T
E
L
KO
M
N
I
KA
T
e
lec
om
m
u
n
icat
ion
,
Com
p
u
t
i
n
g,
E
lec
t
r
on
ics
an
d
Cont
r
ol
Vol.
18
,
No.
3
,
J
une
2020
,
pp.
1
2
8
5
~
1
29
1
I
S
S
N:
1693
-
6930,
a
c
c
r
e
dit
e
d
F
ir
s
t
G
r
a
de
by
Ke
me
nr
is
tekdikti
,
De
c
r
e
e
No:
21/E
/KP
T
/2018
DO
I
:
10.
12928/
T
E
L
KO
M
NI
KA
.
v18i3.
15225
1285
Jou
r
n
al
h
omepage
:
ht
tp:
//
jour
nal.
uad
.
ac
.
id/
index
.
php/T
E
L
K
OM
N
I
K
A
T
r
ai
n
in
g
f
e
e
d
f
or
w
ar
d
n
e
u
r
al
n
e
t
w
or
k
u
si
n
g g
e
n
e
t
ic
al
go
r
it
h
m
t
o d
ia
gn
os
e
l
e
f
t
v
e
n
t
r
i
c
u
la
r
h
y
p
e
r
t
r
o
p
h
y
J.
Re
vat
h
i
1
,
J
.
Ani
t
h
a
2
,
D.
J
u
d
e
Hem
a
n
t
h
3
1
D
ep
ar
t
men
t
o
f
Bi
o
med
i
cal
E
n
g
i
n
eer
i
n
g
,
K
I
T
-
K
a
l
ai
g
n
ar
k
aru
n
an
i
d
h
i
In
s
t
i
t
u
t
e
o
f
T
ech
n
o
l
o
g
y
,
In
d
i
a
2,
3
D
e
p
art
me
n
t
o
f
E
l
ec
t
ro
n
i
c
s
an
d
C
o
mmu
n
i
ca
t
i
o
n
E
n
g
i
n
eeri
n
g
,
K
aru
n
y
a
In
s
t
i
t
u
t
e
o
f
T
ech
n
o
l
o
g
y
a
n
d
Sc
i
en
ce
s
,
In
d
i
a
Ar
t
icle
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
ived
J
a
n
8
,
20
20
R
e
vis
e
d
F
e
b
6
,
2020
Ac
c
e
pted
F
e
b
24
,
2020
In
t
h
i
s
res
earc
h
w
o
rk
,
a
n
ew
t
ec
h
n
i
q
u
e
w
as
p
r
o
p
o
s
e
d
fo
r
t
h
e
d
i
ag
n
o
s
i
s
o
f
l
eft
v
en
t
ri
c
u
l
ar
h
y
p
ert
r
o
p
h
y
(L
V
H
)
fro
m
t
h
e
E
CG
s
i
g
n
a
l
.
T
h
e
ad
v
an
ce
d
i
mag
i
n
g
t
ech
n
i
q
u
e
s
can
b
e
u
s
e
d
t
o
d
i
a
g
n
o
s
e
l
eft
v
en
t
ri
c
u
l
ar
h
y
p
ert
ro
p
h
y
,
b
u
t
i
t
l
ead
s
t
o
t
i
m
e
-
co
n
s
u
mi
n
g
an
d
mo
re
e
x
p
e
n
s
i
v
e.
T
h
i
s
p
r
o
p
o
s
e
d
t
ech
n
i
q
u
e
o
v
erco
me
s
t
h
e
s
e
f
i
s
s
u
es
an
d
may
s
erv
e
as
an
effi
c
i
en
t
t
o
o
l
t
o
d
i
a
g
n
o
s
e
t
h
e
L
V
H
d
i
s
ea
s
e.
T
h
e
L
V
H
cau
s
es
ch
a
n
g
e
s
i
n
t
h
e
p
a
t
t
er
n
s
o
f
E
CG
s
i
g
n
a
l
w
h
i
ch
i
n
cl
u
d
e
s
R
w
av
e,
Q
RS
an
d
T
w
av
e.
T
h
i
s
p
r
o
p
o
s
e
d
ap
p
ro
ac
h
i
d
en
t
i
f
i
es
t
h
e
ch
a
n
g
e
s
i
n
t
h
e
p
at
t
ern
an
d
ex
t
ract
s
t
h
e
t
emp
o
ral
,
s
p
a
t
i
a
l
a
n
d
s
t
at
i
s
t
i
ca
l
feat
u
res
o
f
t
h
e
E
CG
s
i
g
n
al
u
s
i
n
g
w
i
n
d
o
w
ed
f
i
l
t
eri
n
g
t
ech
n
i
q
u
e.
T
h
es
e
fea
t
u
re
s
w
ere
ap
p
l
i
ed
t
o
t
h
e
co
n
v
e
n
t
i
o
n
al
cl
a
s
s
i
fi
er
a
n
d
al
s
o
t
o
t
h
e
n
e
u
ral
n
e
t
w
o
rk
cl
as
s
i
f
i
er
w
i
t
h
t
h
e
mo
d
i
f
i
ed
w
e
i
g
h
t
s
u
s
i
n
g
a
g
en
et
i
c
al
g
o
r
i
t
h
m.
T
h
e
w
ei
g
h
t
s
w
ere
mo
d
i
f
i
ed
b
y
c
o
mb
i
n
i
n
g
t
h
e
cr
o
s
s
o
v
er
o
p
erat
o
r
s
s
u
ch
a
s
cro
s
s
o
v
er
ari
t
h
met
i
c
an
d
cro
s
s
o
v
er
t
w
o
-
p
o
i
n
t
o
p
erat
o
r.
T
h
e
re
s
u
l
t
s
w
ere
co
m
p
ared
w
i
t
h
t
h
e
v
ar
i
o
u
s
cl
as
s
i
fi
ers
an
d
t
h
e
p
erf
o
rman
ce
o
f
t
h
e
n
e
u
ral
n
et
w
o
r
k
w
i
t
h
t
h
e
mo
d
i
f
i
ed
w
ei
g
h
t
s
u
s
i
n
g
a
g
en
e
t
i
c
a
l
g
o
ri
t
h
m
i
s
o
u
t
p
erfo
rme
d
.
T
h
e
accu
rac
y
o
f
t
h
e
w
e
i
g
h
t
s
mo
d
i
fi
ed
feed
f
o
rw
ar
d
n
e
u
ra
l
n
et
w
o
r
k
i
s
9
7
.
5
%
.
K
e
y
w
o
r
d
s
:
E
C
G
s
ignal
F
e
e
df
or
wa
r
d
ne
ur
a
l
ne
twor
k
Ge
ne
ti
c
a
lgor
it
hm
L
e
f
t
ve
ntr
icula
r
hype
r
t
r
ophy
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
ce
n
s
e
.
C
or
r
e
s
pon
din
g
A
u
th
or
:
D.
J
ude
He
manth,
De
pa
r
tm
e
nt
of
E
lec
tr
onics
a
nd
C
omm
unica
ti
on
E
n
ginee
r
ing,
Ka
r
unya
I
ns
ti
tut
e
of
T
e
c
hnology
a
nd
S
c
ienc
e
s
,
Ka
r
unya
Na
ga
r
,
C
oim
ba
tor
e
,
T
a
mi
l
Na
du,
I
ndia
.
E
mail:
judehe
manth@ka
r
unya
.
e
du
1.
I
NT
RODU
C
T
I
ON
T
he
pa
ti
e
nts
who
s
uf
f
e
r
e
d
f
r
om
lef
t
ve
ntr
icula
r
hype
r
tr
ophy
may
lea
d
to
the
r
is
k
of
mor
talit
y
.
S
udde
n
c
a
r
diac
de
a
th
may
a
ls
o
c
a
us
e
s
be
c
a
us
e
of
lef
t
ve
nt
r
icula
r
h
ype
r
t
r
ophy
[
1
]
.
L
e
f
t
ve
ntr
icula
r
hype
r
tr
op
hy
is
a
r
is
ing
due
to
p
r
olonged
high
blood
pr
e
s
s
ur
e
.
T
he
high
blood
pr
e
s
s
ur
e
is
a
ls
o
c
a
ll
e
d
a
s
c
a
r
diovas
c
ular
s
yndr
ome
[
2]
.
I
t
make
s
the
c
ha
mber
wa
ll
s
to
be
thi
c
k
a
nd
it
e
nlar
ge
s
the
c
ha
mber
.
I
t
is
a
ls
o
s
tate
d
that
the
incr
e
a
s
e
in
lef
t
ve
ntr
icula
r
m
a
s
s
lea
d
s
to
c
ha
nge
s
in
the
s
tr
uc
tur
e
of
myoca
r
d
ium
[
3]
.
T
he
f
a
c
tor
s
in
f
luenc
e
s
the
L
VH
dis
e
a
s
e
a
r
e
a
ge
,
incr
e
a
s
e
in
blood
pr
e
s
s
ur
e
,
glucos
e
int
oler
a
nc
e
a
nd
obe
s
it
y
[
4]
.
I
t
a
lt
e
r
s
ve
ntr
icle
r
e
polar
iza
ti
on
a
nd
de
polar
iza
ti
on
whic
h
lea
ds
to
the
c
ha
nge
s
in
QR
S
a
nd
T
pa
tt
e
r
ns
in
the
E
C
G
s
ignal
[
5]
.
T
he
va
r
i
ous
L
VH
-
E
C
G
c
r
it
e
r
ia
we
r
e
r
e
por
ted
whic
h
include
s
Co
r
ne
ll
Voltage
s
,
R
omhi
ll
t
E
s
tee
s
a
nd
S
okolov
L
y
on
(
S
L
)
[
6]
.
T
he
E
C
G
is
a
c
omm
on
method
on
the
a
c
c
ount
of
wide
a
va
il
a
bil
it
y
[
7]
.
T
hus
,
the
e
lec
tr
oc
a
r
diogr
a
phy
c
a
n
be
us
e
d
a
s
a
tool
to
diagnos
e
L
VH
in
a
n
e
a
r
ly
s
tage
.
T
he
L
VH
c
a
n
be
diagnos
e
d
us
ing
im
a
ging
modalit
ies
with
a
dva
nc
e
d
tec
hniques
.
T
his
method
is
high
pr
ice
d
a
nd
the
pr
oc
e
s
s
is
de
laye
d.
T
he
s
hor
tage
of
f
inanc
ial
r
e
s
our
c
e
s
to
a
c
quir
e
the
e
c
hoc
a
r
di
ogr
a
phic
f
a
c
il
it
y
with
s
kil
led
pe
r
s
onne
l
is
a
c
ha
ll
e
nging
is
s
ue
in
the
hos
pit
a
ls
pr
e
s
e
nt
in
the
r
ur
a
l
a
r
e
a
[
8]
.
T
o
o
ve
r
c
ome
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
3
,
J
une
2020:
1
2
8
5
-
1
29
1
1286
thes
e
is
s
ue
s
,
a
s
uit
a
ble
a
lgor
it
hm
is
pr
opos
e
d
to
diagnos
e
lef
t
ve
ntr
icula
r
hype
r
tr
oph
y
us
ing
E
C
G
s
ignal.
T
he
va
r
ious
Ne
ur
a
l
Ne
twor
k
tec
hniques
a
r
e
wide
ly
us
e
d
to
diagnos
e
c
a
r
diovas
c
ular
e
ve
nts
[
9]
.
I
n
thi
s
r
e
s
e
a
r
c
h
wor
k
,
the
c
las
s
if
ica
ti
on
of
L
VH
dis
e
a
s
e
is
im
pleme
nted
us
ing
f
e
e
df
or
wa
r
d
ne
ur
a
l
ne
twor
k
with
modi
f
ied
we
ight
s
.
T
he
we
ight
s
w
e
r
e
modi
f
ied
by
us
ing
a
c
r
os
s
ove
r
ope
r
a
tor
in
a
ge
ne
ti
c
a
lgor
it
hm.
S
e
ve
r
a
l
r
e
s
e
a
r
c
he
r
s
h
a
ve
uti
li
z
e
d
a
ge
n
e
ti
c
a
lgor
it
hm
a
nd
a
f
e
e
df
or
wa
r
d
ne
ur
a
l
ne
twor
k
f
or
s
ignal
pr
oc
e
s
s
ing
a
ppli
c
a
ti
ons
.
Z
e
inab
Ar
a
ba
s
a
di
e
t
a
l
.
p
r
opos
e
d
a
hybr
id
ne
ur
a
l
ne
twor
k
a
nd
ge
ne
ti
c
a
lgor
it
hm
f
or
the
diagnos
is
of
he
a
r
t
dis
e
a
s
e
s
.
T
he
ne
ur
a
l
ne
twor
k
pe
r
f
or
manc
e
wa
s
inc
r
e
a
s
e
d
by
10%
[
10]
.
Ar
p
it
B
ha
r
dwa
j
e
t
a
l
pr
opos
e
d
a
ne
twor
k
model
f
o
r
mul
t
icla
s
s
c
las
s
if
ica
ti
on.
T
his
a
lgor
it
hm
pr
ovides
a
5
%
to
8
%
be
tt
e
r
r
e
s
ult
[
11
]
.
Kinjal
J
a
da
v
e
t
a
l
r
e
por
ted
that
the
ne
ur
a
l
ne
twor
k
t
r
a
ini
ng
may
lea
d
to
loca
l
mi
nim
a
a
nd
s
low
c
onve
r
ge
nc
e
.
T
o
ove
r
c
ome
thi
s
is
s
ue
a
uthor
a
ppli
e
d
a
ge
ne
ti
c
a
lgor
it
hm
f
or
global
opti
mi
z
a
ti
on
[
12
]
.
Z
he
n
-
Guo
C
he
e
t
a
l
c
ompar
e
d
the
ne
ur
a
l
ne
twor
k
a
nd
ge
ne
ti
c
a
lgor
it
hm
pe
r
f
or
manc
e
a
nd
f
ound
that
the
B
P
N
pe
r
f
or
ms
be
tt
e
r
[
13]
.
J
a
na
ti
I
dr
is
s
i
e
t
a
l
.
opti
mi
z
e
d
a
n
e
ur
a
l
ne
twor
k
ba
s
e
d
on
meta
he
ur
is
ti
c
a
lgor
it
hm
by
f
o
r
mul
a
ti
ng
mul
ti
-
objec
ti
ve
mathe
matica
l
f
unc
ti
on
[
14]
.
Z
e
L
i
e
t
a
l
.
r
e
por
ted
a
Ge
ne
ti
c
Algor
it
hm
ba
c
kpr
o
pa
ga
ti
on
ne
twor
k
whic
h
incr
e
a
s
e
s
the
c
onve
r
ge
nc
e
s
pe
e
d
[
15]
.
Hongqia
ng
L
i
e
t
a
l
.
r
e
po
r
ted
the
GA
-
B
P
NN
ne
twor
k
f
or
the
s
ignal
c
las
s
if
ica
ti
on
i
t
pr
oduc
e
d
ve
r
y
go
od
r
e
s
ult
s
[
16]
.
A
r
pit
B
ha
r
dwa
j
e
t
a
l
.
de
ve
lope
d
a
ne
w
c
r
os
s
ove
r
a
nd
mut
a
ti
on
ope
r
a
tor
to
s
olve
a
c
las
s
if
ica
ti
on
pr
oblem
[
17
]
.
V.
S
.
R
.
Kuma
r
i
e
t
a
l
r
e
po
r
ted
that
the
ge
ne
ti
c
a
lgor
i
thm
opt
im
ize
s
th
e
lea
r
n
ing
r
a
te
to
c
las
s
if
y
a
r
r
hyth
mi
a
dis
e
a
s
e
[
18]
.
Ali
B
a
ha
dor
i
nia
e
t
a
l
.
de
ve
loped
a
P
S
O
a
nd
ge
ne
ti
c
a
lgo
r
it
hm
-
ba
s
e
d
ne
ur
a
l
ne
twor
k
f
o
r
the
c
las
s
if
ica
ti
on
o
f
a
r
r
hythm
i
a
s
[
19]
.
M
a
ns
our
a
S
e
kka
l
e
t
a
l
de
ve
loped
a
ne
ur
o
-
ge
n
e
ti
c
a
lgor
it
hm
to
c
las
s
if
y
c
a
r
diac
a
r
r
hythm
ia
d
is
e
a
s
e
[
20]
.
M
a
na
b
Kuma
r
Da
s
e
t
a
l
us
e
d
S
tr
a
ns
f
or
m
to
e
xtr
a
c
t
the
f
e
a
tur
e
s
a
nd
opti
mi
z
e
d
us
ing
a
ge
ne
ti
c
a
lgo
r
it
hm
to
obtain
a
be
tt
e
r
r
e
s
ult
[
21]
.
S
e
c
ti
on
2
e
xplains
the
d
a
tas
e
t,
M
e
thodol
ogy,
E
C
G
pr
e
-
pr
oc
e
s
s
ing
a
nd
E
C
G
s
ignal
a
na
lyzing
a
nd
C
l
a
s
s
if
ica
ti
on
tec
hniques
.
F
ur
ther
,
t
his
a
r
ti
c
le
r
e
s
ult
s
a
r
e
dis
c
us
s
e
d
a
nd
c
onc
luded
in
s
e
c
ti
on
3
a
nd
s
e
c
ti
on
4.
2.
RE
S
E
AR
CH
M
E
T
HO
D
T
his
s
e
c
ti
on
is
de
s
c
r
ibed
with
the
da
tas
e
t
uti
li
z
e
d
in
thi
s
s
tudy,
metho
dology,
s
ignal
pr
e
-
pr
oc
e
s
s
ing,
a
nd
f
e
a
tur
e
e
xtr
a
c
ti
on
a
nd
c
las
s
if
ica
ti
on
methods
.
2.
1.
E
CG
s
ign
al
d
at
as
e
t
T
he
P
T
B
Dia
gnos
ti
c
E
C
G
da
taba
s
e
is
uti
li
z
e
d
in
thi
s
wor
k
to
a
c
quir
e
12
lea
d
E
C
G
s
ignals
of
he
a
lt
hy
pe
r
s
ons
a
nd
the
pa
ti
e
nts
who
a
r
e
a
f
f
e
c
ted
by
lef
t
ve
ntr
icula
r
hype
r
tr
ophy
.
T
h
is
da
taba
s
e
a
ls
o
c
ontains
the
E
C
G
s
ignal
with
bundle
b
r
a
nc
h
block,
myoca
r
dit
is
,
a
nd
c
a
r
diom
yopa
thy
[
22]
.
T
he
s
a
mpl
ing
f
r
e
que
nc
y
of
the
s
ignal
is
1000
Hz
.
2.
2.
M
e
t
h
od
ology
T
he
F
igu
r
e
1
s
hows
the
f
low
of
p
r
opos
e
d
wor
k
to
diagnos
e
lef
t
ve
ntr
icula
r
hype
r
tr
ophy
.
I
t
include
s
f
e
a
tur
e
e
xtr
a
c
ti
on
a
nd
c
las
s
if
ica
ti
on.
T
he
f
a
s
t
F
our
ier
tr
a
ns
f
or
m
wa
s
im
pleme
nted
to
r
e
move
the
low
f
r
e
que
nc
ies
a
nd
inver
s
e
f
a
s
t
F
our
ier
tr
a
ns
f
or
m
wa
s
a
ppli
e
d
to
r
e
s
tor
e
the
E
C
G
s
ignal.
T
his
wa
s
f
oll
owe
d
by
a
windowe
d
f
il
ter
to
de
ter
mi
ne
the
loca
l
maxima.
T
he
s
ize
of
the
f
il
ter
wa
s
a
djus
ted
a
nd
f
il
ter
ing
is
r
e
pe
a
ted
to
identif
y
the
pe
a
ks
pr
e
s
e
nt
in
the
s
ignal.
T
he
tempor
a
l,
s
pe
c
tr
a
l
a
nd
s
tatis
ti
c
a
l
f
e
a
tur
e
s
we
r
e
obtaine
d.
T
he
s
e
f
e
a
tur
e
s
we
r
e
f
e
d
to
the
va
r
ious
c
las
s
if
ier
s
to
diagnos
e
the
L
VH
dis
e
a
s
e
.
F
igur
e
1.
Dia
gnos
is
of
L
e
f
t
Ve
ntr
icula
r
Hype
r
tr
op
hy
f
r
om
E
C
G
s
ignal
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
T
r
aini
ng
fee
dfor
w
ar
d
ne
ur
al
ne
tw
or
k
us
ing
ge
ne
ti
c
al
g
o
r
it
h
m…
(
J
.
R
e
v
athi
)
1287
2.
3.
E
CG
s
ign
al
p
r
e
-
p
r
oc
e
s
s
in
g
T
he
E
C
G
s
ignals
a
r
e
c
or
r
upted
with
the
a
r
ti
f
a
c
t
s
due
to
other
moni
tor
ing
de
vice
s
,
c
a
bles
[
23]
.
T
he
e
li
mi
na
ti
on
of
thes
e
a
r
ti
f
a
c
ts
is
e
s
s
e
nti
a
l
to
diagnos
e
the
dis
e
a
s
e
mor
e
a
c
c
ur
a
tely.
T
he
F
F
T
is
im
pleme
nted
to
r
e
move
the
low
-
f
r
e
que
nc
y
nois
e
pr
e
s
e
nt
in
the
s
ignal.
2.
4.
F
e
at
u
r
e
e
xt
r
ac
t
ion
T
he
e
s
s
e
nti
a
l
f
e
a
tur
e
s
that
a
r
e
e
xtr
a
c
ted
to
diagnos
e
L
VH
a
r
e
va
r
ianc
e
,
kur
tos
is
,
f
o
r
m
f
a
c
tor
,
a
ve
r
a
ge
He
a
r
t
r
a
te,
mea
n
R
R
int
e
r
va
l,
R
R
int
e
r
va
l
r
o
ot
mea
n
s
qua
r
e
dis
tanc
e
,
s
tanda
r
d
de
v
iation,
h
e
a
r
t
r
a
te
va
r
iabili
ty,
a
nd
powe
r
s
pe
c
tr
a
l
E
nt
r
opy.
I
n
th
is
s
t
udy,
the
E
C
G
lea
ds
s
uc
h
a
s
L
e
a
d
I
,
I
I
a
nd
I
I
I
,
a
V
L
,
V1
to
V6
obtaine
d
f
r
om
the
P
T
B
da
taba
s
e
.
T
he
low
-
f
r
e
que
nc
y
s
ignals
r
e
moved
by
us
ing
a
f
a
s
t
F
our
ier
tr
a
ns
f
or
m.
T
his
is
f
oll
owe
d
by
a
n
i
nve
r
s
e
f
a
s
t
F
our
ie
r
t
r
a
ns
f
or
m
to
r
e
s
tor
e
the
E
C
G
s
ignal
a
s
s
hown
in
F
igu
r
e
2
.
F
igur
e
2.
E
C
G
s
ignal
a
f
ter
a
pplyi
ng
F
F
T
T
his
is
f
oll
owe
d
by
a
windowe
d
f
il
ter
ing
tec
hni
que
to
f
ind
the
loca
l
maxima.
T
h
is
he
lps
to
f
ind
the
pe
a
ks
of
the
s
ignal.
T
he
window
s
ize
s
we
r
e
a
djus
ted
f
il
ter
ing
is
a
ppli
e
d
to
the
s
ignal
to
e
xt
r
a
c
t
f
e
a
tur
e
s
s
uc
h
a
s
kur
tos
is
,
f
or
m
f
a
c
tor
,
a
ve
r
a
ge
He
a
r
t
r
a
t
e
,
mea
n
R
R
int
e
r
va
l
,
R
R
int
e
r
va
l
r
oot
mea
n
s
qua
r
e
dis
tanc
e
,
s
tanda
r
d
de
viation,
he
a
r
t
r
a
te
va
r
iabili
ty
,
a
nd
pow
e
r
s
pe
c
tr
a
l
E
ntr
opy
.
T
his
is
f
u
r
ther
a
ppli
e
d
to
the
va
r
ious
c
las
s
if
ier
to
diagnos
e
the
dis
e
a
s
e
.
2.
5.
Clas
s
if
icat
ion
I
n
thi
s
r
e
s
e
a
r
c
h
wor
k,
c
las
s
if
ier
include
s
K
-
ne
a
r
e
s
t
ne
ighbor
(
KN
N)
,
s
uppor
t
ve
c
tor
mac
hine
(
S
VM
)
,
f
e
e
df
o
r
wa
r
d
ne
twor
k
a
nd
modi
f
ied
we
ight
f
e
e
df
or
wa
r
d
ne
twor
k
wa
s
e
mpl
oye
d
f
or
d
iagnos
is
a
nd
c
las
s
if
ica
ti
on
of
L
VH
.
2.
5.
1.
S
u
p
p
or
t
ve
c
t
or
m
ac
h
in
e
T
his
c
las
s
if
ier
s
e
gr
e
ga
tes
the
f
e
a
tur
e
by
plot
ti
ng
th
e
hype
r
plane
that
s
e
pa
r
a
tes
a
ll
the
f
e
a
tur
e
s
o
f
one
c
las
s
to
a
nother
.
S
VM
e
mp
loys
mapping
to
c
onv
e
r
t
nonli
ne
a
r
f
e
a
tu
r
e
s
to
high
dim
e
ns
ions
[
24]
.
T
he
S
VM
is
a
na
lyze
d
us
ing
the
Ga
us
s
ian
ke
r
ne
l
f
unc
ti
on
in
thi
s
wo
r
k
[
25
]
.
T
he
a
dva
ntage
is
that
it
is
f
lexible.
I
t
s
uppor
ts
c
ompl
e
x
models
.
T
his
f
unc
ti
on
is
uti
li
z
e
d
by
the
e
qua
ti
on
is
;
(
,
)
=
−
‖
−
‖
2
2
2
(1
)
2.
5.
2.
K
-
Ne
ar
e
s
t
n
e
igh
b
or
c
las
s
if
ier
T
he
c
las
s
if
ier
K
-
ne
a
r
e
s
t
ne
ighbor
is
a
n
ins
tant
b
a
s
e
d
lea
r
ning
method
[
26]
.
I
t
is
c
ons
ider
e
d
to
be
e
f
f
e
c
ti
ve
,
to
de
c
r
e
a
s
e
the
mi
s
c
las
s
if
ica
ti
on
of
e
r
r
or
,
whe
n
ther
e
is
a
la
r
ge
number
of
c
ompo
ne
nts
in
the
tr
a
ini
ng
s
e
t
[
27]
.
Dur
ing
c
las
s
if
ica
ti
on,
a
g
r
ou
p
of
'k
'
f
e
a
tur
e
s
f
r
om
the
tr
a
ini
ng
s
e
t
a
r
e
pr
opos
e
d
that
a
r
e
c
los
e
to
the
f
e
a
tur
e
s
that
a
r
e
to
be
tes
ted
[
28
]
.
T
he
c
las
s
if
ier
pe
r
f
or
manc
e
is
ba
s
e
d
on
the
va
lue
K
a
nd
dis
tanc
e
metr
ics
.
I
n
th
is
s
tudy,
thes
e
va
lues
a
r
e
im
pleme
nted
by
c
r
os
s
-
va
li
da
ti
on
f
or
the
a
s
s
e
s
s
ment
of
c
las
s
if
ica
ti
on
a
c
c
ur
a
c
y
[
27]
.
F
or
tes
ti
ng,
a
s
ingl
e
da
ta
s
a
mpl
e
is
e
mpl
oye
d
f
or
va
li
da
ti
on
a
nd
r
e
maining
K
-
1
da
ta
s
a
mpl
e
s
a
r
e
us
e
d
f
or
tr
a
ini
ng
[
27]
.
T
his
is
r
e
pe
a
ted
a
K
ti
m
e
with
e
a
c
h
of
the
k
da
ta
s
a
mpl
e
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
3
,
J
une
2020:
1
2
8
5
-
1
29
1
1288
2.
5.
3.
F
e
e
d
f
or
war
d
n
e
u
r
al
n
e
t
wor
k
wi
t
h
ge
n
e
t
ic
algor
it
h
m
T
his
wor
k
pr
opos
e
d
that
the
ge
ne
ti
c
a
lgor
i
thm
is
e
mpl
oye
d
to
tr
a
in
the
f
e
e
d
-
f
or
wa
r
d
ne
ur
a
l
ne
twor
k
.
T
he
c
las
s
if
ica
ti
on
wa
s
e
mpl
oye
d
us
ing
the
L
e
ve
nbe
r
g
-
M
a
r
qua
r
t
ne
twor
k,
a
s
c
a
led
c
onjugate
gr
a
dient
ba
c
kpr
opa
ga
ti
on
ne
ur
a
l
ne
twor
k.
T
he
tot
a
l
nu
mber
of
ne
ur
ons
pr
e
s
e
nted
in
the
hidden
laye
r
is
12
,
15
,
a
nd
20.
T
he
tr
ial
a
nd
e
r
r
or
method
is
im
pleme
nted
f
or
t
he
s
e
lec
ti
on
of
ne
ur
ons
in
the
hidden
laye
r
.
I
n
t
he
output
laye
r
,
one
ne
ur
on
wa
s
e
mpl
oye
d
to
diagnos
e
the
L
VH
ba
s
e
d
E
C
G
s
ignal.
70
%
o
f
the
d
a
tas
e
t
is
im
pl
e
mente
d
f
or
t
r
a
ini
ng
a
nd
15
%
o
f
the
da
tas
e
t
is
us
e
d
f
or
va
l
i
da
ti
on
a
nd
the
las
t
15
%
o
f
da
ta
is
us
e
d
f
or
tes
ti
ng.
T
his
pe
r
f
or
manc
e
of
thes
e
c
las
s
if
ier
s
wa
s
c
ompar
e
d
with
the
ne
ur
a
l
ne
twor
k
c
las
s
if
ier
with
the
modi
f
ied
we
ight
s
.
I
n
thi
s
r
e
s
e
a
r
c
h
wor
k,
the
we
ight
s
of
the
f
e
e
df
or
wa
r
d
ne
u
r
a
l
ne
twor
ks
we
r
e
upda
ted
ba
s
e
d
on
the
ge
ne
ti
c
a
lgor
it
hm.
I
n
a
ge
ne
ti
c
a
lgor
it
hm,
a
population
is
c
ons
ider
e
d
a
s
a
we
ight
.
T
he
c
hr
omos
omes
a
r
e
p
r
e
s
c
r
ibed
by
us
ing
a
n
o
bjec
ti
ve
f
unc
ti
on
.
T
his
is
f
oll
owe
d
by
a
s
e
lec
t
ion
of
two
c
hr
omos
omes
to
pe
r
f
or
m
the
ge
ne
ti
c
ope
r
a
ti
on
.
T
h
e
of
f
s
pr
ing
is
ge
ne
r
a
ted
whic
h
a
lt
e
r
s
the
ini
ti
a
l
po
pulation.
T
his
is
r
e
pe
a
ted
unti
l
the
de
s
ir
e
d
output
is
a
c
hieve
d.
T
he
objec
ti
ve
f
unc
ti
on
is
s
tate
d
a
s
a
no
r
malize
d
mea
n
s
qua
r
e
e
r
r
or
.
e
(
x)=
(
O
i
-
ti)
2
(2
)
(
)
=
(
(
)
)
(
(
)
(
3)
W
he
r
e
Oi
r
e
pr
e
s
e
nts
the
de
s
ir
e
d
output
,
t
i
r
e
p
r
e
s
e
nts
output
obtaine
d
f
r
om
ne
t.
ALGORITHM
Step 1: The size of the weight matrix is defined and considered as a population.
Step
2:
Population
set
is
produced.
Each
component
present
i
n
this
set
represents
weights.
This weight depends on the number of layers in the network
[29].
Step 3: Every matrix of weights is applied to the fitness function.
Step 4: The weight matrix with the best
solution is chosen for the modification.
Step
5:
Genetic
operations
mainly
based
on
crossover
and
mutation.
This
process
creates
new
offspring.
Step 6: Crossover operator:
In
this
work
crossover
arithmetic
and
the
crossover,
two
-
point
is
implemented
to
generate
new
offspring.
The
obtained
results
were
taken
and
the
new
offspring
is
generated
which
is
followed by mutation.
The
crossover
arithmetic
is
a
linear
comb
ination
of
two
parents
who
are
selected
r
andomly.
The
crossover
two
-
point
copied
values
base
d
on
the
two
points.
In
-
between
these
two
points
the genes get swapped. The weights are further computed with the formula (X+X1)/3.
Where
X
represents
the
weights
obtained
fro
m
the
crossover
arithmetic
and
X1
represents
the weights obtained from the crossover two
-
point.
Step
7:
Stopping
criteria
is
based
on
the
number
of
generations
and
the
convergence
rate
the network reaches the objective.
T
hus
f
e
e
df
o
r
wa
r
d
ne
ur
a
l
ne
twor
k
wa
s
im
pleme
n
ted
by
a
djus
ti
ng
the
we
ight
s
of
ne
ur
a
l
ne
twor
ks
us
ing
a
ge
ne
ti
c
a
lgor
it
hm.
3.
RE
S
UL
T
S
AN
D
AN
AL
YSI
S
I
n
thi
s
r
e
s
e
a
r
c
h
wor
k
,
the
f
a
s
t
F
our
ier
t
r
a
ns
f
or
m
wa
s
a
ppli
e
d
to
the
s
ignals
to
e
li
mi
na
te
lowe
r
f
r
e
que
nc
ies
.
T
his
is
f
oll
owe
d
by
inver
s
e
f
a
s
t
F
our
ier
tr
a
ns
f
or
m
to
r
e
s
tor
e
the
E
C
G
s
ignal.
F
ur
ther
,
the
windowe
d
f
il
te
r
wa
s
im
pleme
nted
to
f
ind
l
oc
a
l
maxima.
T
he
pe
a
ks
of
the
s
ignals
we
r
e
id
e
nti
f
ied.
T
he
window
s
ize
is
a
djus
ted
a
nd
a
ga
in
f
il
te
r
a
ppli
e
d
to
f
ind
the
pe
a
ks
.
I
t
im
pr
ove
s
the
qua
l
it
y
of
the
f
il
ter
ing.
T
he
f
e
a
tur
e
s
e
xtr
a
c
ted
f
r
om
thi
s
s
tudy
include
s
va
r
ianc
e
,
kur
tos
is
,
f
or
m
f
a
c
tor
,
a
ve
r
a
ge
He
a
r
t
r
a
t
e
,
mea
n
R
R
int
e
r
va
l,
R
R
int
e
r
va
l
r
oot
mea
n
s
qua
r
e
dis
tanc
e
,
s
tanda
r
d
de
viation,
he
a
r
t
r
a
te
va
r
iabili
ty
,
a
n
d
powe
r
s
pe
c
tr
a
l
e
ntr
opy.
T
his
a
c
quir
e
d
f
e
a
tur
e
s
f
e
d
to
th
e
va
r
ious
c
las
s
if
ier
s
to
de
te
r
mi
ne
L
VH
ba
s
e
d
E
C
G
s
ignal
f
r
om
the
he
a
lt
hy
indi
vidual
E
C
G
s
igna
l.
T
he
432
E
C
G
s
ignals
a
c
quir
e
d
f
r
om
the
da
taba
s
e
whic
h
include
s
both
L
VH
a
nd
he
a
lt
hy
E
C
G
s
ignals
.
T
he
tot
a
l
f
e
a
tur
e
s
obtaine
d
f
r
om
thi
s
s
tudy
we
r
e
(
432x12=
5184
f
e
a
tur
e
s
)
.
T
he
s
e
f
e
a
tur
e
s
we
r
e
a
ppli
e
d
to
the
c
las
s
if
ier
s
s
uc
h
a
s
S
VM
,
KN
N,
L
e
ve
nbe
r
g
-
M
a
r
qua
r
dt
ne
ur
a
l
ne
two
r
k
(
L
M
NN
)
,
s
c
a
led
c
onjugate
gr
a
dient
ne
ur
a
l
ne
twor
k
(
S
C
G
NN
)
,
a
nd
we
ight
s
modi
f
ied
f
e
e
df
or
wa
r
d
ne
ur
a
l
ne
twor
ks
us
ing
GA
.
T
his
e
xpe
r
im
e
nt
s
hows
the
potenc
y
of
modi
f
ied
GA
ba
s
e
d
f
e
e
df
or
wa
r
d
ne
ur
a
l
ne
twor
k
.
I
n
GA
,
the
c
r
os
s
ove
r
ope
r
a
tor
wa
s
im
p
leme
nted
whic
h
include
s
a
r
it
h
metic
c
r
os
s
ove
r
a
nd
two
-
point
c
r
os
s
ove
r
to
c
r
e
a
te
a
ne
w
of
f
s
pr
ing.
F
ur
the
r
ne
w
we
ight
s
obtaine
d
by
c
ombi
ning
the
we
ight
s
f
r
om
thes
e
c
r
os
s
ove
r
ope
r
a
tor
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
T
r
aini
ng
fee
dfor
w
ar
d
ne
ur
al
ne
tw
or
k
us
ing
ge
ne
ti
c
al
g
o
r
it
h
m…
(
J
.
R
e
v
athi
)
1289
T
o
de
ter
mi
ne
the
c
ons
tr
uc
ti
ve
ne
s
s
of
e
xtr
a
c
ted
f
e
a
tur
e
s
,
a
r
e
c
e
iver
-
ope
r
a
t
ing
c
ha
r
a
c
ter
is
ti
c
a
na
lys
is
wa
s
im
pleme
nted.
T
he
a
r
e
a
unde
r
the
c
ur
ve
(
AU
C
)
de
notes
the
goodne
s
s
of
f
it
[
30]
.
T
he
a
ll
oc
a
ti
on
o
f
f
e
a
tur
e
ve
c
tor
s
is
s
im
il
a
r
in
two
c
las
s
e
s
f
or
the
AU
C
va
lu
e
is
0
.
5
[
30
]
.
T
he
a
ll
oc
a
ti
on
of
f
e
a
tur
e
s
of
two
c
las
s
e
s
doe
s
not
ove
r
lap
with
e
a
c
h
other
f
or
the
AU
C
va
lue
is
1
[
30]
.
T
his
va
lue
is
mo
r
e
than
0
.
5
then
the
gue
s
s
ing
model
is
good.
I
n
thi
s
wor
k,
the
f
ive
f
old
c
r
os
s
-
va
li
d
a
ti
on
s
c
he
me
is
e
mpl
oye
d
f
o
r
the
t
r
a
ini
ng
a
nd
tes
ti
ng
of
c
las
s
if
ier
s
.
T
he
pa
r
a
mete
r
s
s
uc
h
a
s
s
e
n
s
it
ivi
ty,
s
pe
c
if
icity,
a
nd
a
c
c
ur
a
c
y
a
r
e
e
va
luate
d
to
identif
y
the
pe
r
f
or
manc
e
of
the
c
las
s
if
ier
.
T
he
s
e
pa
r
a
mete
r
s
a
r
e
judged
by
identif
ying
t
r
ue
pos
it
ives
,
tr
ue
n
e
ga
ti
ve
s
,
f
a
ls
e
pos
it
ives
,
a
nd
f
a
ls
e
ne
ga
ti
v
e
va
lues
.
T
he
T
P
is
tr
ue
pos
it
ive
whic
h
indi
c
a
tes
the
numbe
r
o
f
e
ve
nts
matc
he
d
a
nd
F
N
is
f
a
ls
e
n
e
ga
ti
ve
whic
h
indi
c
a
tes
the
number
of
e
ve
nts
that
a
r
e
not
r
e
c
ognize
d
[
31]
.
T
he
F
P
is
f
a
ls
e
p
os
it
ive
whic
h
r
e
pr
e
s
e
nt
s
the
number
of
e
ve
nts
that
a
r
e
not
matc
he
d
a
nd
tr
ue
ne
ga
ti
ve
(
T
N)
r
e
pr
e
s
e
nts
the
e
ve
nts
that
a
r
e
e
xa
c
tl
y
r
e
c
ognize
d
a
s
not
de
f
e
c
ti
ve
s
[
31
]
.
T
a
ble
1
s
hows
the
c
las
s
if
ier
pe
r
f
or
manc
e
e
va
luate
d
in
ter
ms
of
s
e
ns
it
ivi
ty,
s
pe
c
if
icity,
a
c
c
ur
a
c
y
a
nd
a
r
e
a
unde
r
the
c
ur
ve
.
T
a
ble
1.
P
e
r
f
o
r
manc
e
of
the
c
las
s
if
ier
s
C
la
s
s
if
ie
r
s
N
umbe
r
of
n
e
ur
ons
S
e
ns
it
iv
it
y
(%)
S
pe
c
if
ic
it
y
(%)
A
c
c
ur
a
c
y
(%)
AUC
S
V
M
-
86.1
92
90.7
0.90
KNN
-
85.1
93.3
91.4
0.91
L
M
N
N
12
89.8
98.8
96.5
0.96
L
M
N
N
15
79.6
96.9
92.6
0.92
L
M
N
N
20
69.4
91.4
85.9
0.86
S
C
N
N
12
43.5
96.6
83.3
0.83
S
C
N
N
15
49.1
96.9
85.0
0.85
S
C
N
N
20
72.2
96.6
90.5
0.90
L
M
N
N
w
e
ig
ht
s
opt
im
iz
e
d us
in
g G
A
12
85.2
98.5
95.1
0.95
L
M
N
N
w
e
ig
ht
s
opt
im
iz
e
d us
in
g G
A
15
84.3
96.6
93.5
0.93
L
M
N
N
w
e
ig
ht
s
opt
im
iz
e
d us
in
g G
A
20
93.5
98.8
97.5
0.97
S
C
G
N
N
w
e
ig
ht
s
opt
im
iz
e
d us
in
g G
A
12
37
98.1
82.9
0.83
S
C
G
N
N
w
e
ig
ht
s
opt
im
iz
e
d us
in
g G
A
15
62
98.1
83.1
0.83
S
C
G
N
N
w
e
ig
ht
s
opt
im
iz
e
d us
in
g G
A
20
60.2
97.8
83.3
0.83
T
he
f
e
e
df
o
r
wa
r
d
NN
s
we
r
e
tr
a
ined
wi
th
12
,
15
a
nd
20
hidden
ne
ur
ons
.
T
he
r
e
s
ult
s
s
how
that
the
a
c
c
ur
a
c
y
of
L
M
NN
with
modi
f
ied
we
ight
s
is
incr
e
a
s
e
d
than
the
L
M
NN
without
modi
f
ied
we
ight
s
.
T
he
a
c
c
ur
a
c
y
of
the
L
e
ve
nbe
r
g
-
M
a
r
qua
r
dt
ne
ur
a
l
ne
twor
k
(
L
M
NN
)
s
ho
ws
be
tt
e
r
r
e
s
ult
s
than
s
c
a
led
c
onjugate
gr
a
dient
ne
ur
a
l
ne
twor
k
(
S
C
G
NN
)
.
T
h
e
a
c
c
ur
a
c
y
of
S
VM
a
nd
KN
N
is
low
a
s
c
ompar
e
d
to
L
M
NN
f
e
e
df
or
wa
r
d
ne
u
r
a
l
ne
twor
k
with
mod
if
ied
we
i
ghts
.
4.
CONC
L
USI
ON
I
n
the
c
ur
r
e
nt
s
c
e
na
r
io,
va
r
ious
im
a
ging
modalit
y
tec
hniques
us
e
d
to
diagnos
e
L
VH
.
B
ut
thi
s
lea
ds
to
a
de
lay
in
tr
e
a
tm
e
nt
a
nd
high
c
os
t.
He
nc
e
,
the
diagnos
is
of
L
VH
us
ing
a
n
E
C
G
s
ignal
is
a
ne
c
e
s
s
a
r
y
pa
r
t
of
e
f
f
e
c
ti
ve
tr
e
a
tm
e
nt.
T
his
s
tudy
invo
lves
the
a
na
lys
is
of
tempor
a
l,
s
pe
c
tr
a
l
a
nd
s
tatis
ti
c
a
l
f
e
a
tur
e
s
of
E
C
G
s
ignal.
T
he
c
las
s
if
ica
ti
on
of
dis
e
a
s
e
with
L
VH
c
ompar
e
d
with
S
VM
,
KN
N
a
nd
va
r
ious
f
e
e
df
or
wa
r
d
ne
ur
a
l
ne
twor
ks
with
a
nd
without
modi
f
ied
we
ig
hts
.
I
n
thi
s
wo
r
k,
Ge
ne
ti
c
a
lgo
r
it
hm
wa
s
im
plem
e
nted
to
opti
mi
z
e
the
we
ight
s
o
f
f
e
e
df
or
wa
r
d
NN
.
T
he
we
ight
s
modi
f
ied
ba
s
e
d
on
the
c
r
os
s
ove
r
ope
r
a
t
or
.
T
hus
,
it
pr
ove
s
that
the
L
e
ve
nbe
r
g
-
M
a
r
qua
r
dt
NN
we
ight
s
opti
mi
z
e
d
us
ing
GA
pe
r
f
o
r
m
be
tt
e
r
than
a
nother
c
las
s
i
f
ier
to
diagnos
e
L
VH
.
RE
F
E
RE
NC
E
S
[1
]
K
.
Po
rt
h
an
e
t
al
.
,
"
E
CG
l
eft
v
e
n
t
ri
c
u
l
ar
h
y
p
ert
r
o
p
h
y
a
s
a
ri
s
k
p
re
d
i
c
t
o
r
o
f
s
u
d
d
en
car
d
i
ac
d
eat
h
,
"
In
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
C
a
r
d
i
o
l
o
g
y,
v
o
l
.
2
7
6
,
p
p
.
1
2
5
-
1
2
9
,
2
0
1
9
.
[2
]
X
.
J
i
a
n
g
et
a
l
.
,
"
E
l
ect
r
o
card
i
o
g
ra
p
h
i
c
cri
t
er
i
a
fo
r
t
h
e
d
i
a
g
n
o
s
i
s
o
f
a
b
n
o
rmal
h
y
p
ert
e
n
s
i
v
e
car
d
i
ac
p
h
e
n
o
t
y
p
es
,
"
Th
e
Jo
u
r
n
a
l
o
f
Cl
i
n
i
ca
l
H
y
p
er
t
en
s
i
o
n
,
v
o
l
.
2
1
,
n
o
.
3
,
p
p
.
3
7
2
-
3
7
8
,
2
0
1
9
.
[3
]
A
.
L
i
n
h
ar
t
an
d
F.
Cecch
i
,
"
Co
mmo
n
p
res
e
n
t
at
i
o
n
o
f
rare
d
i
s
ea
s
es
:
L
eft
v
en
t
r
i
cu
l
ar
h
y
p
ert
r
o
p
h
y
a
n
d
d
i
a
s
t
o
l
i
c
d
y
s
fu
n
ct
i
o
n
,
"
In
t
er
n
a
t
i
o
n
a
l
Jo
u
r
n
a
l
o
f
Ca
r
d
i
o
l
o
g
y
,
v
o
l
.
2
5
7
,
p
p
.
3
4
4
-
3
5
0
,
2
0
1
8
.
[4
]
H
.
L
ee
et
al
.
,
"
Cl
i
n
i
ca
l
ch
arac
t
er
i
s
t
i
c
s
as
s
o
ci
a
t
ed
w
i
t
h
e
l
ect
r
o
card
i
o
g
rap
h
i
c
l
eft
v
e
n
t
ri
c
u
l
ar
h
y
p
er
t
ro
p
h
y
i
n
cl
i
n
i
cal
n
o
rm
o
t
e
n
s
i
v
e
s
w
i
t
h
o
u
t
a
h
i
s
t
o
ry
o
f
h
y
p
e
rt
e
n
s
i
o
n
:
A
cr
o
s
s
-
s
ec
t
i
o
n
a
l
s
t
u
d
y
,
"
Ko
r
ea
n
J
o
u
r
n
a
l
o
f
F
a
m
i
l
y
M
ed
i
c
i
n
e
,
v
o
l
.
4
0
,
n
o
.
2
,
p
p
.
1
0
6
-
1
1
5
,
2
0
1
9
.
[5
]
G
.
Sch
i
l
l
aci
et
al
.
,
"
Imp
ro
v
ed
el
ec
t
ro
car
d
i
o
g
ra
p
h
i
c
d
i
a
g
n
o
s
i
s
o
f
l
eft
v
e
n
t
r
i
cu
l
ar
h
y
p
ert
r
o
p
h
y
,
"
A
m
er
i
ca
n
Jo
u
r
n
a
l
o
f
Ca
r
d
i
o
l
o
g
y,
v
ol
.
7,
n
o
.
7
,
p
p
.
7
1
4
-
7
1
9
,
1
9
9
4
.
[6
]
W
.
B.
N
u
rd
i
n
,
H
.
J
an
u
ar
i
,
S.
Su
ry
an
i
,
an
d
S.
D
ew
an
g
,
"
Bes
t
L
V
H
-
E
CG
cri
t
er
i
a
fo
r
In
d
o
n
es
i
an
h
y
p
ert
e
n
s
i
v
es
,
"
Jo
u
r
n
a
l
o
f
P
h
ys
i
cs
:
Co
n
f
e
r
en
ce
S
er
i
es
,
v
o
l
.
1
2
4
2
,
n
o
.
1
,
p
p
.
1
-
4
,
2
0
1
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
3
,
J
une
2020:
1
2
8
5
-
1
29
1
1290
[7
]
V
.
N
o
ms
a
w
ad
i
an
d
R.
K
ri
t
t
ay
a
p
h
o
n
g
,
"
D
i
ag
n
o
s
t
i
c
p
erf
o
rman
ce
o
f
e
l
ect
r
o
card
i
o
g
rap
h
i
c
cri
t
eri
a
fo
r
l
eft
v
en
t
ri
c
u
l
a
r
h
y
p
ert
r
o
p
h
y
am
o
n
g
v
ar
i
o
u
s
b
o
d
y
ma
s
s
i
n
d
e
x
g
r
o
u
p
s
co
mp
ar
e
d
t
o
d
i
ag
n
o
s
i
s
b
y
car
d
i
ac
ma
g
n
e
t
i
c
re
s
o
n
a
n
ce
i
mag
i
n
g
,
"
A
n
n
a
l
s
o
f
N
o
n
i
n
v
a
s
i
ve
E
l
e
ct
r
o
c
a
r
d
i
o
l
o
g
y,
v
o
l
.
2
4
,
n
o
.
2
,
p
p
.
1
-
9
,
2
0
1
9
.
[8
]
J
.
Pi
n
t
o
,
P.
G
eo
rg
e,
an
d
N
.
H
e
g
d
e,
"
St
u
d
y
i
n
So
u
t
h
ern
I
n
d
i
a
amo
n
g
h
y
p
er
t
e
n
s
i
v
e
p
a
t
i
en
t
s
u
s
i
n
g
E
CG
t
o
s
creen
l
ef
t
v
en
t
ri
c
u
l
ar
h
y
p
ert
r
o
p
h
y
-
ca
n
w
e
d
o
i
t
i
n
r
u
ral
h
eal
t
h
ce
n
t
res
?,
"
Jo
u
r
n
a
l
o
f
Cl
i
n
i
ca
l
a
n
d
D
i
a
g
n
o
s
t
i
c
R
e
s
ea
r
ch
:
JC
D
R
,
v
o
l
.
8
,
n
o
.
3
,
p
p
.
59
-
62
,
2
0
1
4
.
[9
]
Z
.
N
.
A
l
i
m
b
ay
e
v
a
et
a
l
,
"
N
eu
ra
l
n
e
t
w
o
rk
a
n
al
y
s
i
s
o
f
el
e
ct
ro
car
d
i
o
g
ram
s
i
g
n
a
l
,
"
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
en
ce
o
f
Y
o
u
n
g
S
p
ec
i
a
l
i
s
t
s
o
n
M
i
c
r
o
/
Na
n
o
t
ech
n
o
l
o
g
i
e
s
a
n
d
E
l
ect
r
o
n
D
e
vi
ces
,
p
p
.
6
0
6
-
6
1
1
,
2
0
1
9
.
[1
0
]
Z
.
A
rab
as
ad
et
al
.
,
"
Co
mp
u
t
er
-
a
i
d
e
d
d
eci
s
i
o
n
mak
i
n
g
fo
r
h
ear
t
d
i
s
eas
e
d
e
t
ect
i
o
n
u
s
i
n
g
a
h
y
b
r
i
d
n
e
u
ra
l
n
et
w
o
r
k
-
G
en
et
i
c
al
g
o
ri
t
h
m,
"
Co
m
p
u
t
e
r
M
et
h
o
d
s
a
n
d
P
r
o
g
r
a
m
s
i
n
B
i
o
m
ed
i
ci
n
e,
v
o
l
.
1
4
1
,
p
p
.
1
9
-
2
6
,
2
0
1
7
.
[1
1
]
A
.
Bh
ard
w
a
j
,
A
.
T
i
w
ar
i
,
H
.
Bh
ard
w
a
j
,
an
d
A
.
Bh
ard
w
aj
,
"
A
g
en
et
i
cal
l
y
o
p
t
i
mi
z
ed
n
e
u
ral
n
e
t
w
o
rk
mo
d
el
fo
r
mu
l
t
i
-
c
l
as
s
cl
as
s
i
f
i
cat
i
o
n
,
"
E
xp
e
r
t
S
y
s
t
e
m
s
w
i
t
h
A
p
p
l
i
ca
t
i
o
n
s
,
v
o
l
.
60,
p
p
.
2
1
1
-
2
2
1
,
2
0
1
6
.
[1
2
]
K.
J
ad
av
an
d
M.
Pan
ch
al
,
"
O
p
t
i
m
i
zi
n
g
w
e
i
g
h
t
s
o
f
art
i
fi
ci
al
n
e
u
ral
n
e
t
w
o
rk
s
u
s
i
n
g
g
en
e
t
i
c
a
l
g
o
ri
t
h
m
,
"
In
t
er
n
a
t
i
o
n
a
l
Jo
u
r
n
a
l
o
f
A
d
va
n
ced
R
e
s
ea
r
ch
i
n
Co
m
p
u
t
e
r
S
c
i
en
ce
a
n
d
E
l
ect
r
o
n
i
c
s
E
n
g
i
n
ee
r
i
n
g
,
v
o
l
.
1
,
n
o
.
1
0
,
p
p
.
4
7
-
5
1
,
2
0
1
2
.
[1
3
]
Z
.
G
.
Ch
e
,
T
.
A
.
Ch
i
an
g
,
an
d
Z
.
H
.
Ch
e,
"
Feed
-
fo
rw
ard
n
eu
ral
n
e
t
w
o
rk
s
t
rai
n
i
n
g
:
a
co
mp
ari
s
o
n
b
e
t
w
ee
n
g
en
et
i
c
al
g
o
ri
t
h
m
an
d
b
ack
-
p
ro
p
ag
a
t
i
o
n
l
earn
i
n
g
al
g
o
ri
t
h
m,
"
In
t
er
n
a
t
i
o
n
a
l
Jo
u
r
n
a
l
o
f
In
n
o
v
a
t
i
ve
Co
m
p
u
t
i
n
g
,
In
f
o
r
m
a
t
i
o
n
,
a
n
d
C
o
n
t
r
o
l
,
v
o
l
.
1
0
,
p
p
.
5
8
3
9
-
5
8
5
0
,
2
0
1
1
.
[1
4
]
M
.
A
.
J
.
Id
ri
s
s
i
,
H
.
Ramch
o
u
n
,
Y
.
G
h
an
o
u
,
an
d
M
.
E
t
t
ao
u
i
l
,
"
G
en
et
i
c
al
g
o
r
i
t
h
m
fo
r
n
eu
ral
n
et
w
o
rk
arch
i
t
ect
u
re
o
p
t
i
mi
zat
i
o
n
,
"
3rd
In
t
er
n
a
t
i
o
n
a
l
C
o
n
f
er
e
n
ce
o
n
Lo
g
i
s
t
i
c
s
O
p
e
r
a
t
i
o
n
s
M
a
n
a
g
e
m
en
t
,
p
p
.
1
-
4,
2
0
1
6
.
[1
5
]
Z
.
L
i
,
D
.
Y
.
Su
n
,
J
.
L
i
,
an
d
Z
.
F
.
Li
,
"
So
ci
al
n
et
w
o
r
k
ch
a
n
g
e
d
et
ec
t
i
o
n
u
s
i
n
g
a
g
e
n
et
i
c
al
g
o
r
i
t
h
m
b
a
s
ed
b
ack
p
ro
p
ag
a
t
i
o
n
n
eu
ra
l
n
e
t
w
o
rk
mo
d
el
,
"
I
E
E
E
/A
C
M
In
t
er
n
a
t
i
o
n
a
l
Co
n
f
er
e
n
ce
o
n
A
d
va
n
ces
i
n
S
o
c
i
a
l
Net
w
o
r
ks
A
n
a
l
y
s
i
s
a
n
d
M
i
n
i
n
g
,
p
p
.
1
3
8
6
-
1
3
8
7
,
2
0
1
6
.
[1
6
]
H
.
L
i
,
D
.
Y
u
an
,
X
.
Ma,
D
.
Cu
i
,
an
d
L
.
Cao
,
"
G
en
et
i
c
al
g
o
ri
t
h
m
fo
r
t
h
e
o
p
t
i
m
i
zat
i
o
n
o
f
feat
u
re
s
an
d
n
e
u
ral
n
e
t
w
o
rk
s
i
n
E
CG
s
i
g
n
al
s
c
l
as
s
i
f
i
cat
i
o
n
,
"
S
ci
e
n
t
i
f
i
c
R
ep
o
r
t
s
,
v
o
l
.
7,
p
p
.
1
-
1
2
,
2
0
1
7
.
[1
7
]
A
.
Bh
ard
w
a
j
an
d
A
.
T
i
w
ari
,
"
Breas
t
ca
n
cer
d
i
ag
n
o
s
i
s
u
s
i
n
g
g
en
e
t
i
ca
l
l
y
o
p
t
i
m
i
ze
d
n
eu
ra
l
n
et
w
o
r
k
mo
d
el
,
"
E
x
p
er
t
S
ys
t
em
s
wi
t
h
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
4
2
,
n
o
.
1
0
,
p
p
.
4
6
1
1
-
4
6
2
0
,
2
0
1
5
.
[1
8
]
V
.
S.
R
.
K
u
mari
,
P.
R.
K
u
mar,
an
d
P.
Raj
es
h
,
"
O
p
t
i
mi
z
a
t
i
o
n
o
f
mu
l
t
i
-
l
ay
er
p
erce
p
t
r
o
n
n
eu
ra
l
n
et
w
o
r
k
u
s
i
n
g
g
en
et
i
c
al
g
o
ri
t
h
m
fo
r
arr
h
y
t
h
m
i
a
cl
as
s
i
f
i
cat
i
o
n
,
”
Co
m
m
u
n
i
ca
t
i
o
n
s
,
v
o
l
.
3
,
n
o
.
5
,
p
p
.
1
5
0
-
1
5
7
,
2
0
1
5
.
[1
9
]
A
.
Bah
a
d
o
r
i
n
i
a,
A
.
D
o
l
at
a
b
ad
i
,
an
d
A
.
H
aj
i
p
o
u
r,
"
A
h
y
b
r
i
d
i
zed
art
i
fi
c
i
al
n
eu
ra
l
n
et
w
o
r
k
an
d
o
p
t
i
m
i
za
t
i
o
n
al
g
o
ri
t
h
m
s
fo
r
t
h
e
d
i
ag
n
o
s
i
s
o
f
car
d
i
ac
arr
h
y
t
h
m
i
as
,
"
A
d
v
a
n
ce
s
i
n
C
o
m
p
u
t
er
S
ci
e
n
ce:
a
n
In
t
er
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
,
v
o
l
.
3
,
n
o
.
4
,
p
p
.
5
1
-
58,
2
0
1
4
.
[2
0
]
M.
Sek
k
al
a
n
d
M.
A
.
Ch
i
k
h
,
“N
eu
r
o
-
g
e
n
et
i
c
ap
p
r
o
ach
t
o
cl
a
s
s
i
fi
ca
t
i
o
n
o
f
card
i
ac
arrh
y
t
h
mi
a
s
”
Jo
u
r
n
a
l
o
f
M
ec
h
a
n
i
c
s
i
n
M
ed
i
ci
n
e
a
n
d
B
i
o
l
o
g
y,
v
o
l
.
1
2
,
n
o
.
1
,
2
0
1
2
.
[2
1
]
M.
K
.
D
as
,
D
.
K
.
G
h
o
s
h
,
an
d
S
.
A
ri
,
"
E
l
ect
r
o
card
i
o
g
r
am
(E
CG
)
s
i
g
n
a
l
cl
as
s
i
f
i
cat
i
o
n
u
s
i
n
g
s
-
t
ran
s
fo
rm,
g
en
et
i
c
al
g
o
ri
t
h
m,
an
d
n
e
u
ral
n
et
w
o
r
k
.
"
1
s
t
I
n
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
er
en
ce
o
n
Co
n
d
i
t
i
o
n
A
s
s
e
s
s
m
en
t
Tech
n
i
q
u
e
s
i
n
E
l
ect
r
i
ca
l
S
ys
t
em
s
,
p
p
.
3
5
3
-
3
5
7
.
2
0
1
3
.
[2
2
]
L
.
N
.
Sh
arma,
R
.
K
.
T
ri
p
at
h
y
,
an
d
S
.
D
an
d
a
p
at
,
"
Mu
l
t
i
s
c
al
e
en
er
g
y
an
d
ei
g
en
s
p
ace
ap
p
ro
ac
h
t
o
d
et
ec
t
i
o
n
an
d
l
o
cal
i
zat
i
o
n
o
f
m
y
o
car
d
i
a
l
i
n
farct
i
o
n
,
"
IE
E
E
T
r
a
n
s
a
c
t
i
o
n
s
o
n
B
i
o
m
e
d
i
c
a
l
E
n
g
i
n
eer
i
n
g
,
v
o
l
.
6
2
,
n
o
.
7
,
pp.
1
8
2
7
-
1
8
3
7
,
2
0
1
5
.
[2
3
]
K
.
S.
K
u
mar,
B
.
Y
az
d
an
p
an
a
h
,
an
d
P.
R
.
K
u
mar,
"
Re
mo
v
a
l
o
f
n
o
i
s
e
fr
o
m
el
ec
t
ro
car
d
i
o
g
ram
u
s
i
n
g
d
i
g
i
t
a
l
F
IR
an
d
IIR
fi
l
t
er
s
w
i
t
h
v
ar
i
o
u
s
me
t
h
o
d
s
,
"
In
t
er
n
a
t
i
o
n
a
l
Co
n
f
e
r
en
ce
o
n
C
o
m
m
u
n
i
c
a
t
i
o
n
s
a
n
d
S
i
g
n
a
l
P
r
o
ces
s
i
n
g
,
p
p
.
0
1
5
7
-
0
1
6
2
,
2
0
1
5
.
[2
4
]
M.
T
u
rh
a
n
,
D
.
Şe
n
g
u
r,
S
.
K
ara
b
at
a
k
,
Y
.
G
u
o
,
a
n
d
.
Smaran
d
ac
h
e,
"
N
eu
t
ro
s
o
p
h
i
c
w
e
i
g
h
t
e
d
s
u
p
p
o
rt
v
ect
o
r
mac
h
i
n
e
s
fo
r
t
h
e
d
e
t
ermi
n
at
i
o
n
o
f
s
c
h
o
o
l
ad
m
i
n
i
s
t
rat
o
rs
w
h
o
at
t
en
d
ed
an
act
i
o
n
l
earn
i
n
g
co
u
rs
e
b
a
s
ed
o
n
t
h
e
i
r
co
n
fl
i
ct
-
h
an
d
l
i
n
g
s
t
y
l
e
s
,
"
S
ym
m
e
t
r
y
,
v
o
l
.
10
,
n
o
.
5
,
p
p
.
1
-
11
,
2
0
1
8
.
[2
5
]
M.
H
o
fman
n
,
"
Su
p
p
o
rt
v
ec
t
o
r
mac
h
i
n
es
-
K
ern
e
l
s
a
n
d
t
h
e
k
ern
el
t
ri
c
k
,
"
No
t
es
,
p
p
.
1
-
1
6
,
2
0
0
6
.
[2
6
]
R.
H
e,
K
.
W
a
n
g
,
Q
.
L
i
,
Y
.
Y
u
an
,
N
.
Z
h
ao
,
Y
.
L
i
u
,
an
d
H
.
Z
h
an
g
,
“
A
n
o
v
e
l
me
t
h
o
d
f
o
r
t
h
e
d
et
ec
t
i
o
n
o
f
R
-
p
ea
k
s
i
n
E
CG
b
as
e
d
o
n
K
-
n
eare
s
t
n
e
i
g
h
b
o
rs
a
n
d
p
ar
t
i
c
l
e
s
w
ar
m
o
p
t
i
m
i
zat
i
o
n
,
”
E
U
R
A
S
I
P
J
o
u
r
n
a
l
o
n
A
d
va
n
ces
i
n
S
i
g
n
a
l
P
r
o
ces
s
i
n
g
,
v
o
l
.
2
0
1
7
,
n
o
.
8
2
,
p
p
.
1
-
1
4
,
2
0
1
7
.
[2
7
]
I.
Sai
n
i
,
D
.
Si
n
g
h
,
an
d
A
.
K
h
o
s
l
a,
"
Q
RS
d
et
ect
i
o
n
u
s
i
n
g
K
-
N
eare
s
t
N
e
i
g
h
b
o
r
al
g
o
r
i
t
h
m
(K
N
N
)
an
d
e
v
al
u
at
i
o
n
o
n
s
t
a
n
d
ar
d
E
CG
d
at
a
b
as
e
s
,
"
Jo
u
r
n
a
l
o
f
a
d
v
a
n
ce
d
r
e
s
ea
r
ch
.
,
v
o
l
.
4
,
n
o
.
4
,
p
p
.
3
3
1
-
3
4
4
,
2
0
1
3
.
[2
8
]
M
.
Su
v
arn
a
an
d
V
en
k
at
e
g
o
w
d
a
N
.
,
“Perfo
rman
ce
meas
u
re
an
d
effi
ci
e
n
cy
o
f
ch
emi
cal
s
k
i
n
b
u
rn
c
l
as
s
i
f
i
cat
i
o
n
u
s
i
n
g
K
N
N
m
et
h
o
d
,
”
P
r
o
ce
d
i
a
Co
m
p
u
t
e
r
S
ci
e
n
ce
,
p
p
.
4
8
-
5
4
,
2
0
1
5
.
[2
9
]
N
.
Mo
h
an
k
u
mar,
B.
Bh
u
v
an
,
N
.
D
ev
i
,
an
d
S.
A
ru
mu
g
am
,
"
A
mo
d
i
fi
ed
g
e
n
et
i
c
al
g
o
ri
t
h
m
fo
r
ev
o
l
u
t
i
o
n
o
f
n
e
u
ra
l
n
et
w
o
r
k
i
n
d
e
s
i
g
n
i
n
g
an
ev
o
l
u
t
i
o
n
ar
y
n
eu
r
o
-
h
ard
w
ar,
"
In
t
er
n
a
t
i
o
n
a
l
Co
n
f
e
r
en
ce
o
n
G
en
e
t
i
c
a
n
d
E
vo
l
u
t
i
o
n
a
r
y
M
et
h
o
d
s
,
p
p
.
1
0
8
-
1
1
1
,
2
0
0
8
.
[3
0
]
M
.
H
.
Imam
,
C
.
K
armak
ar,
H
.
F.
J
el
i
n
ek
,
M
.
Pal
a
n
i
s
w
ami
,
an
d
A
.
H
.
K
h
a
n
d
o
k
er
,
"
D
et
ect
i
n
g
s
u
b
c
l
i
n
i
ca
l
d
i
ab
et
i
c
card
i
ac
a
u
t
o
n
o
mi
c
n
eu
r
o
p
a
t
h
y
b
y
an
a
l
y
z
i
n
g
v
e
n
t
r
i
cu
l
ar
rep
o
l
ar
i
zat
i
o
n
d
y
n
am
i
cs
,
"
J.
B
i
o
m
ed
i
ca
l
a
n
d
H
e
a
l
t
h
In
f
o
r
m
a
t
i
c
s
,
v
o
l
.
2
0
,
n
o
.
1
,
p
p
.
6
4
-
7
2
,
2
0
1
6
.
[3
1
]
P
.
C
.
Ch
an
g
,
J
.
J
.
L
i
n
,
J
.
C
.
H
s
i
eh
,
an
d
J
.
W
en
g
,
"
My
o
car
d
i
a
l
i
n
farct
i
o
n
cl
a
s
s
i
fi
ca
t
i
o
n
w
i
t
h
m
u
l
t
i
-
l
ead
E
CG
u
s
i
n
g
h
i
d
d
e
n
Mark
o
v
mo
d
el
s
an
d
G
au
s
s
i
a
n
mi
x
t
u
re
mo
d
e
l
s
,
"
A
p
p
l
i
ed
S
o
f
t
Co
m
p
u
t
i
n
g
,
v
o
l
.
1
2
,
n
o
.
1
0
,
p
p.
3
1
6
5
-
3
1
7
5
,
2
0
1
2
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
T
r
aini
ng
fee
dfor
w
ar
d
ne
ur
al
ne
tw
or
k
us
ing
ge
ne
ti
c
al
g
o
r
it
h
m…
(
J
.
R
e
v
athi
)
1291
B
I
OG
RA
P
H
I
E
S
OF
AU
T
HO
RS
M
rs
.
R
ev
a
th
i
J
o
th
i
ra
m
a
l
i
n
g
a
m
i
s
an
A
s
s
i
s
t
an
t
Pro
fe
s
s
o
r
i
n
t
h
e
D
e
p
art
me
n
t
o
f
Bi
o
me
d
i
cal
E
n
g
i
n
eeri
n
g
at
K
I
T
-
K
a
l
ai
g
n
ar
k
aru
n
an
i
d
t
h
i
In
s
t
i
t
u
t
e
o
f
T
ech
n
o
l
o
g
y
.
Sh
e
h
as
mo
re
t
w
e
l
v
e
y
ears
o
f
p
ro
fe
s
s
i
o
n
al
ex
p
eri
e
n
ce
i
n
Bi
o
med
i
cal
fi
e
l
d
.
H
er
res
earch
t
o
p
i
c
s
i
n
c
l
u
d
e
Bi
o
me
d
i
ca
l
In
s
t
ru
me
n
t
a
t
i
o
n
,
B
i
o
med
i
cal
Ima
g
e
an
d
Si
g
n
a
l
Pro
c
es
s
i
n
g
a
n
d
N
an
o
t
ec
h
n
o
l
o
g
y
.
Sh
e
h
as
p
u
b
l
i
s
h
ed
mo
re
t
h
a
n
1
0
re
s
earch
art
i
c
l
es
i
n
j
o
u
r
n
al
s
an
d
co
n
fere
n
ce
p
ro
ce
ed
i
n
g
s
.
D
r.
A
ni
tha
J
ude
Hem
a
nth
i
s
an
A
s
s
o
c
i
at
e
Pro
fe
s
s
o
r
i
n
t
h
e
D
ep
ar
t
men
t
o
f
E
l
ect
r
o
n
i
cs
a
n
d
Co
mmu
n
i
ca
t
i
o
n
E
n
g
i
n
eer
i
n
g
at
K
a
ru
n
y
a
In
s
t
i
t
u
t
e
o
f
T
e
ch
n
o
l
o
g
y
an
d
Sc
i
en
ce
s
.
Sh
e
co
m
p
l
e
t
ed
h
er
Ph
.
D
at
t
h
e
K
ar
u
n
y
a
I
n
s
t
i
t
u
t
e
o
f
T
ec
h
n
o
l
o
g
y
,
Co
i
mb
at
o
re.
H
er
re
s
earch
i
n
t
ere
s
t
i
n
c
l
u
d
es
Mach
i
n
e
l
earn
i
n
g
,
Cl
as
s
i
f
i
cat
i
o
n
,
N
eu
ra
l
N
et
w
o
r
k
s
a
n
d
A
rt
i
fi
c
i
al
In
t
el
l
i
g
en
ce,
Imag
e
A
n
a
l
y
s
i
s
,
Med
i
ca
l
Imag
e
a
n
d
s
i
g
n
al
Pr
o
ces
s
i
n
g
.
S
h
e
h
a
s
p
re
s
en
t
e
d
p
ap
e
rs
at
c
o
n
fer
en
ce
s
b
o
t
h
h
o
me
a
n
d
ab
ro
a
d
,
p
u
b
l
i
s
h
ed
ar
t
i
c
l
es
a
n
d
p
ap
er
s
i
n
v
ar
i
o
u
s
j
o
u
r
n
al
s
,
an
d
c
o
n
t
ri
b
u
t
ed
a
c
h
ap
t
er
t
o
t
h
e
b
o
o
k
s
.
Sh
e
p
u
b
l
i
s
h
ed
mo
re
t
h
a
n
9
2
res
earch
p
ap
er
s
i
n
t
h
es
e
areas
.
D
r.
D
.
J
ude
Hem
a
nth
recei
v
e
d
h
i
s
B.
E
d
e
g
ree
i
n
E
C
E
fro
m
B
h
ara
t
h
i
ar
U
n
i
v
ers
i
t
y
i
n
2
0
0
2
,
M.
E
d
eg
ree
i
n
co
mmu
n
i
ca
t
i
o
n
s
y
s
t
ems
fro
m
A
n
n
a
U
n
i
v
ers
i
t
y
i
n
2
0
0
6
an
d
Ph
.
D
.
fro
m
K
aru
n
y
a
U
n
i
v
er
s
i
t
y
i
n
2
0
1
3
.
H
i
s
res
earch
areas
i
n
cl
u
d
e
Co
mp
u
t
a
t
i
o
n
a
l
In
t
el
l
i
g
en
ce
an
d
Imag
e
p
ro
ce
s
s
i
n
g
.
H
e
h
as
au
t
h
o
re
d
mo
re
t
h
an
1
0
0
res
ear
ch
p
ap
er
s
i
n
re
p
u
t
e
d
SCIE
i
n
d
ex
e
d
In
t
er
n
at
i
o
n
al
J
o
u
r
n
al
s
an
d
In
t
ern
a
t
i
o
n
a
l
Co
n
fere
n
ces
w
i
t
h
l
ea
d
i
n
g
p
u
b
l
i
s
h
ers
s
u
ch
a
s
E
l
s
ev
i
er,
Sp
ri
n
g
er,
IE
E
E
,
et
c.
H
i
s
Cu
mu
l
a
t
i
v
e
Imp
act
Fact
o
r
i
s
mo
re
t
h
an
1
0
0
.
H
e
h
as
au
t
h
o
re
d
1
b
o
o
k
w
i
t
h
(V
D
M
-
V
er
l
a
g
,
G
erman
y
)
an
d
2
1
ed
i
t
e
d
b
o
o
k
s
w
i
t
h
rep
u
t
ed
p
u
b
l
i
s
h
ers
s
u
c
h
as
E
l
s
ev
i
er,
Sp
ri
n
g
er,
IE
T
an
d
I
O
S
Pres
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.