a
tu
r
e
s
on mode
l
pe
r
f
or
m
a
nc
e
:
‒
W
it
hout
t
im
e
:
m
ode
ls
w
e
r
e
t
r
a
in
e
d us
in
g t
he
c
li
ni
c
a
l
f
e
a
tu
r
e
s
e
xc
lu
di
ng “
ti
m
e
”
.
‒
W
it
h t
im
e
:
th
e
r
a
w
“
ti
m
e
”
va
r
ia
bl
e
w
a
s
i
nc
lu
d
e
d a
s
a
n a
ddi
ti
ona
l
f
e
a
tu
r
e
.
‒
W
it
h
de
r
iv
e
d
ti
m
e
f
e
a
tu
r
e
s
:
th
e
r
a
w
“
ti
m
e
”
va
lu
e
is
tr
a
ns
f
o
r
m
e
d
us
in
g
lo
ga
r
it
hm
ic
a
nd
c
a
te
gor
ic
a
l
tr
a
ns
f
or
m
a
ti
ons
(
lo
g(
ti
m
e
)
)
,
a
nd
bi
ns
r
e
pr
e
s
e
nt
in
g
s
hor
t,
m
e
di
um
,
a
nd
lo
ng
f
ol
lo
w
-
up
du
r
a
ti
ons
w
e
r
e
in
c
or
por
a
te
d t
o c
a
pt
ur
e
t
e
m
por
a
l
dyna
m
ic
s
.
A
ll
m
ode
ls
w
e
r
e
tr
a
in
e
d
us
in
g
th
e
s
a
m
e
pr
e
pr
oc
e
s
s
in
g
pi
pe
li
ne
,
hype
r
pa
r
a
m
e
te
r
s
,
a
nd
RF
c
onf
ig
ur
a
ti
on
to
e
ns
ur
e
c
om
pa
r
a
bi
li
ty
. T
he
R
O
C
-
A
U
C
w
a
s
us
e
d a
s
t
h
e
pr
im
a
r
y e
va
lu
a
ti
on me
tr
ic
.
3.4.
E
xt
e
r
n
al
v
al
id
at
io
n
T
o
a
s
s
e
s
s
ge
ne
r
a
li
z
a
bi
li
ty
,
a
n
in
de
pe
nd
e
nt
e
xt
e
r
na
l
da
ta
s
e
t
w
a
s
us
e
d
f
or
va
li
da
ti
on.
T
he
U
C
I
he
a
r
t
di
s
e
a
s
e
da
ta
s
e
t
(n
=
920)
[
28]
w
a
s
s
e
le
c
te
d
a
s
it
c
ont
a
in
s
c
o
m
pa
r
a
bl
e
c
li
ni
c
a
l
pr
e
di
c
to
r
s
s
uc
h
a
s
a
ge
,
s
e
x,
c
hol
e
s
te
r
ol
,
r
e
s
ti
ng
bl
ood
pr
e
s
s
ur
e
,
f
a
s
ti
ng
bl
ood
s
ug
a
r
,
e
xe
r
c
i
s
e
a
ngi
na
,
a
nd
m
a
xi
m
um
he
a
r
t
r
a
te
.
T
he
ta
r
ge
t
va
r
ia
bl
e
w
a
s
c
onve
r
te
d
to
a
bi
na
r
y
in
di
c
a
to
r
.
F
e
a
tu
r
e
m
a
ppi
ng
s
w
e
r
e
a
li
gne
d
w
it
h
th
e
m
a
in
K
a
ggl
e
da
ta
s
e
t,
a
nd
m
is
s
in
g
va
lu
e
s
w
e
r
e
h
a
ndl
e
d
by
m
e
di
a
n
im
put
a
ti
on. T
he
R
e
s
ti
ngE
C
G
f
e
a
tu
r
e
w
a
s
e
nt
ir
e
ly
m
is
s
in
g
in
th
e
U
C
I
da
ta
s
e
t,
he
nc
e
,
a
ne
ut
r
a
l
ba
s
e
li
ne
va
lu
e
(
0=
nor
m
a
l
E
C
G
)
w
a
s
a
s
s
ig
ne
d
to
m
a
in
ta
in
c
om
pa
ti
bi
li
ty
w
it
h
th
e
tr
a
in
e
d
m
ode
ls
.
T
he
da
ta
w
e
r
e
s
c
a
le
d
us
in
g
th
e
s
a
m
e
nor
m
a
li
z
a
ti
on
pa
r
a
m
e
te
r
s
,
w
hi
c
h
w
e
r
e
de
r
iv
e
d
f
r
om
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
E
x
pl
ai
nabl
e
hy
br
id
m
ode
ls
f
or
c
ar
di
ov
a
s
c
ul
ar
di
s
e
as
e
d
e
te
c
ti
o
n and mor
ta
li
ty
pr
e
di
c
ti
on
(
A
li
A
l
-
A
ta
by
)
195
th
e
m
a
in
da
ta
s
e
t
to
e
ns
ur
e
c
ons
is
t
e
nc
y.
A
ll
m
ode
ls
w
e
r
e
r
e
tr
a
i
ne
d
on
th
e
m
a
in
da
ta
s
e
t
a
nd
e
va
lu
a
t
e
d
w
it
hout
f
ur
th
e
r
t
uni
ng on the
e
xt
e
r
na
l
da
ta
s
e
t
to
m
e
a
s
ur
e
t
r
ue
pe
r
f
or
m
a
nc
e
.
3.5.
F
u
ll
p
i
p
e
li
n
e
f
or
h
e
ar
t
d
is
e
as
e
an
d
m
or
t
al
it
y p
r
e
d
ic
t
io
n
F
ig
ur
e
1
s
how
s
th
e
e
nd
-
to
-
e
nd
pr
e
pr
oc
e
s
s
in
g
a
nd
e
va
lu
a
ti
on
pi
pe
li
ne
f
or
he
a
r
t
di
s
e
a
s
e
a
nd
m
or
ta
li
ty
pr
e
di
c
ti
on.
R
a
w
da
ta
s
e
ts
a
r
e
c
le
a
ne
d,
e
nc
od
e
d,
im
put
e
d
w
h
e
r
e
ne
c
e
s
s
a
r
y,
a
nd
s
ta
nd
a
r
di
z
e
d.
M
ode
l
s
a
r
e
tr
a
in
e
d
us
in
g
s
tr
a
ti
f
ie
d
tr
a
in
/t
e
s
t
s
pl
it
s
a
nd
s
ub
s
e
que
nt
ly
e
va
lu
a
te
d
on
bot
h
th
e
in
te
r
na
l
te
s
t
pa
r
ti
ti
on
a
nd
a
n
in
de
pe
nde
nt
U
C
I
da
ta
s
e
t
to
a
s
s
e
s
s
e
xt
e
r
na
l
ge
ne
r
a
li
z
a
ti
on.
F
ig
ur
e
1. P
r
e
pr
oc
e
s
s
in
g a
nd e
xt
e
r
na
l
va
li
da
ti
on pipe
li
ne
f
or
he
a
r
t
di
s
e
a
s
e
a
nd mor
ta
li
ty
pr
e
di
c
ti
on
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
15
, N
o.
1
,
F
e
br
ua
r
y
20
26
:
191
-
212
196
T
he
f
ol
lo
w
in
g
s
t
e
ps
s
um
m
a
r
iz
e
th
e
e
nd
-
to
-
e
nd
pr
e
pr
oc
e
s
s
in
g
a
nd
e
va
lu
a
ti
on
pi
pe
li
ne
f
or
he
a
r
t
di
s
e
a
s
e
a
nd mor
ta
li
ty
pr
e
di
c
ti
on:
i)
D
a
ta
s
our
c
e
s
‒
K
a
ggl
e
he
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on
(
C
V
D
de
te
c
ti
on)
[
26]
.
‒
K
a
ggl
e
he
a
r
t
f
a
il
ur
e
c
li
ni
c
a
l
r
e
c
or
ds
(
m
or
ta
li
ty
)
[
27]
.
‒
U
C
I
he
a
r
t
di
s
e
a
s
e
da
ta
s
e
t
(
e
xt
e
r
na
l
va
li
da
ti
on)
[
28]
.
ii)
D
a
ta
c
le
a
ni
ng
‒
F
or
t
he
C
V
D
de
te
c
ti
on da
ta
s
e
t,
r
e
m
ove
R
e
s
ti
ngB
P
=
0, C
hol
e
s
te
r
ol
=
0;
t
ype
c
he
c
ks
.
‒
F
or
t
he
m
or
ta
li
ty
da
ta
s
e
t,
r
a
nge
c
he
c
ks
, t
ype
c
a
s
ti
ng;
no r
ow
s
r
e
m
ove
d.
‒
F
or
t
he
e
xt
e
r
na
l
da
ta
s
e
t,
ha
ndl
e
N
a
N
s
:
m
e
di
a
n i
m
put
a
ti
on;
R
e
s
t
in
gE
C
G
t
o “
nor
m
a
l”
.
iii)
E
nc
odi
ng a
nd
f
e
a
tu
r
e
s
e
le
c
ti
on
‒
E
nc
ode
c
a
te
gor
ic
a
l
va
r
ia
bl
e
s
(
s
e
x
, C
he
s
tP
a
in
T
ype
,
R
e
s
ti
ngE
C
G
, E
xe
r
c
is
e
A
ngi
na
, S
T
_S
lo
pe
)
.
‒
S
e
le
c
t
s
ha
r
e
d
c
li
ni
c
a
l
f
e
a
tu
r
e
s
(
a
ge
,
s
e
x,
B
P
,
c
hol
e
s
te
r
o
l
,
M
a
xH
R
,
O
ld
pe
a
k,
F
a
s
ti
ngB
S
,
R
e
s
ti
ngE
C
G
, E
xe
r
c
is
e
A
ngi
na
)
.
‒
F
or
m
or
ta
li
ty
p
r
e
di
c
ti
on, r
e
ta
in
a
ll
12 pr
e
di
c
to
r
s
.
iv
)
S
c
a
li
ng
‒
F
it
S
ta
nda
r
dS
c
a
le
r
on t
he
m
a
in
t
r
a
in
in
g s
e
t.
‒
A
ppl
y t
he
s
a
m
e
s
c
a
le
r
t
o i
nt
e
r
na
l
te
s
t
a
nd e
xt
e
r
na
l
U
C
I
da
ta
.
v)
T
r
a
in
/
te
s
t
s
pl
it
:
s
tr
a
ti
f
ie
d
tr
a
in
/t
e
s
t
s
pl
it
(
80/
20)
a
nd pr
e
s
e
r
ve
d c
la
s
s
ba
la
n
c
e
.
vi
)
M
ode
l
tr
a
in
in
g:
t
r
a
in
L
R
, R
F
, G
B
(
a
nd s
ta
c
ki
ng f
or
m
or
ta
li
ty
)
.
vi
i)
E
va
lu
a
ti
on
‒
I
nt
e
r
na
l
e
va
lu
a
ti
on (
m
a
in
da
ta
s
e
t)
:
R
O
C
A
U
C
,
a
c
c
ur
a
c
y, pr
e
c
is
io
n, r
e
c
a
ll
, a
nd F
1
-
s
c
or
e
.
‒
E
xt
e
r
na
l
e
va
lu
a
ti
on (
U
C
I
)
:
s
a
m
e
m
e
tr
ic
s
, c
om
pa
r
e
ge
ne
r
a
li
z
a
ti
o
n.
4.
I
M
P
L
E
M
E
N
T
A
T
I
O
N
4.1. Dat
as
e
t
e
xp
lo
r
at
or
y d
at
a an
al
ys
is
E
D
A
w
a
s
pe
r
f
or
m
e
d
on
bot
h
da
ta
s
e
ts
to
ge
t
a
be
tt
e
r
id
e
a
a
bout
th
e
da
ta
s
e
t
be
f
or
e
de
ve
lo
pi
ng
th
e
m
ode
ls
. T
hi
s
i
nc
lu
de
s
f
e
a
tu
r
e
c
or
r
e
la
ti
on he
a
tm
a
p, pa
ir
pl
ot
s
, a
n
d f
e
a
tu
r
e
i
m
por
ta
nc
e
a
na
ly
s
is
. F
ig
ur
e
2 s
how
s
f
e
a
tu
r
e
c
or
r
e
la
ti
on
he
a
tm
a
p
f
or
he
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on
d
a
ta
s
e
t.
F
r
om
F
ig
ur
e
2,
it
c
a
n
be
s
e
e
n
th
a
t
th
e
s
tr
onge
s
t
c
or
r
e
la
ti
ons
w
it
h
H
e
a
r
tDi
s
e
a
s
e
f
e
a
tu
r
e
a
r
e
:
i)
E
xe
r
c
is
e
A
ngi
na
:
pos
it
iv
e
c
or
r
e
la
ti
on,
ii
)
O
ld
pe
a
k:
pos
it
iv
e
c
or
r
e
la
ti
on,
ii
i)
S
T
_S
lo
pe
:
ne
ga
ti
ve
c
or
r
e
la
ti
on,
a
nd
iv
)
C
he
s
tP
a
in
T
ype
:
ne
ga
ti
ve
c
or
r
e
la
ti
on.
F
ig
ur
e
3
s
how
s
pa
ir
pl
ot
f
o
r
he
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on
da
ta
s
e
t.
F
r
om
th
is
f
ig
ur
e
,
th
e
f
ol
lo
w
in
g
ke
y
obs
e
r
va
ti
ons
a
r
e
obt
a
in
e
d:
i)
a
ge
:
pa
ti
e
nt
s
w
it
h
he
a
r
t
di
s
e
a
s
e
a
ppe
a
r
s
li
ght
ly
ol
de
r
on
a
ve
r
a
ge
,
ii
)
c
hol
e
s
te
r
ol
:
no
c
le
a
r
vi
s
ua
l
s
e
pa
r
a
ti
on
be
twe
e
n
th
e
c
la
s
s
e
s
,
ii
i)
M
a
xH
R
:
ge
ne
r
a
ll
y
lo
w
e
r
in
he
a
r
t
di
s
e
a
s
e
pa
ti
e
nt
s
,
iv
)
ol
dpe
a
k:
c
le
a
r
ly
hi
ghe
r
in
pa
ti
e
nt
s
w
it
h
he
a
r
t
di
s
e
a
s
e
,
a
nd
v)
R
e
s
ti
ngB
P
:
ove
r
la
ppi
ng
di
s
tr
ib
ut
io
ns
,
s
o
it
m
a
y
not
be
us
e
f
ul
be
f
or
e
f
e
a
tu
r
e
e
ngi
ne
e
r
in
g.
F
ig
ur
e
2. F
e
a
tu
r
e
c
or
r
e
la
ti
on he
a
tm
a
p f
or
he
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on da
ta
s
e
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
E
x
pl
ai
nabl
e
hy
br
id
m
ode
ls
f
or
c
ar
di
ov
a
s
c
ul
ar
di
s
e
as
e
d
e
te
c
ti
o
n and mor
ta
li
ty
pr
e
di
c
ti
on
(
A
li
A
l
-
A
ta
by
)
197
F
ig
ur
e
3. P
a
ir
pl
ot
f
o
r
he
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on da
ta
s
e
t
F
ig
ur
e
4
s
how
s
f
e
a
tu
r
e
im
po
r
ta
nc
e
a
na
ly
s
is
ba
r
c
ha
r
t.
F
r
om
F
ig
ur
e
4,
it
c
a
n
be
s
e
e
n
th
a
t
th
e
to
p
5
m
os
t
im
por
ta
nt
f
e
a
tu
r
e
s
f
or
he
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on a
r
e
a
s
li
s
te
d
in
T
a
bl
e
5
w
it
h
th
e
ir
pe
r
c
e
nt
a
g
e
im
por
ta
nc
e
.
F
r
om
F
ig
ur
e
4 a
nd T
a
bl
e
5, i
t
c
a
n be
c
onc
lu
de
d t
ha
t:
‒
S
T
_S
lo
pe
a
nd
ol
dpe
a
k
f
e
a
tu
r
e
s
(
r
e
la
te
d
to
S
T
s
e
gm
e
nt
of
th
e
E
C
G
du
r
in
g
e
xe
r
c
is
e
)
a
r
e
im
por
ta
n
t
in
di
c
a
to
r
s
.
‒
C
he
s
t
pa
in
a
nd
a
ngi
na
a
ls
o pl
a
y
a
s
ig
ni
f
ic
a
nt
r
ol
e
i
n he
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on
‒
H
e
a
r
t
r
a
te
(
M
a
xH
R
)
is
a
ls
o
a
n
im
por
ta
nt
f
e
a
tu
r
e
,
w
hi
le
ba
s
ic
f
e
a
tu
r
e
s
s
uc
h
a
s
a
ge
,
c
hol
e
s
te
r
ol
,
or
R
e
s
ti
ngB
P
c
ont
r
ib
ut
e
l
e
s
s
.
F
ig
ur
e
4. F
e
a
tu
r
e
i
m
por
ta
nc
e
a
na
ly
s
is
f
or
he
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
15
, N
o.
1
,
F
e
br
ua
r
y
20
26
:
191
-
212
198
T
a
bl
e
5.
T
op 5 mos
t
im
por
ta
nt
f
e
a
tu
r
e
s
f
or
he
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on da
ta
s
e
t
R
a
nk
F
e
a
t
ur
e
I
m
por
t
a
nc
e
(
%
)
1
S
T
_S
l
ope
25.12
2
O
l
dpe
a
k
13.33
3
C
he
s
t
P
a
i
nT
ype
11.57
4
M
a
xH
R
10.24
5
E
xe
r
c
i
s
e
A
ngi
na
9.13
F
ig
ur
e
5
s
how
s
th
e
f
e
a
tu
r
e
c
or
r
e
la
ti
on
h
e
a
tm
a
p
f
or
m
or
ta
li
ty
pr
e
di
c
ti
on
f
r
om
th
e
he
a
r
t
di
s
e
a
s
e
da
ta
s
e
t.
F
r
om
F
ig
ur
e
5, i
t
c
a
n be
s
e
e
n t
ha
t
th
e
s
tr
onge
s
t
c
or
r
e
la
t
io
ns
w
it
h D
E
A
T
H
_E
V
E
N
T
t
a
r
ge
t
va
r
ia
bl
e
a
r
e
i)
s
e
r
um
_c
r
e
a
ti
ni
ne
:
pos
it
iv
e
c
or
r
e
la
ti
on
,
ii
)
a
ge
:
pos
it
iv
e
c
or
r
e
la
ti
on
,
ii
i)
ti
m
e
:
ne
ga
ti
ve
c
or
r
e
la
ti
on
,
a
n
d
iv
)
e
je
c
ti
on_f
r
a
c
ti
on:
ne
ga
ti
ve
c
or
r
e
la
ti
on
.
F
ig
ur
e
6
s
how
s
th
e
f
e
a
tu
r
e
im
por
ta
nc
e
a
na
ly
s
is
b
a
r
c
ha
r
t.
F
r
om
F
ig
ur
e
6,
it
c
a
n
be
s
e
e
n
th
a
t
th
e
to
p
5
m
o
s
t
im
por
ta
nt
f
e
a
tu
r
e
s
f
or
he
a
r
t
di
s
e
a
s
e
m
or
ta
li
ty
pr
e
di
c
ti
on
a
r
e
a
s
li
s
te
d i
n T
a
bl
e
6 w
it
h t
he
ir
pe
r
c
e
nt
a
g
e
i
m
por
ta
nc
e
. F
r
om
F
ig
ur
e
6 a
nd T
a
bl
e
6, i
t
c
a
n b
e
c
onc
lu
de
d t
ha
t:
‒
T
he
t
im
e
f
e
a
tu
r
e
domi
na
te
s
t
he
pr
e
di
c
ti
on, with patient
s
w
ho s
u
r
vi
ve
l
onge
r
a
r
e
l
e
s
s
l
ik
e
ly
t
o di
e
.
‒
K
id
ne
y r
e
la
te
d f
e
a
tu
r
e
s
(
s
e
r
um
_c
r
e
a
ti
ni
ne
, c
r
e
a
ti
ni
ne
_phos
pho
ki
na
s
e
)
a
r
e
s
tr
ong c
li
ni
c
a
l
in
di
c
a
to
r
s
.
‒
P
la
te
le
ts
a
ls
o i
nf
lu
e
nc
e
r
is
k.
F
ig
ur
e
5. F
e
a
tu
r
e
c
or
r
e
la
ti
on he
a
tm
a
p f
or
he
a
r
t
di
s
e
a
s
e
m
or
ta
li
ty
pr
e
di
c
ti
on da
ta
s
e
t
F
ig
ur
e
6. F
e
a
tu
r
e
i
m
por
ta
nc
e
a
na
ly
s
is
f
or
m
or
ta
li
ty
pr
e
di
c
ti
on
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
E
x
pl
ai
nabl
e
hy
br
id
m
ode
ls
f
or
c
ar
di
ov
a
s
c
ul
ar
di
s
e
as
e
d
e
te
c
ti
o
n and mor
ta
li
ty
pr
e
di
c
ti
on
(
A
li
A
l
-
A
ta
by
)
199
T
a
bl
e
6.
T
op 5 mos
t
im
por
ta
nt
f
e
a
tu
r
e
s
f
or
he
a
r
t
m
or
ta
li
ty
pr
e
di
c
ti
on
R
a
nk
F
e
a
t
ur
e
I
m
por
t
a
nc
e
(
%
)
1
t
i
m
e
(
f
ol
l
ow
-
up da
ys
)
58.69
2
s
e
r
um
_c
r
e
a
t
i
ni
ne
12.13
3
e
j
e
c
t
i
on_f
r
a
c
t
i
on
10.39
4
pl
a
t
e
l
e
t
s
6.23
5
c
r
e
a
t
i
ni
ne
_phos
phoki
na
s
e
5.41
4.2. M
od
e
l
im
p
le
m
e
n
t
at
io
n
I
n
th
is
w
or
k,
P
yt
hon
3
pr
ogr
a
m
m
in
g
la
ngua
ge
w
a
s
us
e
d
in
th
e
G
oogl
e
C
ol
a
b
a
nd
J
upyt
e
r
N
ot
e
book
de
ve
lo
pm
e
nt
e
nvi
r
onm
e
nt
s
to
im
pl
e
m
e
nt
th
e
s
ugge
s
te
d
m
od
e
ls
.
M
or
e
ove
r
,
th
e
f
ol
lo
w
in
g
P
yt
hon
li
br
a
r
ie
s
w
e
r
e
ut
il
iz
e
d i
n t
he
de
ve
lo
pm
e
nt
:
‒
S
c
ik
it
-
le
a
r
n l
ib
r
a
r
y:
us
e
d f
or
t
r
a
in
in
g a
nd a
s
s
e
s
s
in
g m
ode
ls
.
‒
S
H
A
P
l
ib
r
a
r
y:
us
e
d t
o m
a
ke
m
ode
ls
i
nt
e
r
pr
e
ta
bl
e
a
nd e
xpl
a
in
a
b
le
.
‒
T
he
pa
nda
s
a
nd
N
um
P
y
li
br
a
r
ie
s
a
r
e
us
e
d t
o m
a
ni
pul
a
te
a
nd pr
e
pr
oc
e
s
s
d
a
ta
.
‒
T
he
m
a
tp
lo
tl
ib
a
nd s
e
a
bor
n l
ib
r
a
r
ie
s
a
r
e
us
e
d t
o vi
s
ua
li
z
e
m
ode
l
pe
r
f
or
m
a
nc
e
, S
H
A
P
va
lu
e
s
, a
nd f
e
a
tu
r
e
im
por
ta
nc
e
.
T
he
da
ta
s
e
t
w
a
s
s
pl
it
in
to
80:
20
r
a
ti
o
f
or
tr
a
in
in
g
a
nd
te
s
ti
ng
s
e
ts
.
R
F
f
e
a
tu
r
e
im
por
ta
nc
e
w
a
s
f
ir
s
t
c
om
put
e
d
ba
s
e
d
on
G
in
i
im
pur
it
y
f
o
r
he
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on.
T
he
n,
gr
a
di
e
nt
boos
t
f
e
a
tu
r
e
im
por
ta
nc
e
w
a
s
c
om
put
e
d
f
or
he
a
r
t
di
s
e
a
s
e
m
or
ta
li
ty
pr
e
di
c
ti
on.
T
o
in
te
r
pr
e
t
th
e
m
ode
ls
a
nd
e
n
s
ur
e
c
li
ni
c
a
l
r
e
le
va
nc
e
,
S
H
A
P
a
na
ly
s
is
w
a
s
a
ppl
ie
d
a
c
r
os
s
a
ll
m
ode
ls
to
pr
ovi
de
a
c
ons
is
t
e
n
t
e
xpl
a
na
ti
on
of
f
e
a
tu
r
e
c
ont
r
ib
ut
io
ns
a
t
bot
h
gl
oba
l
a
nd
lo
c
a
l
(
pa
ti
e
nt
-
s
pe
c
if
ic
)
le
ve
ls
.
C
om
pa
r
a
ti
ve
S
H
A
P
pl
ot
s
ha
ve
s
how
n
di
f
f
e
r
e
nc
e
s
in
how
L
R
,
R
F
,
a
nd G
B
ut
il
iz
e
c
li
ni
c
a
l
f
e
a
tu
r
e
s
(
e
.g., e
je
c
ti
on f
r
a
c
ti
on, s
e
r
um
c
r
e
a
ti
ni
ne
,
a
nd
a
ge
)
.
4.3. Ap
p
li
c
at
io
n
d
e
p
lo
ym
e
n
t
T
he
be
s
t
pe
r
f
or
m
in
g
m
ode
l
f
r
om
e
a
c
h
s
ta
ge
w
a
s
de
pl
oye
d
a
s
in
te
r
a
c
ti
ve
w
e
b
a
ppl
ic
a
ti
ons
u
s
in
g
S
tr
e
a
m
li
t.
T
he
a
ppl
ic
a
ti
on
s
a
ll
ow
c
li
ni
c
ia
ns
to
in
put
pa
ti
e
nt
d
a
ta
a
nd
r
e
c
e
iv
e
im
m
e
di
a
te
he
a
r
t
di
s
e
a
s
e
a
nd
m
or
ta
li
ty
r
is
k
p
r
e
di
c
ti
ons
.
T
he
s
e
a
ppl
ic
a
ti
ons
e
nha
nc
e
a
c
c
e
s
s
ib
il
it
y
a
nd
pr
ovi
de
r
e
a
l
-
ti
m
e
de
c
is
io
n
s
uppor
t
in
he
a
lt
hc
a
r
e
c
li
ni
c
s
.
5.
R
E
S
U
L
T
S
A
N
D
D
I
S
C
U
S
S
I
O
N
5.1. Re
s
u
lt
s
of
h
e
ar
t
d
is
e
a
s
e
p
r
e
d
ic
t
io
n
F
ig
ur
e
7
s
how
s
th
e
R
O
C
c
ur
ve
f
or
th
e
LR
m
ode
l,
w
hi
c
h
w
a
s
us
e
d
a
s
th
e
ba
s
e
li
ne
.
T
he
obt
a
in
e
d
A
U
C
w
a
s
0.93 in t
hi
s
c
a
s
e
. T
he
r
e
s
ul
ti
ng A
U
C
s
c
or
e
i
ndi
c
a
te
s
e
xc
e
ll
e
nt
di
s
c
r
im
in
a
ti
on be
twe
e
n he
a
r
t
di
s
e
a
s
e
a
nd non
-
di
s
e
a
s
e
c
a
s
e
s
.
F
ig
ur
e
7. R
O
C
c
ur
ve
f
or
L
R
m
ode
l
T
a
bl
e
7
s
how
s
th
e
pe
r
f
or
m
a
nc
e
of
th
e
f
our
m
ode
ls
th
a
t
w
e
r
e
de
ve
lo
pe
d
f
or
th
is
s
t
a
ge
.
F
r
om
th
e
ta
bl
e
,
it
c
a
n
be
s
e
e
n
th
a
t
th
e
GB
m
ode
l
pe
r
f
or
m
s
th
e
be
s
t
ov
e
r
a
ll
.
T
he
R
F
m
ode
l
i
s
a
ls
o
s
tr
ong
a
nd
of
te
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
15
, N
o.
1
,
F
e
br
ua
r
y
20
26
:
191
-
212
200
e
a
s
ie
r
to
in
te
r
pr
e
t.
T
he
S
V
M
m
ode
l
unde
r
pe
r
f
or
m
s
s
ig
ni
f
ic
a
nt
ly
in
th
is
s
e
tu
p.
A
c
c
or
di
ngl
y,
th
e
G
B
m
ode
l
is
r
e
c
om
m
e
nde
d f
or
de
pl
oym
e
nt
or
f
ur
th
e
r
t
uni
ng.
T
a
bl
e
7. M
od
e
l
pe
r
f
or
m
a
nc
e
c
om
pa
r
is
on f
or
he
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on
M
ode
l
R
O
C
A
U
C
A
c
c
ur
a
c
y
P
r
e
c
i
s
i
on
R
e
c
a
l
l
F1
-
s
c
or
e
GB
0.958
0.893
0.932
0.861
0.895
RF
0.953
0.887
0.931
0.848
0.887
LR
0.930
0.867
0.893
0.848
0.870
S
V
M
0.721
0.673
0.727
0.608
0.662
5.2. Re
s
u
lt
s
of
m
or
t
al
it
y p
r
e
d
ic
t
io
n
d
u
e
t
o h
e
ar
t
d
is
e
as
e
F
or
th
is
s
ta
ge
,
th
r
e
e
m
ode
ls
w
e
r
e
te
s
te
d. T
he
s
e
a
r
e
G
B
,
L
R
,
a
nd
R
F
.
T
he
S
V
M
w
a
s
e
xc
lu
de
d
in
th
is
pa
r
t
be
c
a
us
e
it
i
s
li
ke
ly
to
unde
r
pe
r
f
or
m
s
im
il
a
r
to
th
e
pr
e
vi
ous
pa
r
t.
T
a
bl
e
8
pr
ovi
de
s
a
s
um
m
a
r
y
of
th
e
pe
r
f
or
m
a
nc
e
c
om
pa
r
is
on
of
th
e
th
r
e
e
te
s
te
d
m
ode
ls
,
a
nd
F
ig
ur
e
8
s
how
s
th
e
R
O
C
c
ur
ve
s
f
or
th
e
c
or
r
e
s
ponding deve
lo
pe
d m
ode
ls
.
T
a
bl
e
8. M
od
e
l
pe
r
f
or
m
a
nc
e
c
om
pa
r
is
on f
or
m
or
ta
li
ty
pr
e
di
c
ti
o
n due
t
o he
a
r
t
di
s
e
a
s
e
M
ode
l
R
O
C
A
U
C
A
c
c
ur
a
c
y
P
r
e
c
i
s
i
on
R
e
c
a
l
l
F1
-
s
c
or
e
RF
0.899
0.833
0.846
0.579
0.688
LR
0.855
0.800
0.733
0.579
0.647
GB
0.827
0.800
0.706
0.632
0.667
F
ig
ur
e
8. R
O
C
c
ur
ve
s
f
or
L
R
, R
F
, a
nd G
B
m
ode
ls
f
or
m
or
ta
li
ty
pr
e
di
c
ti
on
F
r
om
th
e
T
a
bl
e
8,
it
c
a
n
be
s
e
e
n
th
a
t
th
e
R
F
m
ode
l
ha
s
th
e
hi
g
he
s
t
R
O
C
A
U
C
,
w
hi
c
h
in
di
c
a
te
s
it
is
th
e
be
s
t
a
t
s
e
pa
r
a
ti
ng
d
e
c
e
a
s
e
d
f
r
om
s
ur
vi
vor
pa
ti
e
nt
s
,
a
nd
a
c
hi
e
vi
ng
th
e
hi
gh
e
s
t
a
c
c
ur
a
c
y,
pr
e
c
is
io
n,
a
nd
F1
-
s
c
or
e
.
L
R
m
ode
l
ha
s
s
how
n
a
good
pe
r
f
or
m
a
nc
e
,
w
it
h
g
ood
R
O
C
A
U
C
a
nd
a
c
c
ur
a
c
y.
G
B
pe
r
f
or
m
s
r
e
a
s
ona
bl
y
in
te
r
m
s
of
R
O
C
A
U
C
a
nd
a
c
c
ur
a
c
y,
w
it
h
th
e
h
i
ghe
s
t
obt
a
in
e
d
r
e
c
a
ll
va
lu
e
.
I
f
in
te
r
pr
e
ta
bi
li
ty
m
a
tt
e
r
s
m
os
t,
th
e
n
L
R
i
s
r
e
c
om
m
e
nde
d,
a
nd
if
pr
e
c
i
s
io
n
(
a
voi
di
ng
f
a
ls
e
a
la
r
m
s
)
m
a
tt
e
r
s
m
os
t,
th
e
R
F
m
ode
l
is
r
e
c
om
m
e
nde
d. H
ow
e
ve
r
, i
f
r
e
c
a
ll
(
c
a
tc
hi
ng a
ll
t
r
ue
de
a
th
s
)
m
a
tt
e
r
s
m
os
t,
t
he
n G
B
c
oul
d be
a
dopt
e
d.
5.3. M
od
e
l
t
u
n
in
g r
e
s
u
lt
s
f
or
m
or
t
al
it
y p
r
e
d
ic
t
io
n
F
r
om
th
e
pr
e
vi
ous
r
e
s
ul
ts
of
th
e
m
ode
ls
pr
e
di
c
ti
ng
m
or
ta
li
ty
d
ue
to
he
a
r
t
di
s
e
a
s
e
,
it
c
a
n
be
s
e
e
n
th
a
t
th
e
pe
r
f
or
m
a
nc
e
of
th
e
s
e
m
ode
ls
ne
e
d
s
to
be
im
pr
ove
d.
S
in
c
e
th
is
is
a
m
or
ta
li
ty
pr
e
di
c
ti
on
ta
s
k,
r
e
c
a
ll
(
i.
e
.,
s
e
ns
it
iv
it
y)
is
m
or
e
c
r
it
ic
a
l
th
a
n
pr
e
c
is
io
n,
s
o
it
is
be
tt
e
r
to
f
la
g
m
or
e
pa
ti
e
nt
s
a
s
a
t
r
is
k
th
a
n
to
m
is
s
a
r
e
a
l
c
a
s
e
.
T
hi
s
c
a
n be
c
a
r
r
ie
d out e
it
he
r
by mode
l
tu
ni
ng or
us
in
g a
s
ta
c
ke
d
e
ns
e
m
bl
e
m
ode
l.
Evaluation Warning : The document was created with Spire.PDF for Python.