I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
9
,
No
.
3
,
J
u
n
e
201
9
,
p
p
.
1832
~
1
8
4
1
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v9
i
3
.
p
p
1
8
3
2
-
1841
1832
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e
.
co
m/
jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
JE
C
E
Auto
m
a
ted
seg
m
e
ntatio
n and
clas
si
ficatio
n t
echniqu
e
for bra
in st
ro
ke
N.
S.
M
.
No
o
r
1
,
N.
M
.
Sa
a
d
2
,
A.
R.
Abdu
lla
h
3
,
N.
M
.
Ali
4
1,
2
F
a
k
u
lt
i
Ke
ju
r
u
tera
a
n
El
e
k
tro
n
ik
d
a
n
Ke
ju
r
u
tera
a
n
Ko
m
p
u
ter,
U
n
i
v
e
rsiti
T
e
k
n
ik
a
l
M
a
la
y
si
a
M
e
lak
a
,
M
a
lay
sia
3,
4
F
a
k
u
lt
i
Ke
ju
r
u
ter
a
a
n
El
e
k
tri
k
,
Un
iv
e
rsiti
T
e
k
n
ik
a
l
M
a
la
y
sia
M
e
l
a
k
a
,
M
a
la
y
si
a
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Sep
2
,
2
0
1
8
R
ev
i
s
ed
Dec
1
6
,
2
0
1
8
A
cc
ep
ted
J
an
2
6
,
2
0
1
9
Dif
u
ss
io
n
-
W
e
ig
h
ted
Im
a
g
in
g
(DW
I)
p
lay
s
a
n
im
p
o
rtan
t
r
o
le
i
n
t
h
e
d
iag
n
o
sis
o
f
b
ra
in
stro
k
e
b
y
p
ro
v
id
in
g
d
e
tailed
in
f
o
rm
a
ti
o
n
re
g
a
rd
in
g
th
e
so
f
t
ti
ss
u
e
c
o
n
tras
t
in
th
e
b
ra
in
o
rg
a
n
.
Co
n
v
e
n
ti
o
n
a
ll
y
,
th
e
d
i
ff
e
re
n
ti
a
l
d
iag
n
o
sis o
f
b
ra
in
stro
k
e
les
io
n
s
is
p
e
rf
o
rm
e
d
m
a
n
u
a
ll
y
b
y
p
ro
f
e
ss
io
n
a
l
n
e
u
ro
r
a
d
io
l
o
g
ists
d
u
ri
n
g
a
h
ig
h
l
y
su
b
jec
ti
v
e
a
n
d
ti
m
e
-
c
o
n
s
u
m
in
g
p
ro
c
e
ss
.
T
h
is
stu
d
y
p
ro
p
o
se
s
a
se
g
m
e
n
tatio
n
a
n
d
c
las
sif
ica
ti
o
n
tec
h
n
iq
u
e
t
o
d
e
tec
t
b
ra
in
stro
k
e
les
io
n
s
b
a
se
d
o
n
d
if
f
u
sio
n
-
w
e
ig
h
ted
i
m
a
g
in
g
(D
W
I).
T
h
e
t
y
p
e
o
f
stro
k
e
les
io
n
s
c
o
n
sists
o
f
a
c
u
te
isc
h
e
m
ic,
su
b
-
a
c
u
te
isc
h
e
m
ic,
c
h
ro
n
ic
isc
h
e
m
ic
a
n
d
a
c
u
te
h
e
m
o
rrh
a
g
e
.
F
o
r
se
g
m
e
n
tatio
n
,
f
u
z
z
y
c
-
M
e
a
n
s
(F
CM
)
a
n
d
a
c
ti
v
e
c
o
n
t
o
u
r
is
p
ro
p
o
se
d
t
o
se
g
m
e
n
t
th
e
les
io
n
’
s
re
g
io
n
.
F
CM
is
im
p
lem
e
n
ted
w
it
h
a
c
ti
v
e
c
o
n
to
u
r
to
se
p
a
ra
te
th
e
c
e
re
b
ra
l
sp
in
a
l
f
lu
id
(CS
F
)
w
it
h
th
e
h
y
p
o
in
ten
se
les
io
n
.
P
re
-
p
ro
c
e
ss
in
g
is
a
p
p
li
e
d
to
th
e
DW
I
f
o
r
im
a
g
e
n
o
r
m
a
li
z
a
ti
o
n
,
b
a
c
k
g
ro
u
n
d
re
m
o
v
a
l
a
n
d
i
m
a
g
e
e
n
h
a
n
c
e
m
e
n
t.
T
h
e
a
lg
o
rit
h
m
p
e
rf
o
r
m
a
n
c
e
h
a
s b
e
e
n
e
v
a
lu
a
ted
u
sin
g
Ja
c
c
a
rd
In
d
e
x
,
Dic
e
Co
e
ff
icie
n
t
(DC) an
d
b
o
t
h
f
a
lse
p
o
siti
v
e
ra
te
(F
P
R)
a
n
d
f
a
lse
n
e
g
a
ti
v
e
ra
te
(F
NR).
T
h
e
a
v
e
r
a
g
e
re
s
u
lt
s
f
o
r
th
e
Ja
c
c
a
rd
in
d
e
x
,
DC,
F
P
R
a
n
d
F
N
R
a
re
0
.
5
5
,
0
.
6
8
,
0
.
2
3
a
n
d
0
.
2
3
,
r
e
sp
e
c
ti
v
e
l
y
.
F
irst
sta
ti
stica
l
o
r
d
e
r
m
e
th
o
d
is
a
p
p
li
e
d
to
th
e
se
g
m
e
n
tatio
n
re
su
lt
to
o
b
tai
n
th
e
f
e
a
tu
re
s
f
o
r
th
e
c
las
si
f
ier
in
p
u
t.
F
o
r
c
las
sif
ic
a
ti
o
n
tec
h
n
i
q
u
e
,
b
a
g
g
e
d
tree
c
las
si
f
ier
is
p
ro
p
o
se
d
t
o
c
las
sify
th
e
ty
p
e
o
f
stro
k
e
.
T
h
e
a
c
c
u
ra
c
y
re
su
lt
s
f
o
r
th
e
c
las
sif
ica
ti
o
n
is
9
0
.
8
%
.
Ba
se
d
o
n
th
e
re
su
l
ts,
t
h
e
p
r
o
p
o
se
d
tec
h
n
i
q
u
e
h
a
s
p
o
ten
ti
a
l
to
se
g
m
e
n
t
a
n
d
c
las
sify
b
ra
in
str
o
k
e
les
io
n
f
ro
m
DW
I
i
m
a
g
e
s.
K
ey
w
o
r
d
s
:
B
r
ain
i
m
a
g
in
g
C
las
s
i
f
icatio
n
Dif
f
u
s
io
n
-
w
e
ig
h
ted
i
m
a
g
i
n
g
(
DW
I
)
Seg
m
en
tatio
n
Stro
k
e
Co
p
y
rig
h
t
©
2
0
1
9
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
N.
M.
Saad
,
Fak
u
lti Ke
j
u
r
u
ter
aa
n
E
le
k
tr
o
n
i
k
d
an
Kej
u
r
u
ter
aa
n
Ko
m
p
u
ter
,
Un
i
v
er
s
iti T
ek
n
ik
al
Ma
la
y
s
i
a
Me
lak
a,
Me
lak
a,
Ma
la
y
s
ia.
E
m
ail:
n
o
r
h
as
h
i
m
a
h
@
u
te
m
.
ed
u
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
No
w
ad
a
y
s
,
m
ed
ical
i
m
a
g
i
n
g
to
o
l
p
lay
s
a
n
i
m
p
o
r
tan
t
r
o
le
i
n
v
ie
w
i
n
g
t
h
e
i
n
ter
n
a
l
ti
s
s
u
e
o
f
h
u
m
a
n
b
r
ain
.
I
n
b
r
ain
s
tr
o
k
e
d
iag
n
o
s
is
,
m
ed
ical
i
m
a
g
i
n
g
s
u
c
h
as
MRI
o
f
f
er
s
f
ast
i
m
a
g
i
n
g
to
o
l
w
it
h
h
ig
h
r
eso
lu
t
io
n
an
d
n
o
n
io
n
izin
g
r
ad
iatio
n
[
1
]
.
T
h
e
MRI
im
a
g
i
n
g
to
o
l
is
ab
le
to
d
etec
t
8
5
%
s
tr
o
k
e
s
u
r
v
iv
o
r
f
r
o
m
is
c
h
e
m
ic
s
tr
o
k
e
an
d
1
5
-
2
0
%
s
tr
o
k
e
s
u
r
v
iv
o
r
f
r
o
m
h
e
m
o
r
r
h
ag
ic
s
tr
o
k
e
[
2
]
.
D
W
I
is
o
n
e
o
f
th
e
MRI
m
o
d
ali
t
y
t
h
at
s
h
o
ws
h
ig
h
s
e
n
s
iti
v
it
y
(
8
8
-
100%
)
an
d
s
p
ec
if
icit
y
(
9
5
-
1
0
0
%)
in
d
et
ec
tin
g
ea
r
l
y
is
c
h
e
m
ic
s
tr
o
k
e
[
3
]
.
T
o
f
u
ll
y
u
t
ilize
m
ed
ical
i
m
a
g
in
g
,
n
e
w
al
g
o
r
ith
m
s
an
d
m
et
h
o
d
s
ar
e
u
r
g
ed
i
n
t
h
e
m
ed
ica
l
i
m
a
g
i
n
g
f
ield
f
o
r
u
tili
z
in
g
v
er
y
lar
g
e
a
m
o
u
n
ts
o
f
i
m
a
g
es
w
i
s
el
y
[
4
]
.
Ho
w
e
v
er
,
d
u
e
to
t
h
e
lar
g
e
s
a
m
p
le
o
f
i
m
a
g
es,
t
h
e
d
iag
n
o
s
i
s
p
r
o
ce
s
s
is
ti
m
e
c
o
n
s
u
m
i
n
g
a
n
d
tire
d
n
e
s
s
[
5
]
.
A
co
m
p
u
ter
aid
ed
d
ia
g
n
o
s
is
is
n
ee
d
ed
f
o
r
n
eu
r
o
r
ad
io
lo
g
is
t
to
in
ter
p
r
et
th
e
i
m
a
g
es
m
o
r
e
ea
s
il
y
.
M
an
y
r
esear
ch
er
s
h
a
v
e
s
tu
d
ie
s
m
ac
h
i
n
e
lear
n
i
n
g
tech
n
iq
u
e
i
n
b
r
ain
i
m
ag
i
n
g
[
6
]
.
Ma
ch
in
e
lear
n
i
n
g
tec
h
n
iq
u
e
is
b
elie
v
ed
to
p
er
f
o
r
m
ta
s
k
s
an
d
s
o
lv
e
p
r
o
b
lem
s
r
elate
d
to
p
o
o
r
q
u
alit
y
b
r
ain
i
m
a
g
in
g
s
a
m
p
le
s
b
y
p
r
o
v
id
i
n
g
ac
c
u
r
ate
r
ep
r
esen
tat
io
n
a
n
d
p
r
io
r
k
n
o
w
led
g
e
m
o
d
eli
n
g
.
T
h
u
s
,
it
p
la
y
s
an
i
m
p
o
r
ta
n
t
r
o
le
in
b
r
ain
i
m
a
g
i
n
g
w
h
er
e
it
h
a
s
b
ec
o
m
e
o
n
e
o
f
th
e
m
o
s
t
p
r
o
m
i
s
in
g
r
esear
ch
ar
ea
s
in
co
m
p
u
ter
aid
ed
d
etec
tio
n
an
d
d
iag
n
o
s
i
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
A
u
to
ma
ted
s
eg
men
ta
tio
n
a
n
d
cla
s
s
ifica
tio
n
tech
n
iq
u
e
fo
r
b
r
a
in
s
tr
o
ke
(
N
.
M.
S
a
a
d
)
1833
Z
h
an
g
Y.
et
a
l
.
[
7
]
p
r
o
p
o
s
ed
k
er
n
el
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
to
d
etec
t
A
lzh
e
i
m
er
’
s
d
is
ea
s
e
(
A
D)
an
d
class
i
f
y
th
e
d
is
ea
s
e
b
et
w
ee
n
n
o
r
m
al
eld
er
co
n
tr
o
l
(
NC
)
.
R
a
m
a
n
i
R
.
G.
et
a
l
.
[
8
]
p
r
o
p
o
s
ed
Naiv
e
B
a
y
es,
Su
p
p
o
r
t
Vec
to
r
Ma
ch
in
e
an
d
r
an
d
o
m
tr
ee
to
an
al
y
ze
t
h
e
m
ac
h
in
e
lear
n
i
n
g
tec
h
n
iq
u
e
i
n
class
i
f
y
in
g
n
o
r
m
al
an
d
ab
n
o
r
m
al
b
r
ain
i
m
a
g
e.
M
u
d
ali
D.
et
a
l
.
[
9
]
p
r
o
p
o
s
ed
d
ec
is
io
n
tr
ee
m
eth
o
d
to
in
ter
p
r
et
th
e
d
iag
n
o
s
i
s
o
f
n
eu
r
o
d
eg
e
n
er
ativ
e
b
r
ain
d
is
ea
s
es.
Me
et
Oza
et
a
l
.
[
1
0
]
p
r
o
p
o
s
ed
r
an
d
o
m
f
o
r
est
cla
s
s
i
f
ie
r
to
class
i
f
y
b
r
ain
tu
m
o
r
i
n
to
b
e
n
ig
n
o
r
m
ali
g
n
ca
s
es.
Fra
tello
M.
et
a
l
.
[
1
1
]
p
r
o
p
o
s
ed
r
an
d
o
m
f
o
r
est
to
cl
ass
i
f
y
t
w
o
t
y
p
e
o
f
d
is
ea
s
e
f
r
o
m
i
n
d
iv
id
u
al
later
al
s
cler
o
s
is
(
AL
S)
p
atien
ts
,
P
ar
k
in
s
o
n
’
s
d
i
s
ea
s
e
(
P
D)
p
a
tien
t
s
an
d
h
ea
lt
h
y
co
n
tr
o
l (
HC
)
s
u
b
j
ec
ts
.
I
n
th
i
s
p
ap
er
,
s
eg
m
en
tatio
n
an
d
class
if
ica
tio
n
tec
h
n
iq
u
e
f
o
r
b
r
ain
s
tr
o
k
e
u
s
i
n
g
DW
I
i
m
a
g
e
i
s
p
r
o
p
o
s
ed
.
T
h
e
p
u
r
p
o
s
e
o
f
th
is
s
tu
d
y
is
to
d
ev
elo
p
an
a
u
to
m
atic
s
tr
o
k
e
s
eg
m
e
n
tatio
n
an
d
class
i
f
icatio
n
u
s
i
n
g
DW
I
im
a
g
es.
T
h
e
p
r
o
p
o
s
ed
s
eg
m
e
n
tat
io
n
an
al
y
s
is
is
b
ased
o
n
th
e
f
u
zz
y
c
-
Me
a
n
s
(
FC
M)
an
d
ac
tiv
e
co
n
to
u
r
tech
n
iq
u
e.
A
c
tiv
e
co
n
to
u
r
is
in
teg
r
ated
in
t
h
e
an
al
y
s
i
s
f
r
a
m
e
w
o
r
k
to
s
ep
ar
ate
th
e
ce
r
eb
r
al
s
p
in
al
f
l
u
id
(
C
SF
)
w
it
h
t
h
e
h
y
p
o
in
te
n
s
e
le
s
io
n
a
n
d
in
cr
ea
s
e
t
h
e
s
e
g
m
e
n
tat
io
n
ac
cu
r
ac
y
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
s
eg
m
e
n
tatio
n
tech
n
iq
u
es
is
e
v
al
u
ated
b
ased
o
n
J
ac
ca
r
d
in
d
ex
,
Dice
C
o
ef
f
icien
t
(
D
C
)
,
f
al
s
e
p
o
s
iti
v
e
r
at
e
(
FP
R
)
an
d
f
al
s
e
n
eg
at
iv
e
r
ate
(
FNR
)
.
Af
ter
t
h
e
s
eg
m
e
n
tatio
n
r
esu
lt
i
s
o
b
tain
,
th
e
r
esu
lt
i
m
a
g
e
is
e
x
tr
ac
t
e
d
u
s
i
n
g
f
ir
s
t
o
r
d
er
s
tatis
t
ical
m
et
h
o
d
.
T
h
ese
f
ea
tu
r
es
ar
e
u
s
ed
f
o
r
th
e
b
ag
g
e
d
tr
ee
class
if
ier
in
p
u
t.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
class
i
f
ier
tech
n
iq
u
e
is
e
v
alu
ate
d
b
ased
o
n
th
e
ac
cu
r
ac
y
(
A
C
C
)
.
T
h
is
p
ap
er
c
o
n
tain
s
f
i
v
e
s
ec
ti
o
n
.
Sectio
n
2
d
is
cu
s
s
ed
th
e
f
l
o
w
p
r
o
c
ess
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
o
lo
g
y
in
d
etail.
Sectio
n
3
p
r
o
p
o
s
ed
th
e
b
r
ain
s
tr
o
k
e
a
n
al
y
s
i
s
tec
h
n
i
q
u
e
u
s
ed
to
s
e
g
m
en
t
an
d
cla
s
s
if
y
t
h
e
b
r
ain
s
tr
o
k
e
lesi
o
n
.
Sec
tio
n
4
d
is
c
u
s
s
e
d
th
e
r
es
u
lt
f
r
o
m
t
h
e
s
e
g
m
e
n
tat
io
n
r
es
u
lt
a
n
d
S
ec
tio
n
5
co
n
cl
u
d
e
th
e
w
o
r
k
p
r
esen
ted
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
2
.
1
.
P
r
o
po
s
ed
a
na
ly
s
is
f
ra
m
ew
o
rk
T
h
e
f
lo
w
p
r
o
ce
s
s
o
f
t
h
is
s
tu
d
y
s
tar
ts
w
i
th
i
m
a
g
e
p
r
ep
r
o
ce
s
s
in
g
w
h
er
e
i
m
a
g
e
n
o
r
m
a
lizati
o
n
,
i
m
a
g
e
en
h
a
n
ce
m
en
t
an
d
b
ac
k
g
r
o
u
n
d
r
e
m
o
v
al
ar
e
ap
p
lied
.
Af
ter
t
h
at,
FC
M
is
ap
p
lied
f
o
r
i
m
a
g
e
s
eg
m
e
n
tatio
n
s
ta
g
e
in
o
r
d
er
to
ex
tr
ac
t
th
e
R
OI
o
f
th
e
s
tr
o
k
e
lesi
o
n
.
T
h
en
,
th
e
ac
tiv
e
co
n
to
u
r
is
u
s
ed
to
s
ep
ar
ate
th
e
C
SF
w
it
h
t
h
e
h
y
p
o
i
n
ten
s
e
s
tr
o
k
e
lesi
o
n
r
eg
io
n
.
H
y
p
er
i
n
ten
s
e
is
r
ef
er
r
ed
to
b
r
ig
h
t
le
s
io
n
s
u
c
h
a
s
ac
u
te
is
ch
e
m
ic
s
tr
o
k
e
,
ac
u
te
h
e
m
o
r
r
h
a
g
e
s
tr
o
k
e
a
n
d
s
u
b
-
ac
u
te
is
c
h
e
m
ic
s
tr
o
k
e
lesi
o
n
w
h
er
e
h
y
p
o
i
n
te
n
s
e
i
s
r
ef
er
r
ed
to
d
ar
k
lesi
o
n
s
u
c
h
as
c
h
r
o
n
ic
is
c
h
e
m
ic
s
tr
o
k
e
lesi
o
n
.
T
h
e
s
e
g
m
e
n
tat
io
n
r
esu
lt
is
e
x
tr
ac
ted
u
s
in
g
f
i
r
s
t
o
r
d
er
s
tatis
tical
m
et
h
o
d
f
o
r
th
e
in
p
u
t
clas
s
i
f
ie
r
.
A
f
ter
th
e
f
ea
tu
r
es
ar
e
o
b
tai
n
ed
,
b
ag
g
ed
tr
ee
class
i
f
ier
is
u
s
ed
to
class
i
f
y
th
e
t
y
p
e
o
f
s
tr
o
k
e
s
lesi
o
n
.
L
as
t
b
u
t
n
o
t
least,
t
h
e
p
er
f
o
r
m
a
n
c
e
o
f
th
e
b
r
ain
i
m
ag
e
i
s
ev
al
u
ated
b
ased
o
n
th
e
J
ac
ca
r
d
in
d
ex
,
DC
,
FP
R
,
FNR
,
A
CC
.
2
.
2
.
I
m
a
g
ing
pa
ra
m
et
er
T
h
is
p
ap
er
f
o
cu
s
ed
o
n
t
w
o
b
r
ain
s
tr
o
k
e
d
atase
ts
f
r
o
m
DW
I
i
m
a
g
es
.
Da
taset
1
co
n
tai
n
s
6
1
s
a
m
p
le
s
f
r
o
m
t
h
e
Gen
er
al
Ho
s
p
ital
o
f
Ku
ala
L
u
m
p
u
r
.
Data
s
et
2
co
n
tai
n
s
6
9
s
a
m
p
le
s
f
r
o
m
th
e
o
n
li
n
e
d
atab
ase
o
f
I
s
ch
e
m
ic
Stro
k
e
L
esio
n
Seg
m
en
tatio
n
(
I
S
L
E
S).
T
h
e
d
ata
s
et
o
n
l
y
f
o
cu
s
ed
o
n
f
o
u
r
t
y
p
es
o
f
b
r
ain
s
tr
o
k
e
lesi
o
n
w
h
ich
ar
e
ac
u
te
is
c
h
e
m
ic
s
tr
o
k
e,
ch
r
o
n
ic
i
s
ch
e
m
ic
s
tr
o
k
e,
ac
u
te
h
e
m
o
r
r
h
a
g
e
s
tr
o
k
e
an
d
s
u
b
-
ac
u
te
is
ch
e
m
ic
s
tr
o
k
e.
T
h
e
ac
u
te
is
ch
e
m
ic
s
tr
o
k
e,
ch
r
o
n
ic
i
s
ch
e
m
ic
s
tr
o
k
e
a
n
d
ac
u
te
h
e
m
o
r
r
h
ag
e
s
tr
o
k
e
i
m
ag
e
s
w
er
e
g
ain
ed
f
r
o
m
t
h
e
Ge
n
er
a
l
Ho
s
p
ital
o
f
K
u
ala
L
u
m
p
u
r
(
HKL
)
.
T
h
e
s
u
b
-
ac
u
t
e
is
c
h
e
m
ic
s
tr
o
k
e
i
m
ag
e
s
w
er
e
g
a
in
ed
f
r
o
m
th
e
p
u
b
lic
o
n
li
n
e
I
s
ch
e
m
ic
S
tr
o
k
e
L
e
s
io
n
Seg
m
en
tatio
n
(
I
S
L
E
S)
ch
a
llen
g
e
2
0
1
5
[
1
2
]
.
T
ab
le
1
.
T
h
e
B
r
ain
Stro
k
e
Data
s
ets
S
t
r
o
k
e
D
a
t
a
se
t
s
D
a
t
a
se
t
1
(
c
l
i
n
i
c
a
l
)
D
a
t
a
se
t
2
(
I
S
L
ES)
[
1
2
]
A
c
u
t
e
I
sch
e
mi
c
C
h
r
o
n
i
c
I
sch
e
mi
c
A
c
u
t
e
H
e
mo
r
r
h
a
g
e
S
u
b
-
a
c
u
t
e
I
sch
e
mi
c
N
u
mb
e
r
o
f
S
a
m
p
l
e
s
20
33
8
69
2
.
3
.
I
m
a
g
e
pre
-
pro
ce
s
s
ing
T
h
e
p
r
e
-
p
r
o
ce
s
s
in
g
s
ta
g
e
i
s
t
h
e
s
ta
g
e
w
h
er
e
t
h
e
i
m
ag
e
s
n
ee
d
to
u
n
d
er
g
o
p
r
e
-
p
r
o
ce
s
s
in
g
s
ta
g
es
to
ac
q
u
ir
e
b
etter
s
eg
m
e
n
tatio
n
[
1
3
]
-
[
1
5
]
.
T
h
r
ee
alg
o
r
ith
m
s
w
er
e
ap
p
lied
to
th
e
DW
I
im
a
g
e
w
h
ic
h
is
i
m
ag
e
n
o
r
m
aliza
t
io
n
,
b
ac
k
g
r
o
u
n
d
r
e
m
o
v
al
a
n
d
i
m
ag
e
e
n
h
an
ce
m
en
t
[
1
6
]
.
T
h
ese
i
m
a
g
es
ar
e
co
n
v
er
ted
in
to
t
h
e
d
esire
d
f
o
r
m
i
n
w
h
ic
h
th
e
i
n
te
n
s
it
y
i
s
ad
j
u
s
ted
an
d
th
e
n
o
i
s
e
is
r
e
m
o
v
ed
.
2
.
4
.
F
uzzy
c
-
m
ea
ns
FC
M
i
s
a
cl
u
s
ter
i
n
g
m
e
th
o
d
w
h
er
e
it
d
i
v
id
es
a
g
r
o
u
p
o
f
p
i
x
els
i
n
to
h
o
m
o
g
en
o
u
s
cl
u
s
ter
an
d
ass
i
g
n
s
th
e
p
ix
el
s
ac
co
r
d
in
g
to
t
h
eir
ca
teg
o
r
y
.
I
t
allo
w
s
p
ix
el
s
’
p
o
in
ts
to
b
e
ass
i
g
n
ed
to
m
u
ltip
l
e
clu
s
ter
s
a
n
d
ea
ch
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
3
,
J
u
n
e
2
0
1
9
:
1
8
3
2
-
1
8
4
1
1834
p
ix
el
p
o
in
t
h
a
s
a
d
eg
r
ee
o
f
m
e
m
b
er
s
h
ip
in
a
cl
u
s
ter
to
w
h
i
ch
it
b
elo
n
g
s
.
I
n
th
is
s
eg
m
e
n
t
atio
n
tec
h
n
iq
u
e,
t
h
e
p
ix
el
p
o
in
t
is
s
elec
t
b
ase
o
n
t
h
e
ce
n
ter
o
f
t
h
r
ee
clu
s
ter
w
h
ic
h
ar
e
lo
w
er
,
m
id
d
le
an
d
h
ig
h
er
clu
s
ter
.
E
ac
h
d
ata
p
o
in
t o
f
th
e
cl
u
s
ter
s
h
o
u
ld
eq
u
al
to
o
n
e.
I
t is b
ased
o
n
m
in
i
m
izatio
n
o
f
t
h
e
f
o
llo
w
i
n
g
o
b
j
ec
t
iv
e
f
u
n
ctio
n
:
=
∑
∑
‖
−
‖
2
=
1
=
1
,
1
≤
<
0
(
1
)
w
h
er
e
m
is
a
n
y
r
ea
l
n
u
m
b
er
g
r
ea
ter
th
an
1
,
ij
u
is
th
e
d
e
g
r
ee
o
f
m
e
m
b
er
s
h
ip
o
f
i
x
in
th
e
cl
u
s
ter
j
,
i
x
is
t
h
e
i
th
o
f
d
-
d
i
m
en
s
io
n
al
m
ea
s
u
r
e
d
d
ata,
j
c
is
th
e
d
-
d
i
m
e
n
s
io
n
ce
n
ter
o
f
th
e
cl
u
s
ter
,
an
d
is
an
y
n
o
r
m
ex
p
r
ess
i
n
g
t
h
e
s
i
m
ilar
it
y
b
et
wee
n
a
n
y
m
ea
s
u
r
ed
d
ata
an
d
th
e
ce
n
ter
.
Fu
zz
y
p
ar
titi
o
n
i
n
g
is
ca
r
r
ied
o
u
t
th
r
o
u
g
h
an
iter
ati
v
e
o
p
ti
m
iza
tio
n
o
f
th
e
o
b
j
ec
tiv
e
f
u
n
ctio
n
s
h
o
w
n
a
b
o
v
e,
w
it
h
t
h
e
u
p
d
ate
o
f
m
e
m
b
er
s
h
ip
ij
u
an
d
th
e
clu
s
ter
ce
n
ter
s
j
c
b
y
:
=
1
∑
(
‖
−
‖
‖
−
‖
)
2
−
1
=
1
;
=
∑
.
=
1
∑
=
1
(
2
)
T
h
is
iter
atio
n
w
ill s
to
p
w
h
en
,
{
|
(
+
1
)
|
−
(
)
}
<
(
3
)
w
h
er
e
is
a
ter
m
i
n
atio
n
cr
iter
io
n
b
et
w
ee
n
0
an
d
1
,
w
h
er
ea
s
k
ar
e
th
e
iter
atio
n
s
tep
s
.
2
.
5
.
Act
i
v
e
co
nto
ur
I
n
DW
I
ch
r
o
n
ic
s
tr
o
k
e
les
io
n
i
m
a
g
e,
th
e
ce
r
eb
r
o
s
p
in
al
f
l
u
i
d
(
C
SF
)
s
h
ar
e
th
e
s
i
m
ilar
i
n
te
n
s
it
y
le
v
el
w
it
h
t
h
e
s
tr
o
k
e
lesi
o
n
.
Du
e
to
th
is
m
at
ter
,
FC
M
m
et
h
o
d
h
as
f
ailed
to
s
e
g
m
e
n
t t
h
e
le
s
io
n
s
i
n
ce
th
e
a
lg
o
r
it
h
m
in
th
e
cl
u
s
ter
ca
n
n
o
t
b
e
d
i
f
f
er
e
n
tiated
.
T
o
im
p
r
o
v
e
t
h
i
s
p
er
f
o
r
m
an
ce
,
th
e
C
S
F
ar
ea
i
s
r
e
m
o
v
ed
b
y
u
s
i
n
g
t
h
e
ac
tiv
e
co
n
to
u
r
m
et
h
o
d
.
A
cti
v
e
co
n
to
u
r
i
s
a
m
eth
o
d
t
h
at
cr
ea
te
b
o
u
n
d
ar
ies
i
n
a
n
i
m
a
g
e.
I
t
u
s
es
co
m
p
u
ter
g
en
er
ate
d
cu
r
v
es
to
d
etec
t
an
d
lo
ca
te
o
b
j
ec
t.
T
h
is
m
et
h
o
d
is
o
f
ten
u
s
in
g
in
m
ed
i
ca
l
i
m
a
g
es
to
f
in
d
t
h
e
b
o
u
n
d
a
r
ies
o
f
a
n
o
r
g
a
n
i
n
th
e
i
m
ag
e
s
.
T
h
e
im
a
g
e
is
cla
s
s
i
f
ied
in
to
t
w
o
p
ar
t
w
h
ich
a
r
e
o
b
j
ec
t
r
eg
io
n
an
d
b
ac
k
g
r
o
u
n
d
r
eg
io
n
.
I
n
th
i
s
m
et
h
o
d
th
e
r
eg
io
n
i
s
r
ep
r
esen
t
ed
as th
e
in
s
id
e
an
d
o
u
ts
id
e
r
eg
io
n
s
o
f
t
h
e
ze
r
o
-
lev
el
s
et.
L
e
v
el
s
et
f
r
a
m
e
w
o
r
k
ta
k
e
t
wo
s
ig
n
s
p
o
s
iti
v
e
an
d
n
eg
a
tiv
e
to
d
iv
id
e
i
m
a
g
e
d
o
m
ai
n
.
T
h
e
i
m
ag
e
d
o
m
ai
n
i
s
s
ep
ar
ated
in
to
t
w
o
d
is
j
o
in
t
r
eg
io
n
Ω
1
an
d
Ω
2
.
L
o
ca
l
in
ten
s
it
y
cl
u
s
ter
i
n
g
p
r
o
p
er
t
y
m
ea
n
s
th
at
th
e
in
te
n
s
it
ies
i
n
th
e
n
ei
g
h
b
o
r
h
o
o
d
0
ca
n
b
e
d
iv
id
ed
i
n
to
N
c
l
u
s
ter
s
,
w
it
h
ce
n
ter
,
=
1
,
2
,
…
,
.
I
t
ca
n
b
e
w
r
itte
n
as:
=
∑
∫
|
(
)
−
|
2
(
)
0
=
1
(
4
)
W
h
er
e
th
e
cl
u
s
ter
i
n
t
h
e
ce
n
ter
o
f
th
e
i
-
t
h
cl
u
s
ter
i
s
,
is
t
h
e
m
e
m
b
er
s
h
ip
f
u
n
ctio
n
o
f
t
h
e
r
eg
io
n
,
i.e
.
(
)
=
1
f
o
r
∈
Ω
an
d
(
)
=
0
f
o
r
∉
Ω
.
Or
th
e
co
r
r
esp
o
n
d
in
g
eq
u
at
io
n
ca
n
b
e
w
itte
n
as
:
=
∫
[
(
(
)
)
+
(
(
)
)
]
0
(
5
)
W
h
er
e
th
e
p
o
s
i
tio
n
o
f
ac
ti
v
e
c
o
n
to
u
r
is
d
escr
ib
e
p
ar
a
m
etr
ica
ll
y
b
y
v
(
s
)
=
(
x
(
s
)
,
y
(
s
)
)
,
r
ep
r
esen
ts
i
n
ter
n
al
p
o
ten
tial
en
er
g
y
o
f
t
h
e
c
o
n
to
u
r
,
is
th
e
e
n
er
g
y
th
at
m
o
d
els
e
x
ter
n
al
co
n
s
tr
ai
n
ts
i
m
p
o
s
e
o
n
t
o
th
e
co
n
to
u
r
s
h
ap
e.
G(
y
-
x)
is
t
h
e
Ga
u
s
s
ian
k
er
n
el
ap
p
li
ed
as
w
in
d
o
w
f
u
n
c
tio
n
s
h
o
w
ed
.
(
)
=
{
1
−
|
|
2
2
2
|
|
≤
0
ℎ
(
6
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
A
u
to
ma
ted
s
eg
men
ta
tio
n
a
n
d
cla
s
s
ifica
tio
n
tech
n
iq
u
e
fo
r
b
r
a
in
s
tr
o
ke
(
N
.
M.
S
a
a
d
)
1835
W
h
er
e
is
a
co
n
s
tan
t,
d
is
a
d
is
tan
ce
b
et
w
ee
n
x
an
d
y
p
o
in
ts
.
is
a
s
tan
d
ar
d
d
ev
iatio
n
o
r
s
ca
le
p
ar
am
eter
o
f
Gau
s
s
ia
n
f
u
n
ct
io
n
,
an
d
is
a
r
ad
iu
s
o
f
n
ei
g
h
b
o
r
in
g
p
ix
el
s
.
T
h
e
r
ad
iu
s
o
f
th
e
n
ei
g
h
b
o
r
h
o
o
d
s
h
o
u
ld
b
e
ch
o
o
s
e
ap
p
r
o
p
r
iately
b
ased
o
n
t
h
e
d
eg
r
ee
o
f
in
ten
s
it
y
in
-
h
o
m
o
g
en
e
it
y
.
T
h
e
n
ex
t
s
tep
is
co
n
to
u
r
co
n
s
tr
u
ctio
n
to
d
e
f
in
e
in
i
tial
s
h
ap
e
ar
o
u
n
d
th
e
o
b
j
ec
t
th
at
w
i
ll
s
er
v
e
as
a
n
in
itial
izatio
n
s
et
u
p
.
L
a
s
t
b
u
t
n
o
t
least,
th
e
g
r
ee
d
y
m
et
h
o
d
is
ap
p
lied
to
s
i
m
p
l
if
y
t
h
e
i
m
p
le
m
e
n
tatio
n
o
f
t
h
e
m
i
n
i
m
izatio
n
o
f
e
n
er
g
y
w
it
h
o
u
t
h
a
v
i
n
g
to
p
er
f
o
r
m
an
o
p
tim
izatio
n
alg
o
r
ith
m
m
eth
o
d
li
k
e
g
r
ad
ien
t
d
esce
n
t
[
1
7
]
-
[
2
1
]
.
T
h
e
f
u
n
c
tio
n
o
f
t
h
i
s
m
eth
o
d
i
s
b
y
f
i
n
d
in
g
f
o
r
ea
ch
p
o
in
t
o
f
t
h
e
co
n
to
u
r
t
h
e
cl
o
s
est
lo
ca
l
e
n
er
g
y
m
i
n
i
m
izi
n
g
n
ei
g
h
b
o
r
w
ill
co
n
v
er
g
e
to
th
e
o
v
er
all
g
lo
b
al
m
i
n
i
m
u
m
o
f
t
h
e
co
n
to
u
r
.
T
h
e
m
eth
o
d
u
s
i
n
g
2
eq
u
atio
n
s
w
h
ic
h
ar
e:
=
(
(
)
|
(
)
|
2
+
(
)
|
(
)
|
2
)
2
(
7
)
W
h
er
e
(
)
r
ep
r
esen
t
ad
j
u
s
ts
th
e
elasticit
y
o
f
th
e
ac
ti
v
e
co
n
to
u
r
an
d
(
)
is
ad
j
u
s
ts
th
e
s
ti
f
f
n
e
s
s
o
f
th
e
ac
tiv
e
co
n
to
u
r
.
=
−
∫
‖
∇
(
∗
)
‖
2
(
(
)
)
(
8
)
W
h
er
e
is
a
r
ea
l
w
ei
g
h
ti
n
g
v
alu
e
w
h
ich
f
o
r
o
b
v
io
u
s
r
ea
s
o
n
w
o
u
ld
b
e
p
o
s
iti
v
e,
i
s
a
Gau
s
s
ia
n
w
ei
g
h
ted
k
er
n
el
o
f
d
i
m
e
n
s
io
n
n
,
I
r
ep
r
esen
t
t
h
e
i
n
p
u
t
i
m
ag
e,
∇
i
s
th
e
s
p
atial
g
r
ad
ien
t
f
u
n
ctio
n
a
n
d
(
)
is
th
e
co
n
to
u
r
f
u
n
ctio
n
.
2
.
6
.
F
irst
o
rder
s
t
a
t
is
t
ica
l
m
et
ho
d
First
o
r
d
er
s
tatis
tical
o
r
d
er
m
eth
o
d
is
ap
p
lied
at
th
e
R
OI
to
o
b
tain
s
ev
er
al
f
ea
t
u
r
es.
Si
n
c
e
th
is
R
OI
d
ep
en
d
s
o
n
t
h
e
s
i
g
n
al
o
f
in
te
n
s
it
y
,
t
h
e
R
OI
is
c
lass
if
ied
in
t
o
t
w
o
p
ar
ts
o
f
t
h
e
i
m
ag
e
w
h
i
ch
ar
e
h
y
p
er
in
te
n
s
e
an
d
h
y
p
o
i
n
te
n
s
e
le
s
io
n
.
Me
a
n
,
m
ed
ian
a
n
d
m
o
d
e
ar
e
u
s
ed
to
s
ep
ar
ate
th
e
i
m
ag
e
b
et
w
ee
n
h
y
p
er
in
te
n
s
e
an
d
h
y
p
o
i
n
ten
s
e
lesi
o
n
.
Fo
r
h
y
p
er
i
n
ten
s
e
les
io
n
,
s
ta
n
d
ar
d
d
ev
iati
o
n
is
as
s
ig
n
ed
w
h
ile
in
h
y
p
o
i
n
ten
s
e
les
io
n
m
ea
n
o
f
r
eg
io
n
b
o
u
n
d
ar
y
i
s
as
s
ig
n
e
d
.
T
h
e
s
tan
d
ar
d
d
ev
iatio
n
a
n
d
m
ea
n
o
f
b
o
u
n
d
ar
y
ar
e
u
s
ed
t
o
d
if
f
er
en
tiate
ea
ch
ch
ar
ac
ter
is
tic
o
f
s
tr
o
k
e
lesi
o
n
.
T
h
e
ca
lcu
latio
n
f
o
r
th
e
f
ir
s
t o
r
d
er
m
et
h
o
d
is
s
h
o
w
s
as b
elo
w
:
=
1
∑
(
)
−
1
=
1
(
9
)
M
e
dia
n
=
1
2
(
+
1
)
ℎ
(
1
0
)
=
√
1
∑
(
(
)
−
)
2
−
1
=
0
(
1
1
)
Pe
a
r
s
on
mode
s
k
e
wn
e
s
s
=
|
−
|
(
1
2
)
C
omp
a
c
tn
e
s
s
=
(
ℎ
(
)
)
2
(
1
3
)
W
h
er
e
P
(
i
)
i
s
t
h
e
p
r
o
b
ab
ilit
y
i
n
te
n
s
it
y
lev
el
o
f
th
e
R
OI
,
N
b
e
t
h
e
to
tal
n
u
m
b
er
o
f
g
r
a
y
lev
el
s
i
n
t
h
e
en
tire
R
OI
.
2
.
7
.
B
a
g
g
ed
t
re
e
cla
s
s
if
ier
B
ag
g
ed
tr
ee
cla
s
s
i
f
ier
i
s
a
m
e
th
o
d
th
at
p
er
f
o
r
m
a
co
llectio
n
o
f
d
ata
b
y
an
al
y
ze
t
h
e
w
h
o
le
d
ata
th
a
n
in
d
iv
id
u
all
y
.
T
h
is
c
lass
if
ier
b
u
ild
s
a
n
d
tr
ai
n
s
a
v
ar
iet
y
o
f
d
ec
is
io
n
tr
ee
s
a
n
d
p
r
o
d
u
ce
s
cla
s
s
es
f
r
o
m
t
h
e
m
o
d
e
o
f
in
d
iv
id
u
al
tr
ee
class
.
I
t
ap
p
lies
g
e
n
er
al
tech
n
iq
u
es
b
y
r
es
a
m
p
lin
g
d
ata
f
r
o
m
th
e
p
r
ev
io
u
s
lear
n
i
n
g
.
Ass
u
m
e
th
at
Fo
r
b
=
1
,
.
.
,
B
;
w
h
er
e
B
is
th
e
b
ag
g
in
g
cla
s
s
i
f
ier
(
B
ti
m
e
s
)
th
at
r
ep
ea
t
th
e
tr
ai
n
s
et
o
f
=
1
,
…
,
w
it
h
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
3
,
J
u
n
e
2
0
1
9
:
1
8
3
2
-
1
8
4
1
1836
th
e
r
esp
o
n
s
e
=
1
,
…
,
.
E
ac
h
in
d
i
v
id
u
al
tr
ee
class
w
il
l
o
b
s
er
v
e
th
e
d
ata
f
r
o
m
th
e
p
r
ev
io
u
s
d
ata
o
f
tr
ee
class
an
d
m
a
k
e
th
eir
o
w
n
d
ec
is
io
n
.
T
h
e
d
ec
is
io
n
th
at
m
a
k
e
is
th
en
is
s
elec
ted
an
d
b
ein
g
co
m
p
ar
e
w
i
th
t
h
e
o
th
er
tr
ee
cl
ass
e
s
b
y
av
er
a
g
e
t
h
e
d
ata.
Fo
r
u
n
s
ee
n
d
ata
(
x
’
)
,
t
h
e
d
ata
i
s
s
elec
t
b
y
u
s
i
n
g
t
h
e
m
o
d
o
f
d
ata
in
th
e
tr
ee
class
.
̂
=
1
∑
(
′
)
=
1
(
1
4
)
I
n
t
h
is
ex
p
er
i
m
e
n
t,
t
h
e
b
a
g
g
ed
tr
ee
clas
s
i
f
ier
s
tr
ain
1
2
9
d
ata
f
o
r
th
e
n
u
m
b
er
o
f
b
r
an
c
h
n
o
d
e
s
p
lits
p
er
tr
ee
.
Fig
u
r
e
1
s
h
o
w
s
t
h
e
b
ag
g
e
d
tr
ee
d
iag
r
am
.
Fig
u
r
e
1
.
B
ag
g
ed
tr
ee
class
i
f
ie
r
d
iag
r
a
m
2
.
8
.
P
er
f
o
rm
a
nce
ev
a
lua
t
io
n
T
h
e
p
er
f
o
r
m
a
n
ce
ev
al
u
atio
n
is
ca
lcu
lated
f
r
o
m
th
e
s
eg
m
e
n
t
atio
n
r
esu
lts
w
it
h
t
h
e
n
e
u
r
o
r
ad
io
lo
g
is
t
m
an
u
al
r
e
f
er
en
ce
.
J
ac
ca
r
d
in
d
ex
,
Di
ce
co
ef
f
icie
n
t
(
DC
)
,
f
alse
p
o
s
itiv
e
r
ate
(
FP
R
)
a
n
d
f
alse
n
e
g
ati
v
e
r
ate
(
FNR
)
ar
e
u
s
ed
as
t
h
e
p
er
f
o
r
m
an
ce
m
etr
ic
s
[
1
6
]
.
Fro
m
th
i
s
ca
lcu
latio
n
,
th
e
DW
I
s
e
g
m
e
n
tatio
n
r
es
u
lt
ca
n
b
e
f
u
ll
y
s
eg
m
e
n
t.
T
h
e
p
er
f
o
r
m
an
ce
v
er
if
icat
io
n
f
o
r
class
i
f
icatio
n
ta
k
e
p
ar
t
in
th
e
s
tati
s
tical
ca
lc
u
latio
n
a
n
d
s
h
o
w
n
i
n
co
n
f
u
s
io
n
m
a
tr
ix
attr
ib
u
tes
u
s
in
g
M
A
T
L
A
B
A
P
P
S.
Nu
m
b
e
r
o
f
o
b
s
er
v
atio
n
s
,
tr
u
e
p
o
s
iti
v
e
r
ate
(
T
P
R
)
,
f
alse
n
eg
at
iv
e
r
ate
(
FNR
)
,
p
o
s
it
iv
e
p
r
ed
ictiv
e
v
al
u
es
(
P
P
V)
an
d
f
alse
d
is
co
v
er
y
r
ate
(
FD
R
)
i
s
u
s
ed
to
id
e
n
ti
f
y
th
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
clas
s
i
f
ier
.
B
elo
w
is
t
h
e
s
tati
s
tical
ca
lc
u
la
tio
n
f
o
r
p
er
f
o
r
m
a
n
ce
v
er
i
f
icati
o
n
.
=
(
1
5
)
=
1
−
(
1
6
)
W
h
er
e
tr
u
e
p
o
s
itiv
e
(
T
P
)
is
th
e
n
u
m
b
er
o
f
s
a
m
p
le
s
ar
e
co
r
r
ec
tl
y
clas
s
i
f
ied
w
ith
in
t
h
eir
t
y
p
e
o
f
s
tr
o
k
e
in
t
h
e
p
o
s
itiv
e
s
a
m
p
le
(
P
)
.
P
o
s
itiv
e
p
r
ed
ictiv
e
v
al
u
e
(
P
P
V)
as
s
h
o
w
n
i
n
E
q
u
atio
n
(
1
7
)
is
ca
lcu
lated
a
s
th
e
p
r
o
b
a
b
ilit
y
th
a
t c
o
r
r
ec
tly
cla
s
s
if
y
th
e
t
y
p
e
o
f
s
tr
o
k
e.
T
h
e
f
al
s
e
p
o
s
itiv
e
(
FP
)
is
m
is
c
lass
if
ied
o
f
s
tr
o
k
e.
PPV
=
+
(
1
7
)
L
ast
l
y
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
class
i
f
ier
w
ill b
e
v
er
i
f
ied
u
s
i
n
g
A
cc
to
as sh
o
w
n
i
n
E
q
u
atio
n
(
1
8
)
.
=
+
+
(
1
8
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
A
u
to
ma
ted
s
eg
men
ta
tio
n
a
n
d
cla
s
s
ifica
tio
n
tech
n
iq
u
e
fo
r
b
r
a
in
s
tr
o
ke
(
N
.
M.
S
a
a
d
)
1837
R
ec
eiv
er
o
p
er
atin
g
ch
ar
ac
ter
i
s
tic
(
R
O
C
)
is
u
s
ed
to
in
d
icat
e
T
P
R
an
d
FP
R
f
o
r
ea
ch
t
y
p
e
o
f
s
tr
o
k
e.
Fro
m
t
h
e
y
ea
r
1
9
4
0
,
R
OC
g
r
a
p
h
is
u
s
ef
u
l
f
o
r
i
n
ter
p
r
etin
g
m
ed
ical
test
r
e
s
u
lt
s
.
T
h
e
r
ec
ip
ie
n
t
o
p
er
atin
g
f
ea
t
u
r
e
is
t
h
e
m
etr
ic
u
s
ed
to
ch
ec
k
th
e
q
u
alit
y
o
f
t
h
e
cla
s
s
i
f
icat
io
n
.
Fo
r
ea
ch
t
y
p
e
o
f
s
tr
o
k
e,
R
O
C
u
s
es
t
h
e
t
h
r
es
h
o
ld
v
alu
e
alo
n
g
t
h
e
in
ter
v
al
0
an
d
1
to
th
e
o
u
tp
u
t.
Fo
r
ea
ch
th
r
esh
o
ld
,
t
w
o
v
al
u
es
ar
e
ca
lcu
late
d
b
ased
o
n
th
e
T
P
R
an
d
th
e
FP
R
f
o
r
m
u
la.
=
+
(
1
9
)
W
h
er
e
f
alse
p
o
s
iti
v
e
(
FP
)
is
th
e
n
u
m
b
er
o
f
s
a
m
p
le
s
ar
e
m
is
c
lass
if
ied
w
it
h
i
n
t
h
eir
t
y
p
e
o
f
s
tr
o
k
e
in
t
h
e
p
o
s
itiv
e
s
a
m
p
le.
T
r
u
e
n
eg
ativ
e
(
TN
)
is
th
e
n
u
m
b
er
o
f
s
a
m
p
l
es
ar
e
m
i
s
clas
s
i
f
ied
.
I
n
th
is
R
OC
cu
r
v
e,
T
P
R
is
a
s
et
o
f
o
u
tp
u
ts
g
r
ea
ter
o
r
eq
u
al
to
th
e
th
r
e
s
h
o
ld
,
d
iv
id
ed
b
y
a
tar
g
et
v
al
u
e
a
n
d
FPR
i
s
a
s
et
o
f
o
u
tp
u
ts
les
s
t
h
a
n
th
e
t
h
r
esh
o
ld
,
d
iv
id
ed
b
y
th
e
z
er
o
-
tar
g
et
v
a
lu
e.
T
h
e
A
UC
i
s
a
r
esu
lt o
f
t
h
e
o
v
er
all
q
u
a
lit
y
o
f
th
e
s
tr
o
k
e.
L
ar
g
er
A
U
C
i
n
d
icate
s
b
etter
p
er
f
o
r
m
an
ce
f
o
r
th
e
cla
s
s
i
f
ier
.
=
(
1
>
0
)
(
2
0
)
W
h
er
e
1
is
th
e
p
r
o
b
ab
ilit
y
o
f
s
co
r
e
f
o
r
th
e
t
y
p
e
o
f
s
tr
o
k
e
th
at
is
cla
s
s
i
f
y
in
to
it
s
class
an
d
0
is
v
ic
e
v
er
s
a
to
1
.
3.
RE
SU
L
T
S
A
ND
AN
AL
Y
SI
S
3
.
1
.
F
C
M
a
nd
a
ct
iv
e
c
o
nto
u
r
s
eg
m
e
nta
t
io
n t
ec
hn
iqu
e
T
h
e
s
eg
m
en
tatio
n
r
e
s
u
l
ts
o
f
t
h
e
o
r
ig
in
al
i
m
ag
e
f
r
o
m
s
tr
o
k
e
l
esio
n
ar
e
s
h
o
w
ed
i
n
F
ig
u
r
e
2
.
Fro
m
t
h
e
s
eg
m
e
n
tatio
n
r
es
u
lts
,
FC
M
ca
n
ac
c
u
r
atel
y
s
e
g
m
en
t
t
h
e
h
y
p
er
in
ten
s
e
le
s
io
n
.
T
h
e
h
y
p
er
i
n
t
en
s
e
lesi
o
n
o
f
DW
I
i
m
a
g
es
s
h
o
w
s
h
i
g
h
ac
c
u
r
ac
y
r
esu
lt
co
m
p
ar
e
to
th
e
h
y
p
o
in
t
en
s
e
les
io
n
.
E
v
e
n
th
o
u
g
h
t
h
e
r
esu
lt
s
h
o
w
s
b
y
t
h
e
FC
M
to
w
ar
d
h
y
p
o
in
te
n
s
e
le
s
i
o
n
is
lo
w
ac
cu
r
ac
y
,
th
e
ac
tiv
e
co
n
to
u
r
tech
n
iq
u
e
ca
n
b
e
u
s
ed
to
s
ep
ar
ate
th
e
h
y
p
o
i
n
ten
s
e
lesi
o
n
w
it
h
th
e
C
SF
T
ab
le
2
s
h
o
w
s
t
h
e
p
er
f
o
r
m
a
n
c
e
an
al
y
s
i
s
a
n
d
ev
a
lu
at
io
n
o
f
t
h
e
p
r
o
p
o
s
ed
FC
M
s
eg
m
e
n
tati
o
n
m
e
th
o
d
f
o
r
DW
I
im
a
g
e
s
f
r
o
m
d
ata
s
e
t
1
an
d
d
ataset
2
.
T
h
e
p
er
f
o
r
m
an
ce
o
f
th
e
al
g
o
r
ith
m
i
s
m
ea
s
u
r
ed
u
s
i
n
g
th
e
m
etr
ics
s
u
c
h
as J
ac
ca
r
d
in
d
ex
,
DC
,
FP
R
an
d
FN
R
.
T
h
e
r
esu
lt
s
h
o
w
s
t
h
e
p
r
o
p
o
s
e
d
m
et
h
o
d
o
f
f
er
s
v
er
y
g
o
o
d
s
e
g
m
e
n
tat
io
n
r
e
s
u
lt
f
o
r
s
u
b
-
ac
u
t
e
is
c
h
e
m
ic
s
tr
o
k
e
ac
co
r
d
in
g
to
h
i
g
h
v
alu
e
o
f
J
ac
ca
r
d
in
d
ex
a
n
d
DC
with
lo
w
v
al
u
e
o
f
FP
R
a
n
d
FN
R
.
T
h
e
h
y
p
o
i
n
te
n
s
e
lesi
o
n
is
le
s
s
ac
cu
r
ate
co
m
p
ar
e
to
h
y
p
er
in
te
n
s
e
lesi
o
n
d
u
e
to
th
e
FC
M
tec
h
n
iq
u
e
in
p
r
o
d
u
cin
g
t
h
e
R
OI
i
m
a
g
e.
T
h
e
s
eg
m
en
tatio
n
also
s
h
o
w
s
th
at
th
e
s
tr
o
k
e
les
io
n
i
s
o
v
er
s
eg
m
e
n
tatio
n
b
ased
o
n
th
e
r
esu
lt
s
h
o
w
i
n
FP
R
.
Ho
w
e
v
er
,
th
e
s
e
g
m
en
tatio
n
p
r
o
p
o
s
ed
s
till
ca
n
s
eg
m
e
n
t
t
h
e
R
OI
o
f
b
r
ain
s
tr
o
k
e.
T
h
e
ac
tiv
e
co
n
to
u
r
h
a
s
s
u
cc
e
s
s
f
u
ll
y
s
ep
ar
ated
th
e
C
S
F a
r
ea
w
it
h
t
h
e
h
y
p
o
i
n
te
n
s
e
le
s
io
n
.
T
ab
le
2
.
T
h
e
P
e
r
f
o
r
m
a
n
ce
Me
tr
i
cs f
o
r
DW
I
Stro
k
e
Data
s
et
s
D
a
t
a
se
t
Ty
p
e
o
f
S
t
r
o
k
e
P
e
r
f
o
r
man
c
e
M
e
t
r
i
c
s
Jac
c
a
r
d
i
n
d
e
x
DC
F
P
R
F
N
R
D
a
t
a
se
t
1
A
c
u
t
e
I
sch
e
mi
c
0
.
5
3
0
.
6
6
0
.
0
9
0
.
3
7
C
h
r
o
n
i
c
I
sch
e
mi
c
0
.
2
8
0
.
4
2
0
.
6
1
0
.
1
0
A
c
u
t
e
H
e
mo
r
r
h
a
g
e
0
.
6
8
0
.
7
9
0
.
1
8
0
.
2
5
D
a
t
a
se
t
2
S
u
b
-
A
c
u
t
e
I
sch
e
mi
c
0
.
7
2
0
.
8
3
0
.
0
5
0
.
2
3
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
3
,
J
u
n
e
2
0
1
9
:
1
8
3
2
-
1
8
4
1
1838
Ty
p
e
o
f
S
t
r
o
k
e
D
W
I
I
mag
e
M
a
n
u
a
l
R
e
f
e
r
e
n
c
e
S
e
g
me
n
t
a
t
i
o
n
R
e
su
l
t
A
c
u
t
e
I
sch
e
mi
c
C
h
r
o
n
i
c
I
sch
e
mi
c
A
c
u
t
e
H
e
mo
r
r
h
a
g
e
S
u
b
-
a
c
u
t
e
I
sch
e
mi
c
Fig
u
r
e
2
.
T
h
e
s
eg
m
e
n
tatio
n
r
esu
lt
s
o
f
t
h
e
o
r
ig
i
n
al
i
m
ag
e
f
r
o
m
DW
I
s
tr
o
k
e
lesi
o
n
3
.
2
.
B
a
g
g
ed
t
re
e
cla
s
s
if
ier
T
h
e
p
er
f
o
r
m
an
ce
o
f
th
e
co
n
f
u
s
io
n
m
atr
ices o
f
t
h
e
clas
s
i
f
icat
io
n
m
et
h
o
d
is
s
h
o
w
n
i
n
T
ab
le
3
.
T
ab
le
3
s
h
o
w
s
t
h
e
n
u
m
b
er
o
f
o
b
s
er
v
ati
o
n
s
o
f
ea
c
h
t
y
p
e
o
f
s
tr
o
k
e
lesi
o
n
.
On
l
y
all
ch
r
o
n
ic
is
c
h
e
m
ic
s
tr
o
k
e
s
a
m
p
les ar
e
c
o
r
r
ec
tly
cla
s
s
i
f
ied
in
to
it
s
g
r
o
u
p
.
5
s
a
m
p
le
o
f
ac
u
te
is
c
h
e
m
ic
is
m
i
s
clas
s
i
f
ied
in
to
s
u
b
-
ac
u
te
is
ch
e
m
ic
les
io
n
.
Fo
r
ac
u
te
h
e
m
o
r
r
h
ag
e
s
tr
o
k
e
o
n
l
y
3
s
a
m
p
le
ar
ec
lass
if
ied
co
r
r
ec
tl
y
.
4
o
th
er
s
a
m
p
le
ar
e
m
is
cl
ass
i
f
ied
in
to
s
u
b
-
ac
u
te
is
c
h
e
m
ic
o
th
er
s
i
n
to
ac
u
te
is
c
h
e
m
ic.
Fo
r
s
u
b
-
ac
u
te
is
c
h
e
m
ic
s
tr
o
k
e
6
7
s
a
m
p
le
i
s
clas
s
if
ied
co
r
r
ec
tl
y
w
it
h
i
n
t
h
eir
g
r
o
u
p
s
,
1
s
a
m
p
le
i
s
m
i
s
clas
s
i
f
ied
in
to
ac
u
te
is
c
h
e
m
ic
an
d
o
th
er
s
1
is
m
i
s
clas
s
i
f
ied
in
to
ac
u
te
h
e
m
o
r
r
h
a
g
e.
T
h
e
h
y
p
er
i
n
ten
s
e
lesi
o
n
is
m
o
s
tl
y
m
is
cla
s
s
i
f
ied
in
to
th
eir
cla
s
s
e
s
d
u
e
to
th
e
r
a
n
g
e
o
f
f
ea
t
u
r
es o
f
ea
ch
class
t
h
at
s
h
ar
es t
h
e
s
a
m
e
f
ea
tu
r
e
s
w
it
h
o
th
er
class
.
T
ab
le
3
.
T
h
e
P
e
r
f
o
r
m
a
n
ce
M
e
tr
ic
f
o
r
C
lass
if
icatio
n
Me
t
h
o
d
P
r
e
d
i
c
t
e
d
C
l
a
ss
A
c
u
t
e
H
e
mo
r
r
h
a
g
e
A
c
u
t
e
I
sch
e
mi
c
C
h
r
o
n
i
c
I
sch
e
mi
c
S
u
b
-
a
c
u
t
e
I
sch
e
mi
c
T
r
u
e
C
l
a
ss
A
c
u
t
e
H
e
mo
r
r
h
a
g
e
3
1
4
A
c
u
t
e
I
sch
e
mi
c
15
5
C
h
r
o
n
i
c
I
sch
e
mi
c
33
S
u
b
-
a
c
u
t
e
I
sch
e
mi
c
1
1
67
T
ab
le
4
s
h
o
w
s
th
e
p
er
f
o
r
m
a
n
ce
v
al
u
e
o
f
T
P
R
,
FNR
,
P
P
V
an
d
FD
R
o
f
ea
ch
t
y
p
e
o
f
s
tr
o
k
e.
C
h
r
o
n
ic
is
ch
e
m
ic
s
tr
o
k
e
s
h
o
w
s
b
etter
p
er
f
o
r
m
a
n
ce
b
ase
o
n
h
ig
h
er
v
alu
e
i
n
T
P
R
an
d
P
P
V
w
i
th
lo
w
er
v
alu
e
o
f
F
N
R
an
d
FDR
.
T
h
e
p
r
o
p
o
s
ed
class
if
ica
tio
n
m
et
h
o
d
s
h
o
w
s
t
h
e
p
er
f
o
r
m
a
n
ce
r
es
u
lt
f
o
r
ac
u
te
is
ch
e
m
ic,
ch
r
o
n
i
c
is
ch
e
m
ic
a
n
d
s
u
b
-
ac
u
te
i
s
ch
e
m
ic
i
s
h
ig
h
co
m
p
ar
e
to
t
h
e
ac
u
te
h
e
m
o
r
r
h
a
g
e.
T
h
e
T
P
R
r
esu
lt
f
o
r
ac
u
te
s
tr
o
k
e
s
h
o
w
lo
w
v
al
u
e
b
ec
au
s
e
th
e
s
a
m
p
le
o
f
ac
u
te
h
e
m
o
r
r
h
ag
e
r
e
f
er
to
s
u
b
-
ac
u
te
i
s
ch
e
m
ic
i
s
h
i
g
h
er
.
Ho
w
ev
er
,
t
h
e
P
P
V
o
f
ac
u
te
h
e
m
o
r
r
h
a
g
e
is
cl
o
s
e
to
1
s
in
ce
th
at
t
h
e
ac
u
te
h
e
m
o
r
r
h
ag
e
i
s
co
r
r
ec
tl
y
d
iag
n
o
s
ed
.
Fig
u
r
e
3
s
h
o
w
s
th
e
R
OC
p
lo
t
f
o
r
ea
ch
t
y
p
e
o
f
s
tr
o
k
e.
Fro
m
th
e
R
OC
p
lo
t,
ch
r
o
n
ic
s
tr
o
k
e
s
h
o
w
s
1
0
0
%
o
n
th
e
T
P
R
ax
is
s
h
o
w
i
n
g
t
h
e
s
tr
o
k
e
is
co
r
r
ec
tl
y
class
if
ied
in
to
its
clas
s
.
T
h
e
p
o
s
itiv
e
ty
p
e
o
f
ea
ch
t
y
p
e
o
f
s
tr
o
k
e
f
o
r
ac
u
te
is
c
h
e
m
ic,
c
h
r
o
n
ic
i
s
ch
e
m
ic
a
n
d
ac
u
te
h
e
m
o
r
r
h
ag
e
is
2
%,
1
%
an
d
1
5
%
r
esp
ec
tiv
el
y
.
B
ased
o
n
t
h
e
r
es
u
lt,
ac
u
te
is
c
h
e
m
ic,
ch
r
o
n
ic
is
c
h
e
m
ic
a
n
d
ac
u
te
h
e
m
o
r
r
h
a
g
e
s
h
o
w
s
g
o
o
d
R
O
C
r
es
u
lt
s
s
in
ce
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
A
u
to
ma
ted
s
eg
men
ta
tio
n
a
n
d
cla
s
s
ifica
tio
n
tech
n
iq
u
e
fo
r
b
r
a
in
s
tr
o
ke
(
N
.
M.
S
a
a
d
)
1839
r
esu
lt
o
f
FP
R
is
clo
s
e
to
t
h
e
th
r
es
h
o
ld
v
al
u
e
1
.
T
h
e
r
es
u
lts
f
o
r
ac
u
te
i
s
c
h
e
m
ic,
ch
r
o
n
ic
is
c
h
e
m
ic,
ac
u
te
h
e
m
o
r
r
h
a
g
e
a
n
d
s
u
b
-
ac
u
te
is
c
h
e
m
ic
s
tr
o
k
e
f
o
r
A
U
C
ar
e
9
4
%,
1
0
0
%,8
9
%an
d
9
7
%,
r
esp
e
ctiv
el
y
.
O
v
er
all
th
e
ac
cu
r
ac
y
o
f
t
h
e
clas
s
i
f
ier
p
r
o
p
o
s
ed
is
9
0
.
8
%.
T
ab
le
4
.
P
er
f
o
r
m
a
n
ce
R
es
u
lt
f
o
r
C
lass
i
f
icatio
n
M
eth
o
d
T
P
R
F
N
R
PPV
F
D
R
A
c
u
t
e
I
sch
e
mi
c
0
.
7
5
0
.
2
5
0
.
8
8
0
.
1
2
C
h
r
o
n
i
c
I
sch
e
mi
c
1
0
1
0
A
c
u
t
e
H
e
mo
r
r
h
a
g
e
0
.
3
8
0
.
6
3
0
.
7
5
0
.
2
5
S
u
b
-
a
c
u
t
e
I
sch
e
mi
c
0
.
9
7
0
.
0
3
0
.
8
8
0
.
1
2
(
a)
R
OC
p
lo
t f
o
r
ac
u
te
is
c
h
e
m
ic
(
b
)
R
OC
p
lo
t f
o
r
ch
r
o
n
ic
is
c
h
e
m
ic
(
c)
R
OC
p
lo
t f
o
r
ac
u
te
h
e
m
o
r
r
h
ag
e
(
d
)
R
OC
p
lo
t f
o
r
s
u
b
-
ac
u
te
is
c
h
e
m
ic
Fig
u
r
e
3
.
R
OC
p
lo
t f
o
r
ea
ch
t
y
p
e
o
f
s
tr
o
k
e
4.
CO
NCLU
SI
O
N
I
n
th
i
s
s
t
u
d
y
,
t
h
e
F
C
M
tech
n
iq
u
e
h
as
b
ee
n
i
m
p
le
m
en
ted
w
it
h
ac
ti
v
e
co
n
to
u
r
to
s
eg
m
en
t
t
h
e
s
tr
o
k
e
lesi
o
n
s
in
DW
I
i
m
a
g
es.
T
h
e
s
eg
m
e
n
tat
io
n
o
f
ac
ti
v
e
co
n
to
u
r
is
i
m
p
le
m
en
ted
w
it
h
FC
M
to
s
ep
ar
ate
th
e
C
SF
ar
ea
w
it
h
t
h
e
h
y
p
o
in
te
n
s
e
lesi
o
n
.
T
h
e
s
eg
m
en
ta
tio
n
r
es
u
lt
s
ar
e
co
m
p
ar
ed
w
ith
th
e
m
a
n
u
a
l
r
ef
er
en
ce
to
v
er
i
f
y
th
e
ac
cu
r
ac
y
.
A
cc
o
r
d
in
g
to
t
h
e
ab
o
v
e
r
esu
l
ts
,
F
C
M
p
r
o
v
id
es
g
o
o
d
s
eg
m
e
n
tat
io
n
r
es
u
l
ts
in
h
y
p
er
i
n
te
n
s
e
lesi
o
n
s
esp
ec
iall
y
to
s
u
b
-
ac
u
te
is
c
h
e
m
ic
le
s
io
n
ac
co
r
d
in
g
to
th
e
h
i
g
h
v
a
lu
e
o
f
J
ac
ca
r
d
in
d
e
x
a
n
d
D
C
w
it
h
l
o
w
v
alu
e
o
f
FP
R
an
d
FN
R
p
r
esen
ted
in
s
e
g
m
en
t
th
e
b
r
ain
s
tr
o
k
e.
Ov
er
all
a
v
er
ag
e
v
a
lu
e
s
o
f
J
ac
ca
r
d
in
d
ex
,
DC
,
FP
R
an
d
FNR
ar
e
0
.
5
8
,
0
.
7
1
,
0
.
1
2
an
d
0
.
3
1
r
esp
ec
tiv
el
y
.
A
cti
v
e
co
n
to
u
r
tech
n
iq
u
e
s
h
o
w
s
t
h
at
th
e
b
r
ain
s
tr
o
k
e
DW
I
i
m
a
g
e
ca
n
s
e
g
m
e
n
t
t
h
e
s
tr
o
k
e
lesi
o
n
an
d
s
ep
ar
ate
th
e
C
SF
w
t
it
h
h
y
p
o
in
ten
s
e
lesi
o
n
w
ell.
B
ased
o
n
th
e
f
ea
t
u
r
es
o
f
t
h
e
R
OI
s
e
g
m
en
tatio
n
r
e
s
u
l
t,
ea
ch
t
y
p
e
o
f
s
tr
o
k
e
is
clas
s
if
ied
ac
co
r
d
in
g
t
o
its
clas
s
.
B
ag
g
ed
tr
ee
class
i
f
ier
u
s
e
e
n
s
e
m
b
le
m
eth
o
d
to
p
r
ed
icted
th
e
t
y
p
e
o
f
ea
ch
s
tr
o
k
e
le
s
io
n
.
A
ll
c
h
r
o
n
ic
s
tr
o
k
e
s
a
m
p
les ar
e
co
r
r
ec
tly
cla
s
s
i
f
ied
w
it
h
i
n
it c
l
ass
.
T
h
e
ac
cu
r
ac
y
s
h
o
w
f
o
r
th
i
s
p
er
f
o
r
m
an
ce
m
et
h
o
d
is
9
0
.
8
%.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
3
,
J
u
n
e
2
0
1
9
:
1
8
3
2
-
1
8
4
1
1840
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
e
au
th
o
r
s
w
o
u
ld
lik
e
to
th
a
n
k
to
th
e
U
n
i
v
er
s
iti
T
ek
n
i
k
al
Ma
la
y
s
ia
Me
la
k
a
(
UT
eM
)
,
R
e
h
ab
ilit
atio
n
E
n
g
i
n
ee
r
i
n
g
&
Ass
i
s
ti
v
e
T
e
ch
n
o
lo
g
y
(
R
E
A
T
)
r
esear
ch
g
r
o
u
p
u
n
d
er
C
en
ter
f
o
r
R
o
b
o
tics
&
I
n
d
u
s
tr
ial
Au
to
m
a
tio
n
(
C
eRI
A
)
,
A
d
v
an
ce
d
Dig
ital
Si
g
n
al
P
r
o
ce
s
s
in
g
(
A
D
SP
)
L
ab
a
n
d
Mi
n
is
tr
y
o
f
Hi
g
h
er
E
d
u
ca
tio
n
(
MO
HE
)
,
Ma
la
y
s
ia
f
o
r
s
p
o
n
s
o
r
in
g
t
h
i
s
w
o
r
k
u
n
d
er
p
r
o
j
ec
t
FR
GS/1
/T
K0
4
/FKE
-
C
e
R
I
A
/
F0
0
3
3
4
an
d
th
e
u
s
e
o
f
th
e
e
x
is
ti
n
g
f
ac
ilit
ie
s
to
co
m
p
lete
th
is
p
r
o
j
ec
t.
RE
F
E
R
E
NC
E
S
[1
]
G
o
m
e
s
J,
W
a
c
h
s
m
a
n
A
M
,
“
Ha
n
d
b
o
o
k
o
f
Cli
n
ica
l
Nu
tri
ti
o
n
a
n
d
S
t
ro
k
e
,
”
Ne
w
Y
o
rk
:
Hu
ma
n
a
Pre
ss
,
T
o
to
w
a
,
NJ
,
p
p
.
15
-
3
1
,
2
0
1
3
.
[2
]
Ka
n
c
h
a
n
a
R,
M
e
n
a
k
a
R.
,
“
Co
m
p
u
ter
Re
in
f
o
rc
e
d
A
n
a
l
y
sis
f
o
r
Is
c
h
e
m
i
c
S
tro
k
e
Re
c
o
g
n
it
io
n
:
A
Re
v
ie
w
”
,
In
d
ia
n
J
o
u
rn
a
l
o
f
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
8
,
n
o
.
3
5
,
p
p
.
1
-
9
,
De
c
e
m
b
e
r
2
0
1
5
.
[3
]
P
o
ten
te
M
,
G
e
rh
a
rd
t
H,
Ca
rm
e
li
e
t
P
.
,
“
Ba
sic
a
n
d
T
h
e
ra
p
e
u
ti
c
A
sp
e
c
ts
o
f
A
n
g
io
g
e
n
e
sis”
,
Ce
ll
,
v
o
l.
1
4
6
,
n
o
.
6
,
p
p
.
873
-
8
7
7
,
S
e
p
2
0
1
1
.
[4
]
W
e
rn
ick
M
N,
Ya
n
g
Y,
Bra
n
k
o
v
JG
,
Yo
u
rg
a
n
o
v
G
,
S
tro
th
e
r
S
C.
,
“
M
a
c
h
in
e
L
e
a
rn
in
g
in
M
e
d
ica
l
Im
a
g
in
g
”
,
IEE
E
S
ig
n
a
l
Pro
c
e
ss
in
g
M
a
g
a
zin
e
,
v
o
l
.
2
7
n
o
.
4
,
p
p
.
2
5
-
3
8
,
J
u
l
2
0
1
0
.
[5
]
Kh
a
d
e
m
i
A
,
V
e
n
e
tsa
n
o
p
o
u
l
o
s
A
,
M
o
o
d
y
A
R,
“
Ro
b
u
st
W
h
it
e
M
a
t
ter
Les
io
n
S
e
g
m
e
n
tatio
n
in
F
L
A
I
R
M
RI”
,
IEE
E
Tr
a
n
sa
c
ti
o
n
s
o
n
Bi
o
me
d
ica
l
En
g
i
n
e
e
rin
g
.
,
v
o
l.
5
9
,
n
o
.
3
,
p
p
.
8
6
0
-
7
1
,
M
a
r
2
0
1
2
.
[6
]
S
u
z
u
k
i
K.,
“
P
ix
e
l
-
b
a
se
d
M
a
c
h
in
e
L
e
a
rn
in
g
in
M
e
d
ica
l
Im
a
g
in
g
”
J
o
u
rn
a
l
o
f
Bi
o
me
d
ica
l
Im
a
g
i
n
g
,
v
o
l
1
,
Ja
n
2
0
1
2
.
[7
]
Zh
a
n
g
Y,
Do
n
g
Z,
P
h
il
li
p
s
P
,
W
a
n
g
S
,
Ji
G
,
Ya
n
g
J,
Yu
a
n
T
F
,
“
De
t
e
c
ti
o
n
o
f
S
u
b
jec
ts
a
n
d
Bra
in
Re
g
io
n
s
Re
late
d
to
A
lzh
e
i
m
e
r'
s
Dis
e
a
se
Us
in
g
3
D
M
RI
S
c
a
n
s
Ba
se
d
o
n
Ei
g
e
n
b
ra
in
a
n
d
M
a
c
h
in
e
L
e
a
rn
in
g
”
,
Fro
n
ti
e
rs
in
Co
mp
u
t
a
ti
o
n
a
l
Ne
u
r
o
sc
ien
c
e
,
v
o
l
.
2
,
J
u
n
2
0
1
5
.
[8
]
Ra
m
a
n
i
RG
,
S
iv
a
s
e
lv
i
K.,
“
Clas
si
f
ica
ti
o
n
o
f
P
a
t
h
o
l
o
g
ica
l
M
a
g
n
e
ti
c
Re
so
n
a
n
c
e
I
m
a
g
e
s
o
f
Bra
in
Us
in
g
Da
ta
M
in
in
g
T
e
c
h
n
iq
u
e
s”
,
In
Rec
e
n
t
T
re
n
d
s
a
n
d
Ch
a
ll
e
n
g
e
s
in
Co
m
p
u
t
a
ti
o
n
a
l
M
o
d
e
ls
(
ICRT
CCM
),
2
0
1
7
S
e
c
o
n
d
I
n
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
,
2
0
1
7
,
p
p
.
7
7
-
8
2
,
IEE
E.
[9
]
M
u
d
a
li
D,
T
e
u
n
e
L
K,
Re
n
k
e
n
RJ,
L
e
e
n
d
e
rs
K
L
,
Ro
e
rd
in
k
JB.,
“
Clas
si
f
ica
ti
o
n
o
f
P
a
rk
in
so
n
ian
S
y
n
d
ro
m
e
s
f
ro
m
F
DG
-
P
ET
Bra
in
Da
ta
Us
in
g
De
c
isio
n
T
re
e
s
w
it
h
S
S
M
/
P
CA
F
e
a
tu
re
s”
,
Co
mp
u
t
a
ti
o
n
a
l
a
n
d
M
a
t
h
e
ma
ti
c
a
l
M
e
th
o
d
s
in
M
e
d
icin
e
,
2
0
1
5
.
[1
0
]
Oz
a
M
,
K
a
p
d
i
R,
S
tu
d
e
n
t
B.
,
“
Bra
in
T
u
m
o
r
Dise
a
se
Id
e
n
ti
f
ic
a
ti
o
n
Us
in
g
Ra
n
d
o
m
F
o
re
st
Cla
ss
i
f
iers
”
,
Bra
in
,
v
o
l.
7
n
o
.
1
,
S
e
p
2
0
1
5
.
[1
1
]
F
ra
tello
M
,
Ca
iaz
z
o
G
,
T
ro
jsi
F
,
Ru
ss
o
A
,
T
e
d
e
sc
h
i
G
,
T
a
g
li
a
fe
rri
R,
Esp
o
sito
F
.
,
“
M
u
lt
i
-
V
iew
En
se
m
b
le
Clas
sif
ic
a
ti
o
n
o
f
Bra
in
C
o
n
n
e
c
ti
v
it
y
I
m
a
g
e
s
f
o
r
Ne
u
ro
d
e
g
e
n
e
ra
ti
o
n
T
y
p
e
Disc
ri
m
in
a
ti
o
n
”
,
Ne
u
ro
i
n
fo
rm
a
t
ics
,
v
o
l
.
1
5
,
n
o
.
2
,
p
p
.
1
9
9
-
2
1
3
,
A
p
ril
2
0
1
7
.
[1
2
]
M
a
ier
O,
M
e
n
z
e
BH,
v
o
n
d
e
r
G
a
b
len
tz
J,
Hä
n
i
L
,
He
in
rich
M
P
,
L
ieb
ra
n
d
M
,
W
in
z
e
c
k
S
,
Ba
sit
A
,
Be
n
tl
e
y
P
,
Ch
e
n
L
,
Ch
risti
a
e
n
s
D,“ISL
ES
2
0
1
5
-
A
P
u
b
li
c
Ev
a
lu
a
ti
o
n
Be
n
c
h
m
a
r
k
f
o
r
Isc
h
e
m
i
c
S
tro
k
e
L
e
sio
n
S
e
g
m
e
n
tatio
n
f
ro
m
M
u
lt
is
p
e
c
tral
M
RI”
,
M
e
d
ica
l
Im
a
g
e
An
a
ly
sis
,
v
o
l.
1
,
n
o
.
3
5
,
p
p
.
2
5
0
-
6
9
0
,
Ja
n
2
0
1
7
.
[1
3
]
L
iu
J,
L
i
M
,
W
a
n
g
J,
W
u
F
,
L
iu
T
,
P
a
n
Y,“A
S
u
rv
e
y
o
f
M
RI
-
b
a
se
d
Bra
in
T
u
m
o
r
S
e
g
m
e
n
tatio
n
M
e
t
h
o
d
s”
,
T
sin
g
h
u
a
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
1
9
,
n
o
.
6
,
p
p
.
5
7
8
-
9
5
,
De
c
2
0
1
4
.
[1
4
]
Ba
ra
i
y
a
N,
M
o
d
i
H.,
“
Co
m
p
a
ra
t
iv
e
stu
d
y
o
f
Di
f
f
e
re
n
t
M
e
th
o
d
s
f
o
r
Bra
in
T
u
m
o
r
Ex
trac
ti
o
n
f
ro
m
M
RI
I
m
a
g
e
s
Us
in
g
Im
a
g
e
P
ro
c
e
ss
in
g
”
,
In
d
i
a
n
J
o
u
rn
a
l
o
f
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
Vo
l.
9
,
n
o
.
4
,
Ja
n
2
0
1
6
.
[1
5
]
Ra
h
m
a
n
N
A
,
S
a
a
d
NM,
A
b
d
u
ll
a
h
A
R,
W
a
h
a
b
F
A
.
,
“
T
h
e
In
tern
e
t
o
f
T
h
in
g
s
Be
v
e
ra
g
e
s
Bo
tt
le
S
h
a
p
e
De
f
e
c
t
De
tec
ti
o
n
u
sin
g
Na
ïv
e
Ba
y
e
s
Clas
sif
i
e
r,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Hu
ma
n
a
n
d
T
e
c
h
n
o
l
o
g
y
In
ter
a
c
ti
o
n
(
IJ
Ha
T
I)
,
v
o
l.
2
,
n
o
.
1
,
p
p
.
7
1
-
6
,
A
p
r
2
0
1
8
.
[1
6
]
S
a
a
d
N.M
,
No
o
r
NS
,
A
b
d
u
ll
a
h
A
R,
M
u
d
a
S
,
M
u
d
a
A
F
,
M
u
sa
H.,
“
S
e
g
m
e
n
tatio
n
a
n
d
Clas
sif
ic
a
ti
o
n
A
n
a
ly
sis
T
e
c
h
n
iq
u
e
s
f
o
r
S
tro
k
e
b
a
s
e
d
o
n
Diff
u
sio
n
-
W
e
ig
h
ted
I
m
a
g
e
s”
,
IA
ENG
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Co
mp
u
ter
S
c
ien
c
e
,
v
o
l.
4
4
,
n
o
.
3
,
A
u
g
2
0
1
7
.
[1
7
]
A
b
d
u
ll
a
h
A
R,
A
b
id
u
ll
a
h
NA
,
S
h
a
m
su
d
in
NH
,
A
h
m
a
d
NH
,
Jo
p
ri
M
H.,
“
P
o
w
e
r
Qu
a
li
ty
S
ig
n
a
ls
Clas
sif
ic
a
ti
o
n
S
y
st
e
m
Us
in
g
T
i
m
e
-
F
re
q
u
e
n
c
y
Distrib
u
ti
o
n
”
,
I
n
A
p
p
li
e
d
M
e
c
h
a
n
ics
a
n
d
M
a
ter
ia
ls,
T
ra
n
s
T
e
c
h
P
u
b
li
c
a
ti
o
n
s,
v
o
l
.
4
9
4
,
p
p
.
1
8
8
9
-
1
8
9
4
,
2
0
1
4
.
[1
8
]
A
b
d
u
ll
a
h
A
R,
A
b
id
u
l
lah
NA
,
S
h
a
m
su
d
in
NH
,
A
h
m
a
d
NH
,
Jo
p
ri
M
H.,
“
P
e
rf
o
rm
a
n
c
e
V
e
rif
ica
ti
o
n
o
f
P
o
w
e
r
Qu
a
li
ty
S
ig
n
a
ls
Clas
sif
ica
ti
o
n
S
y
ste
m
”
,
In
Ap
p
li
e
d
M
e
c
h
a
n
ics
a
n
d
M
a
ter
i
a
ls,
T
ra
n
s
T
e
c
h
Pu
b
li
c
a
ti
o
n
s,
v
o
l
.
7
5
2
,
p
p
.
1
1
5
8
-
1
1
6
3
,
2
0
1
5
.
[1
9
]
Jo
p
ri
M
H
,
A
b
d
u
ll
a
h
A
R,
Ka
ss
i
m
NM,
M
a
n
a
p
M
,
Ng
a
ti
m
a
n
N
A
,
Yu
so
f
f
M
R.
,
“
L
o
c
a
li
z
a
ti
o
n
o
f
M
u
lt
ip
le
Ha
rm
o
n
ic
S
o
u
rc
e
s
f
o
r
In
v
e
rter
L
o
a
d
s
Util
izin
g
P
e
ri
o
d
o
g
ra
m
”
,
J
o
u
rn
a
l
o
f
T
e
lec
o
mm
u
n
ica
t
io
n
,
El
e
c
tro
n
ic
a
n
d
Co
m
p
u
te
r
En
g
i
n
e
e
rin
g
(
J
T
EC
),
v
o
l
.
8
,
n
o
.
2
,
p
p
8
7
-
9
1
,
M
a
y
2
0
1
6
.
[2
0
]
Ha
b
b
a
n
,
M
.
F
.
,
M
a
n
a
p
,
M
.
,
A
b
d
u
ll
a
h
,
A
.
R.
,
Jo
p
ri,
M
.
H.
a
n
d
S
u
ti
k
n
o
,
T
.
,
“
A
n
E
v
a
lu
a
ti
o
n
o
f
li
n
e
a
r
ti
m
e
f
re
q
u
e
n
c
y
d
istri
b
u
ti
o
n
A
n
a
l
y
sis
f
o
r
V
S
I
sw
it
c
h
f
a
u
lt
s
id
e
n
ti
f
ica
ti
o
n
”
,
I
n
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
P
o
we
r
El
e
c
tr
o
n
ics
a
n
d
Dr
ive
S
y
ste
ms
(
IJ
PE
DS
)
,
v
o
l.
8
,
n
o
.
1
,
p
p
.
1
-
9.
[2
1
]
Jo
p
ri
M
H
,
A
b
d
u
ll
a
h
A
R,
M
a
n
a
p
M
,
Ha
b
b
a
n
M
F
,
S
u
ti
k
n
o
T
.
,
“
A
n
A
c
c
u
ra
te
Clas
si
f
ic
a
ti
o
n
M
e
th
o
d
o
f
Ha
rm
o
n
ic
S
ig
n
a
ls i
n
P
o
w
e
r
Distrib
u
ti
o
n
S
y
s
tem
b
y
Util
isin
g
S
-
T
ra
n
s
f
o
r
m
”
,
T
e
lko
mn
ika
(
,
v
o
l.
1
5
,
n
o
.
1
,
M
a
r
2
0
1
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
A
u
to
ma
ted
s
eg
men
ta
tio
n
a
n
d
cla
s
s
ifica
tio
n
tech
n
iq
u
e
fo
r
b
r
a
in
s
tr
o
ke
(
N
.
M.
S
a
a
d
)
1841
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
N.
S.
M
.
No
o
r
re
c
e
iv
e
d
B.
En
g
o
f
e
lec
tro
n
ic
a
n
d
c
o
m
p
u
ter
e
n
g
i
n
e
e
rin
g
f
ro
m
Un
iv
e
rsiti
T
e
k
n
ik
a
l
M
a
la
y
sia
M
e
l
a
k
a
(U
T
e
M
)
in
2
0
1
6
.
S
h
e
is
c
u
rre
n
tl
y
p
u
rsu
ig
M
.
S
c
d
e
g
re
e
in
El
e
c
tro
n
ic
En
g
in
e
e
rin
g
a
t
UT
e
M
.
He
r
m
a
in
re
se
a
rc
h
in
tere
st i
s in
t
h
e
f
ield
o
f
im
a
g
e
p
ro
c
e
ss
in
g
a
n
d
m
a
c
h
in
e
lea
rn
in
g
.
N.
M
.
S
a
a
d
is
a
se
n
io
r
lec
tu
re
r
a
t
De
p
a
rtm
e
n
t
o
f
Co
m
p
u
ter
En
g
in
e
e
rin
g
,
F
a
c
u
lt
y
o
f
El
e
c
tro
n
ics
a
n
d
Co
m
p
u
ter
En
g
in
e
e
rin
g
,
Un
iv
e
rsiti
T
e
k
n
ik
a
l
M
a
l
a
y
sia
M
e
lak
a
(U
TeM)
.
S
h
e
re
c
e
iv
e
d
h
e
r
B.
En
g
.
in
M
e
d
ica
l
El
e
c
tro
n
ics
(
2
0
0
1
),
M
a
ste
r
En
g
.
in
T
e
lec
o
m
m
u
n
ica
ti
o
n
s
(2
0
0
4
)
a
n
d
P
h
D
in
Dig
it
a
l
Im
a
g
e
P
r
o
c
e
ss
in
g
(2
0
1
5
)
f
ro
m
Un
iv
e
rsi
ti
T
e
k
n
o
lo
g
i
M
a
lay
sia
(U
T
M
).
He
r
re
se
a
rc
h
a
re
a
in
v
o
lv
e
s
Dig
it
a
l
Im
a
g
e
a
n
d
S
ig
n
a
l
P
ro
c
e
ss
in
g
,
Co
m
p
u
ter
V
i
sio
n
a
n
d
M
e
d
ica
l
I
m
a
g
in
g
.
S
h
e
is
re
g
istere
d
w
it
h
Bo
a
r
d
o
f
En
g
in
e
e
r
M
a
la
y
sia
(BEM
),
M
a
la
y
sia
Bo
a
rd
o
f
T
e
c
h
n
o
lo
g
ist
(
M
BOT
),
In
stit
u
te
f
o
r
El
e
c
tri
c
a
l
a
n
d
El
e
c
tro
n
ics
En
g
in
e
e
rs
(IE
EE
)
S
i
g
n
a
l
P
ro
c
e
ss
in
g
S
o
c
iety
(S
P
S
),
I
EE
E
En
g
in
e
e
rin
g
in
M
e
d
ici
n
e
a
n
d
Bio
lo
g
y
S
o
c
iety
(EM
BS
)
a
n
d
In
te
rn
a
ti
o
n
a
l
A
ss
o
c
iatio
n
o
f
En
g
in
e
e
r
s (IA
EN
G
).
A.R.
Ab
d
u
ll
a
h
re
c
e
iv
e
d
h
is
B.
En
g
.
in
E
lec
tri
c
a
l
En
g
in
e
e
rin
g
(
2
0
0
1
),
M
a
ste
r
d
e
g
re
e
in
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
(2
0
0
4
)
a
n
d
P
h
D
d
e
g
re
e
in
P
o
w
e
r
En
g
in
e
e
rin
g
a
n
d
Dig
it
a
l
S
ig
n
a
l
P
r
o
c
e
ss
in
g
(2
0
1
1
)
f
ro
m
Un
iv
e
rsiti
Tek
n
o
lo
g
i
M
a
la
y
sia
(U
T
M
)
.
He
is
c
u
rre
n
tl
y
a
n
A
ss
o
c
iate
P
ro
f
e
ss
o
r
in
th
e
De
p
a
rtme
n
t
o
f
P
o
w
e
r
El
e
c
tro
n
ic
s
&
Dri
v
e
,
F
a
c
u
lt
y
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
Un
iv
e
rsiti
Tek
n
ik
a
l
M
a
la
y
sia
M
e
la
k
a
(U
TeM)
a
n
d
He
a
d
o
f
A
d
v
a
n
c
e
d
Dig
it
a
l
S
ig
n
a
l
P
r
o
c
e
ss
in
g
(AD
S
P
)
L
a
b
.
He
is
c
u
rre
n
tl
y
a
c
ti
v
e
in
th
e
Ce
n
ter
o
f
E
x
c
e
ll
e
n
c
e
(Co
E)
a
n
d
lea
d
f
o
r
E
-
Be
c
a
f
o
r
Un
iv
e
rsiti
T
e
k
n
ik
a
l
M
a
la
y
sia
M
e
l
a
k
a
(U
T
e
M
).
His
f
i
e
ld
o
f
sp
e
c
i
a
li
z
a
ti
o
n
in
c
lu
d
e
s
a
d
v
a
n
c
e
d
d
ig
it
a
l
sig
n
a
l
p
ro
c
e
ss
in
g
f
o
r
p
o
w
e
r
q
u
a
li
ty
,
re
h
a
b
il
it
a
ti
o
n
e
n
g
in
e
e
rin
g
,
a
ss
isti
v
e
tec
h
n
o
lo
g
y
a
n
d
p
o
w
e
r
e
lec
tro
n
ics
a
n
d
d
riv
e
s.
N.
M.
Ali
re
c
e
iv
e
d
B.
En
g
(
Ho
n
s.)
(M
e
c
h
a
tr
o
n
ic
E
n
g
in
e
e
rin
g
)
f
ro
m
th
e
Un
iv
e
rsiti
Tek
n
ik
a
l
M
a
la
y
sia
M
e
la
k
a
in
2
0
0
9
a
n
d
M
.
S
c
(M
e
c
h
a
tro
n
ic
En
g
in
e
e
rin
g
)
f
r
o
m
Un
iv
e
rsiti
Isla
m
A
n
tara
b
a
n
g
s
a
in
2
0
1
4
.
S
h
e
is
c
u
rre
n
tl
y
p
u
rsu
i
n
g
P
h
D
d
e
g
re
e
in
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
a
t
M
u
lt
im
e
d
ia
Un
iv
e
rsity
,
M
e
lak
a
.
He
r
m
a
in
re
se
a
rc
h
in
te
re
st
is
in
th
e
f
ield
o
f
i
m
a
g
e
p
ro
c
e
ss
in
g
,
in
telli
g
e
n
t
s
y
ste
m
a
n
d
m
a
c
h
in
e
lea
rn
in
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.