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.
7
,
No
.
1
,
Feb
r
u
ar
y
201
7
,
p
p
.
3
8
3
~
3
9
1
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v7
i
1
.
p
p
3
8
3
-
3
9
1
383
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
jo
u
r
n
a
l.c
o
m/o
n
lin
e/in
d
ex
.
p
h
p
/I
JE
C
E
Co
m
pa
riso
n of
T
hree
Seg
m
en
tatio
n Metho
ds
for
Br
ea
st
Ultras
o
und I
m
a
g
es
Ba
sed o
n
Lev
el
Set
a
nd
Mo
rpho
lo
g
ica
l
O
pera
tions
Dew
i P
utr
ie
L
esta
ri
,
Sa
rif
ud
din
M
a
dend
a
,
E
rna
s
t
uti
,
E
ri
P
ra
s
et
y
o
Wibo
w
o
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
a
n
d
I
n
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
,
G
u
n
a
d
a
rm
a
Un
iv
e
rsit
y
,
In
d
o
n
e
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
5
,
2
0
1
6
R
ev
i
s
ed
No
v
2
5
,
2
0
1
6
A
cc
ep
ted
Dec
1
5
,
2
0
1
6
Bre
a
st
c
a
n
c
e
r
is
o
n
e
o
f
th
e
m
a
jo
r
c
a
u
se
s
o
f
d
e
a
th
a
m
o
n
g
w
o
m
e
n
a
ll
o
v
e
r
th
e
w
o
rld
.
T
h
e
m
o
st
f
r
e
q
u
e
n
tl
y
u
se
d
d
iag
n
o
sis
to
o
l
to
d
e
tec
t
b
re
a
st
c
a
n
c
e
r
is
u
lt
ra
so
u
n
d
.
Ho
w
e
v
e
r,
t
o
se
g
m
e
n
t
th
e
b
re
a
st
u
lt
ra
so
u
n
d
im
a
g
e
s
is
a
d
iff
icu
lt
th
in
g
.
S
o
m
e
stu
d
ies
sh
o
w
th
a
t
th
e
a
c
ti
v
e
c
o
n
to
u
r
m
o
d
e
ls
h
a
v
e
b
e
e
n
p
ro
v
e
d
to
b
e
t
h
e
m
o
st
su
c
c
e
ss
f
u
l
m
e
th
o
d
s
f
o
r
m
e
d
ica
l
i
m
a
g
e
se
g
m
e
n
tatio
n
.
T
h
e
le
v
e
l
se
t
m
e
th
o
d
is
a
c
las
s
o
f
c
u
rv
e
e
v
o
lu
ti
o
n
m
e
th
o
d
s
b
a
se
d
o
n
th
e
g
e
o
m
e
tri
c
a
c
ti
v
e
c
o
n
to
u
r
m
o
d
e
l.
M
o
rp
h
o
l
o
g
ica
l
o
p
e
ra
ti
o
n
d
e
sc
rib
e
s
a
ra
n
g
e
o
f
i
m
a
g
e
p
ro
c
e
ss
in
g
tec
h
n
iq
u
e
th
a
t
d
e
a
l
w
it
h
th
e
sh
a
p
e
o
f
f
e
a
tu
re
s
in
a
n
ima
g
e
.
M
o
rp
h
o
l
o
g
ica
l
o
p
e
ra
ti
o
n
s
a
re
a
p
p
li
e
d
to
re
m
o
v
e
i
m
p
e
r
f
e
c
ti
o
n
s
t
h
a
t
in
tro
d
u
c
e
d
d
u
rin
g
se
g
m
e
n
tatio
n
.
I
n
t
h
i
s
p
a
p
e
r
,
w
e
h
a
v
e
e
v
a
l
u
a
te
d
t
h
r
e
e
l
e
v
e
l
s
e
t
m
e
t
h
o
d
s
t
h
a
t
c
o
m
b
i
n
e
d
w
i
t
h
m
o
r
p
h
o
l
o
g
i
c
a
l
o
p
e
r
a
t
i
o
n
s
t
o
se
g
m
e
n
t
t
h
e
b
r
e
a
s
t
l
e
s
i
o
n
s
.
T
h
e
lev
e
l
se
t
m
e
th
o
d
s th
a
t
u
se
d
i
n
o
u
r
re
se
a
rc
h
a
re
th
e
Ch
a
n
V
e
se
(
C
-
V
)
m
o
d
e
l,
th
e
S
e
lec
ti
v
e
Bin
a
ry
a
n
d
G
a
u
ss
ian
F
i
lt
e
rin
g
Re
g
u
lariz
e
d
L
e
v
e
l
S
e
t
(
S
BG
F
RL
S
)
m
o
d
e
l
a
n
d
th
e
D
istan
c
e
Re
g
u
lariz
e
d
L
e
v
e
l
S
e
t
Ev
o
lu
ti
o
n
(
DRL
S
E
)
m
o
d
e
l.
F
u
rth
e
rm
o
re
,
to
e
v
a
lu
a
te
th
e
m
e
th
o
d
,
w
e
c
o
m
p
a
re
d
th
e
se
g
m
e
n
ted
b
re
a
st
les
io
n
th
a
t
o
b
t
a
in
e
d
b
y
e
a
c
h
m
e
th
o
d
w
it
h
th
e
les
io
n
t
h
a
t
o
b
tai
n
e
d
m
a
n
u
a
ll
y
b
y
ra
d
io
lo
g
ist
s.
T
h
e
e
v
a
lu
a
ti
o
n
is
d
o
n
e
b
y
f
o
u
r
m
e
tri
c
s
:
Dic
e
S
i
m
il
a
rit
y
Co
e
ff
icie
n
t
(
DSC),
T
ru
e
-
P
o
siti
v
e
Ra
ti
o
(T
P
R)
,
T
ru
e
-
Ne
g
a
ti
v
e
Ra
ti
o
(T
NR),
a
n
d
A
c
c
u
ra
c
y
(
A
CC).
Ou
r
e
x
p
e
ri
m
e
n
tal
re
su
lt
s
w
it
h
3
0
b
re
a
st
u
lt
ra
so
u
n
d
im
a
g
e
s
sh
o
w
e
d
th
a
t
t
h
e
C
-
V
m
o
d
e
l
th
a
t
c
o
m
b
in
e
d
w
it
h
m
o
rp
h
o
lo
g
ica
l
o
p
e
ra
ti
o
n
s
h
a
v
e
b
e
tt
e
r
p
e
rf
o
r
m
a
n
c
e
th
a
n
th
e
o
th
e
r
tw
o
m
e
th
o
d
s ac
c
o
rd
i
n
g
to
m
e
a
n
v
a
lu
e
o
f
DSC
m
e
tri
c
s.
K
ey
w
o
r
d
:
B
r
ea
s
t le
s
io
n
L
e
v
el
s
et
Mo
r
p
h
o
lo
g
ical
o
p
er
atio
n
s
Seg
m
en
tatio
n
Ultr
aso
u
n
d
Co
p
y
rig
h
t
©
2
0
1
7
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
:
De
w
i P
u
tr
ie
L
e
s
tar
i
,
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
I
n
f
o
r
m
atio
n
T
ec
h
n
o
lo
g
y
,
Gu
n
ad
ar
m
a
Un
iv
er
s
it
y
,
Ma
r
g
o
n
d
a
R
a
y
a
R
o
ad
,
P
o
n
d
o
k
C
i
n
a,
Dep
o
k
,
W
est J
av
a
1
6
4
2
4
,
I
n
d
o
n
esia.
E
m
ail:
d
e
w
ip
u
tr
ie2
6
@
g
m
ail.
c
o
m
1.
I
NT
RO
D
UCT
I
O
N
B
r
ea
s
t
ca
n
ce
r
is
o
n
e
o
f
th
e
le
ad
in
g
ca
u
s
e
s
o
f
d
ea
th
a
m
o
n
g
w
o
m
en
b
o
t
h
i
n
d
ev
elo
p
ed
co
u
n
tr
ie
s
a
n
d
in
d
e
v
elo
p
in
g
co
u
n
tr
ie
s
[
1
]
.
T
h
e
ea
r
l
y
d
etec
tio
n
ca
n
r
ed
u
ce
th
e
d
ea
t
h
r
ate
an
d
h
elp
f
u
l
t
h
e
tr
ea
t
m
e
n
t
o
f
b
r
ea
s
t
ca
n
ce
r
[
2
]
.
Ho
w
e
v
er
,
ea
r
ly
d
etec
tio
n
r
eq
u
ir
es
an
ac
c
u
r
ate
an
d
r
eliab
le
d
iag
n
o
s
is
t
h
at
b
e
ab
le
to
d
is
tin
g
u
i
s
h
b
en
ig
n
an
d
m
a
lig
n
a
n
t
lesi
o
n
s
,
w
h
er
e
i
m
a
g
i
n
g
tec
h
n
iq
u
es
p
lay
a
n
i
m
p
o
r
tan
t
ta
s
k
[
3
]
.
On
e
o
f
th
e
m
o
s
t
f
r
eq
u
en
tl
y
u
s
ed
d
iag
n
o
s
i
s
to
o
ls
to
d
etec
t b
r
ea
s
t c
an
ce
r
is
u
ltr
aso
u
n
d
[
4
]
.
Ultr
aso
u
n
d
i
m
a
g
e
s
e
g
m
e
n
tatio
n
is
a
d
i
f
f
ic
u
lt
ta
s
k
b
ec
au
s
e
t
h
ese
i
m
a
g
es
co
n
tai
n
s
tr
o
n
g
s
p
e
ck
le
n
o
i
s
e
an
d
atte
n
u
a
tio
n
ar
ti
f
ac
t
s
[
5
]
.
S
ev
er
al
s
tu
d
ies
s
h
o
w
t
h
at
th
e
a
ctiv
e
co
n
to
u
r
m
o
d
els
h
a
v
e
b
e
en
p
r
o
v
ed
to
b
e
t
h
e
m
o
s
t s
u
cc
es
s
f
u
l
m
et
h
o
d
s
f
o
r
i
m
ag
e
s
eg
m
e
n
tatio
n
[
6
]
.
T
h
e
b
asic id
ea
o
f
th
e
ac
ti
v
e
co
n
to
u
r
m
o
d
el
i
s
to
d
ev
elo
p
a
cu
r
v
es
u
n
d
er
s
o
m
e
co
n
s
tr
a
in
ts
t
o
ex
tr
ac
t
t
h
e
d
esire
d
o
b
j
ec
t
[
7
]
.
T
h
er
e
ar
e
tw
o
t
y
p
es
o
f
ac
tiv
e
co
n
to
u
r
m
o
d
el
s
:
p
ar
a
m
etr
ic
ac
ti
v
e
c
o
n
to
u
r
m
o
d
els
a
n
d
g
e
o
m
etr
i
c
ac
tiv
e
co
n
to
u
r
m
o
d
els
[
8
]
.
T
h
e
d
if
f
er
en
ce
o
f
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
C
o
mp
a
r
is
o
n
o
f
Th
r
ee
S
eg
men
t
a
tio
n
Meth
o
d
s
fo
r
B
r
ea
s
t U
ltr
a
s
o
u
n
d
I
ma
g
es B
a
s
ed
o
n
…
(
Dewi P
.
L
.
)
384
p
ar
am
etr
ic
ac
ti
v
e
co
n
to
u
r
m
o
d
els
an
d
g
eo
m
e
tr
ic
ac
tiv
e
co
n
to
u
r
m
o
d
els
i
s
t
h
at
th
e
f
o
r
m
er
r
ep
r
esen
t
cu
r
v
es
an
d
s
u
r
f
ac
es e
x
p
licitl
y
in
t
h
eir
p
ar
am
etr
ic
f
o
r
m
s
,
b
u
t th
e
latt
er
r
ep
r
esen
t c
u
r
v
es a
n
d
s
u
r
f
ac
es i
m
p
lici
tl
y
[
9
].
Geo
m
etr
ic
ac
ti
v
e
co
n
to
u
r
m
o
d
els
h
av
e
m
a
n
y
ad
v
an
ta
g
e
s
o
v
er
p
ar
am
etr
ic
ac
ti
v
e
co
n
to
u
r
m
o
d
els,
s
u
c
h
as
co
m
p
u
tatio
n
al
s
i
m
p
lic
it
y
a
n
d
t
h
e
ab
ilit
y
to
c
h
an
g
e
c
u
r
v
e
to
p
o
lo
g
y
d
u
r
i
n
g
d
ef
o
r
m
a
tio
n
[
9
]
.
T
h
e
lev
el
s
et
m
et
h
o
d
is
a
c
lass
o
f
c
u
r
v
e
ev
o
lu
tio
n
m
et
h
o
d
b
ased
o
n
th
e
g
eo
m
etr
ic
ac
ti
v
e
co
n
to
u
r
m
o
d
el
s
,
w
h
ic
h
t
h
i
s
m
et
h
o
d
in
tr
o
d
u
ce
d
b
y
Os
h
er
-
Set
h
ia
n
th
at
o
v
er
ca
m
e
th
e
p
r
o
b
lem
s
o
f
th
e
as
s
o
ciate
d
w
i
th
class
ical
en
er
g
y
m
i
n
i
m
izatio
n
ap
p
r
o
ac
h
es
[
9
]
.
T
h
e
b
asic
id
ea
o
f
th
e
lev
el
s
e
t
m
et
h
o
d
is
to
r
ep
r
esen
t
a
co
n
to
u
r
as
th
e
ze
r
o
lev
e
l
s
et
o
f
a
h
i
g
h
er
d
i
m
e
n
s
io
n
al
f
u
n
ctio
n
,
ca
ll
ed
a
le
v
el
s
et
f
u
n
ct
io
n
(
L
S
F),
an
d
f
o
r
m
u
la
te
th
e
m
o
tio
n
o
f
t
h
e
co
n
to
u
r
as
t
h
e
e
v
o
lu
t
io
n
o
f
th
e
lev
el
s
et
f
u
n
ctio
n
[
10
]
.
A
d
esira
b
le
ad
v
a
n
tag
e
o
f
t
h
e
le
v
e
l
s
et
m
e
th
o
d
i
s
t
h
at
th
e
y
ca
n
r
ep
r
esen
t
co
n
to
u
r
s
o
f
co
m
p
le
x
to
p
o
lo
g
y
a
n
d
ar
e
ab
le
to
h
an
d
le
to
p
o
lo
g
ical
ch
a
n
g
e
s
,
s
u
c
h
as
s
p
litt
i
n
g
an
d
m
er
g
in
g
,
in
a
n
at
u
r
al
an
d
ef
f
icien
t
w
a
y
,
w
h
ic
h
is
n
o
t a
ll
o
w
ed
in
p
ar
a
m
etr
ic
ac
ti
v
e
co
n
to
u
r
m
o
d
els
[
10
].
I
n
th
is
p
a
p
e
r
,
w
e
c
o
m
p
a
r
e
d
th
r
e
e
l
e
v
e
l
s
e
t
m
et
h
o
d
s
th
a
t
c
o
m
b
i
n
e
d
w
ith
m
o
r
p
h
o
l
o
g
ic
a
l
o
p
er
a
t
i
o
n
s
f
o
r
a
u
t
o
m
at
i
c
s
eg
m
en
t
at
i
o
n
o
f
b
r
ea
s
t
u
lt
r
a
s
o
u
n
d
im
ag
es
.
T
h
r
e
e
lev
e
l
s
et
m
o
d
e
l
s
u
s
e
d
in
th
is
p
a
p
e
r
a
r
e
th
e
C
h
an
V
e
s
e
(C
-
V
)
m
o
d
e
l
[
1
1
]
,
t
h
e
S
e
l
ec
t
iv
e
B
in
a
r
y
an
d
G
au
s
s
i
an
Fi
l
te
r
i
n
g
R
eg
u
l
a
r
i
z
e
d
L
ev
el
S
et
(
SB
G
F
R
L
S
)
m
o
d
e
l
[
7
]
a
n
d
th
e
D
is
t
an
ce
R
eg
u
l
a
r
i
ze
d
L
ev
e
l
Se
t
E
v
o
lu
ti
o
n
(
DR
L
S
E
)
m
o
d
e
l
[
10
]
.
Fu
r
th
e
r
m
o
r
e
,
t
o
e
v
a
l
u
at
e
d
th
e
m
e
th
o
d
s
,
w
e
co
m
p
ar
ed
th
e
s
e
g
m
en
ted
b
r
ea
s
t
les
io
n
t
h
at
o
b
tain
ed
b
y
ea
ch
m
eth
o
d
w
i
th
t
h
e
le
s
io
n
t
h
at
o
b
tain
ed
m
an
u
all
y
b
y
r
ad
io
lo
g
is
ts
u
s
i
n
g
ar
ea
-
b
ased
s
e
g
m
e
n
tat
io
n
as
s
ess
m
en
t
m
etr
ic
s
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
T
u
m
o
r
ar
ea
o
n
b
r
ea
s
t
u
ltra
s
o
u
n
d
i
m
ag
e
s
o
b
tain
ed
b
y
m
ea
n
s
p
er
f
o
r
m
t
h
e
s
e
g
m
en
ta
ti
o
n
p
r
o
ce
s
s
.
I
n
tak
e
o
f
th
e
t
u
m
o
r
ar
ea
s
h
o
u
l
d
b
e
d
o
n
e
p
r
o
p
er
ly
,
b
ec
a
u
s
e
th
e
co
n
to
u
r
o
f
s
eg
m
e
n
tatio
n
r
es
u
lt
is
v
er
y
estab
li
s
h
th
e
tr
u
t
h
o
f
t
h
e
g
r
o
u
p
in
g
p
r
o
ce
s
s
t
h
e
t
y
p
es
o
f
t
u
m
o
r
s
.
T
h
er
ef
o
r
e,
w
e
n
ee
d
to
m
ea
s
u
r
e
th
e
ac
c
u
r
ac
y
o
f
th
e
s
eg
m
e
n
tatio
n
m
et
h
o
d
s
th
at
u
s
ed
in
th
is
r
esear
c
h
.
T
h
is
is
d
o
n
e
b
y
p
er
f
o
r
m
i
n
g
v
alid
atio
n
o
f
th
e
s
e
g
m
e
n
tatio
n
m
et
h
o
d
p
r
o
p
o
s
ed
o
r
R
eg
io
n
o
f
I
n
ter
est
(
R
OI
)
i
m
a
g
es
ag
a
in
s
t
th
e
s
eg
m
e
n
tatio
n
m
a
n
u
a
ll
y
b
y
an
e
x
p
er
ien
ce
d
r
ad
io
lo
g
is
t
o
r
Gr
o
u
n
d
T
r
u
th
(
GT
)
im
ag
e.
T
h
e
r
esear
ch
s
ta
g
es
in
t
h
is
p
ap
er
ar
e
s
h
o
w
n
b
y
a
r
esear
ch
s
ch
e
m
e
i
n
Fig
u
r
e
1.
Fig
u
r
e
1
.
Sch
e
m
e
o
f
th
e
R
esea
r
ch
Stag
e
s
B
ased
o
n
Fig
u
r
e
1
,
w
e
ca
n
co
n
clu
d
e
t
h
at
t
h
is
r
esear
c
h
is
d
i
v
id
ed
in
to
t
w
o
s
ta
g
es.
T
h
e
f
ir
s
t
s
ta
g
e
is
th
e
s
eg
m
e
n
tatio
n
o
f
t
h
e
b
r
ea
s
t
u
ltra
s
o
u
n
d
i
m
ag
e
an
d
th
e
s
ec
o
n
d
s
ta
g
e
i
s
v
alid
atio
n
t
h
e
r
es
u
lts
o
f
s
e
g
m
en
ta
tio
n
m
et
h
o
d
.
T
h
e
f
ir
s
t
s
tag
e
b
e
g
i
n
s
w
it
h
a
s
e
g
m
en
tatio
n
p
r
o
ce
s
s
o
f
b
r
ea
s
t
u
ltra
s
o
u
n
d
i
m
a
g
es
u
s
i
n
g
lev
el
s
et
alg
o
r
ith
m
.
T
h
e
n
e
x
t
p
r
o
ce
s
s
is
th
e
ap
p
licatio
n
o
f
m
o
r
p
h
o
lo
g
i
ca
l
o
p
er
atio
n
s
to
th
e
i
m
ag
e
o
f
th
e
r
es
u
lts
o
f
le
v
el
s
et
m
et
h
o
d
.
T
h
is
s
e
g
m
e
n
tat
i
o
n
p
r
o
ce
s
s
w
il
l
p
r
o
d
u
ce
th
e
R
e
g
io
n
o
f
I
n
ter
est
(
R
OI
)
i
m
ag
e.
T
h
en
i
n
t
h
e
u
ltra
s
o
u
n
d
i
m
a
g
es
also
d
o
n
e
s
e
g
m
e
n
tatio
n
m
a
n
u
a
ll
y
t
h
at
p
r
o
d
u
ce
d
th
e
Gr
o
u
n
d
T
r
u
t
h
(
GT
)
i
m
ag
e.
T
h
e
v
alid
atio
n
p
r
o
ce
s
s
i
s
d
o
n
e
b
y
co
m
p
ar
in
g
t
h
e
R
OI
i
m
a
g
e
with
th
e
GT
i
m
a
g
e
u
s
in
g
ar
ea
-
b
ased
s
eg
m
e
n
tatio
n
ass
es
s
m
en
t
m
etr
ics
m
et
h
o
d
.
2
.
1
.
Seg
m
ent
a
t
io
n
M
et
ho
ds
I
m
ag
e
s
e
g
m
en
ta
tio
n
is
a
p
r
o
c
ess
o
f
d
iv
id
in
g
th
e
i
m
a
g
e
in
t
o
s
ev
er
al
ar
ea
s
h
o
m
o
g
e
n
eo
u
s
b
ased
o
n
ce
r
tain
s
i
m
i
lar
it
y
cr
iter
ia
s
u
ch
as
i
n
ten
s
it
y
,
co
lo
r
an
d
tex
tu
r
e
[
1
2
]
.
T
h
e
au
to
m
atic
s
eg
m
e
n
tatio
n
p
r
o
ce
s
s
i
s
a
U
l
t
r
a
so
u
n
d
I
mag
e
G
T
I
mag
e
R
O
I
I
mag
e
L
e
v
e
l
S
e
t
M
o
r
p
h
o
l
o
g
i
c
a
l
O
p
e
r
a
t
i
o
n
s
A
r
e
a
-
B
a
se
d
S
e
g
me
n
t
a
t
i
o
n
A
ss
e
ssm
e
n
t
M
e
t
r
i
c
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
7
:
3
8
3
–
391
385
n
ec
es
s
ar
y
s
tep
in
an
a
l
y
s
is
o
f
m
ed
ical
i
m
a
g
es.
Ho
w
e
v
er
,
th
e
ac
cu
r
ate
an
d
r
eliab
le
s
eg
m
e
n
t
atio
n
m
et
h
o
d
s
ar
e
a
k
e
y
r
eq
u
ir
e
m
en
t
f
o
r
th
e
ex
tr
ac
tio
n
o
f
q
u
alitati
v
e
o
r
q
u
an
titativ
e
i
n
f
o
r
m
at
io
n
f
r
o
m
i
m
ag
es
[
13
]
.
I
n
th
is
r
esear
ch
,
to
s
eg
m
en
t
th
e
b
r
ea
s
t
lesi
o
n
w
e
u
s
ed
th
r
ee
m
e
th
o
d
s
b
ased
o
n
lev
el
s
et
t
h
at
ea
ch
m
et
h
o
d
co
m
b
i
n
ed
w
it
h
m
o
r
p
h
o
lo
g
ical
o
p
er
atio
n
s
.
T
h
r
e
e
le
v
e
l
s
e
t
m
o
d
e
l
s
u
s
e
d
in
th
is
p
a
p
e
r
a
r
e
th
e
C
h
a
n
V
e
s
e
m
o
d
e
l
,
th
e
S
e
l
e
ct
iv
e
B
in
a
r
y
an
d
G
au
s
s
i
an
F
i
l
t
e
r
i
n
g
R
eg
u
l
a
r
i
z
e
d
L
ev
el
Se
t
m
o
d
e
l
an
d
th
e
D
is
t
an
c
e
R
eg
u
l
a
r
iz
e
d
L
ev
e
l
S
e
t
E
v
o
lu
ti
o
n
m
o
d
e
l
.
2
.
1
.
1.
Cha
n Ve
s
e
M
o
del
C
h
a
n
an
d
V
ese
p
r
o
p
o
s
ed
a
r
eg
io
n
-
b
ased
ac
ti
v
e
co
n
to
u
r
m
o
d
els
w
h
ic
h
is
a
s
p
ec
ial
ca
s
e
o
f
Mu
m
f
o
r
d
-
Sh
a
h
f
o
r
m
u
latio
n
to
h
a
n
d
le
p
r
o
b
lem
s
o
f
ed
g
e
-
b
ased
ac
t
iv
e
c
o
n
to
u
r
m
o
d
els [
1
]
.
L
e
t
is
a
d
o
m
ai
n
f
o
r
a
g
i
v
en
i
m
a
g
e
y
x
I
,
,
th
ey
r
ep
r
esen
ted
a
co
n
to
u
r
C
i
m
p
licitl
y
th
r
o
u
g
h
L
i
p
s
ch
itz
f
u
n
ctio
n
:
,
y
x
th
at
d
ef
in
ed
as
f
o
llo
w
s
:
0
,
:
,
0
,
:
,
0
,
:
,
y
x
y
x
C
o
u
t
s
i
d
e
y
x
y
x
C
i
n
s
i
d
e
y
x
y
x
C
(
1
)
Seg
m
en
tatio
n
u
s
i
n
g
C
h
a
n
V
es
e
(
C
-
V
)
m
o
d
e
l
i
s
d
o
n
e
b
y
m
i
n
i
m
izi
n
g
a
n
en
er
g
y
f
u
n
ct
i
o
n
[
7
]
.
T
h
e
f
o
r
m
u
la
f
u
n
ctio
n
t
h
is
en
e
r
g
y
ex
p
r
ess
ed
b
y
u
s
in
g
Hea
v
i
s
id
e
f
u
n
ct
io
n
t
h
at
d
e
f
i
n
ed
in
E
q
u
atio
n
2
an
d
Dir
ac
f
u
n
ctio
n
th
a
t
d
ef
i
n
ed
in
E
q
u
ati
o
n
3
.
0
,
0
0
,
1
j
i
k
a
j
i
k
a
H
(
2
)
H
d
d
(
3
)
Fu
r
t
h
er
m
o
r
e,
C
h
an
a
n
d
Vese
f
o
r
m
u
late
e
n
er
g
y
f
u
n
ctio
n
t
h
at
s
h
o
w
n
b
y
th
e
f
o
r
m
u
la
[
9
]
:
d
x
d
y
y
x
H
c
y
x
I
d
x
d
y
y
x
H
c
y
x
I
,
1
,
,
,
2
2
2
2
1
1
(
4
)
w
h
er
e
≥
0
,
v
0
,
1
0
,
2
0
,
ar
e
f
ix
ed
p
ar
a
m
eter
s
,
co
n
tr
o
ls
th
e
s
m
o
o
t
h
n
e
s
s
o
f
ze
r
o
le
v
el
s
et
,
v
in
cr
ea
s
e
s
th
e
p
r
o
p
ag
atio
n
s
p
ee
d
,
1
co
n
t
r
o
l
th
e
i
m
ag
e
d
ata
d
r
iv
en
f
o
r
ce
in
s
id
e,
2
co
n
tr
o
l
th
e
im
a
g
e
d
ata
d
r
iv
en
f
o
r
ce
o
u
ts
id
e
th
e
co
n
to
u
r
,
is
th
e
g
r
ad
ien
t
o
p
er
ato
r
,
c
1
is
is
th
e
av
er
ag
e
i
n
ten
s
itie
s
in
s
i
d
e
th
e
co
n
to
u
r
th
at
r
ep
r
esen
ted
in
E
q
u
atio
n
(
5
)
,
an
d
c
2
is
th
e
av
er
ag
e
in
te
n
s
it
ies
o
u
t
s
id
e
th
e
co
n
to
u
r
t
h
at
r
ep
r
esen
ted
in
E
q
u
atio
n
6
[
7
].
d
x
d
y
y
x
H
d
x
d
y
y
x
H
y
x
I
c
,
,
,
1
(
5
)
d
x
d
y
y
x
H
d
x
d
y
y
x
H
y
x
I
c
,
1
,
1
,
2
(
6
)
B
y
in
co
r
p
o
r
atin
g
th
e
le
n
g
t
h
a
n
d
ar
ea
en
er
g
y
ter
m
s
i
n
to
E
q
u
atio
n
4
an
d
m
i
n
i
m
izi
n
g
th
e
m
,
t
h
e
co
r
r
esp
o
n
d
in
g
v
ar
iatio
n
al
le
v
e
l se
t f
o
r
m
u
lat
i
o
n
is
as
f
o
llo
w
s
[
7
]
:
2
2
2
2
1
1
,
,
c
y
x
I
c
y
x
I
v
t
(
7
)
d
x
d
y
y
x
H
v
d
x
d
y
y
x
y
x
c
c
F
,
,
,
,
,
2
1
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
C
o
mp
a
r
is
o
n
o
f
Th
r
ee
S
eg
men
t
a
tio
n
Meth
o
d
s
fo
r
B
r
ea
s
t U
ltr
a
s
o
u
n
d
I
ma
g
es B
a
s
ed
o
n
…
(
Dewi P
.
L
.
)
386
2
.
1
.
2
.
Select
iv
e
B
ina
r
y
a
nd
G
a
us
s
ia
n F
ilte
ring
Reg
ula
ri
ze
d L
ev
el
Set
M
o
del
Selectiv
e
B
i
n
ar
y
a
n
d
Gau
s
s
i
an
Fil
ter
i
n
g
R
e
g
u
lar
ized
L
e
v
el
Set
(
SB
GFR
L
S)
m
o
d
el
u
t
ilizes
t
h
e
ad
v
an
ta
g
es
o
f
b
o
th
ed
g
e
-
b
a
s
ed
ac
tiv
e
co
n
to
u
r
m
o
d
els
an
d
r
eg
io
n
-
b
ased
ac
ti
v
e
co
n
to
u
r
m
o
d
els
[
14
]
.
E
d
g
e
-
b
ased
m
o
d
el
s
u
s
e
i
m
ag
e
g
r
ad
ien
t
s
to
co
n
s
tr
u
ct
a
n
E
d
g
e
Sto
p
p
in
g
Fu
n
ctio
n
(
E
SF
)
to
s
to
p
th
e
co
n
to
u
r
ev
o
lu
t
io
n
o
n
t
h
e
o
b
j
ec
t
b
o
u
n
d
ar
ies
[
7
]
.
I
n
SB
GFR
L
S
m
o
d
e
l
h
as
b
ee
n
d
ev
elo
p
ed
a
r
eg
io
n
b
ased
Si
n
g
ed
P
r
ess
u
r
e
Fo
r
ce
(
SP
F)
f
u
n
c
tio
n
to
s
u
b
s
tit
u
te
t
h
e
E
SF
,
w
h
ic
h
is
ab
le
to
co
n
tr
o
l
th
e
d
ir
ec
ti
o
n
o
f
ev
o
l
u
tio
n
[
7
]
.
T
h
e
SP
F
f
u
n
ct
io
n
h
a
s
o
p
p
o
s
ite
s
ig
n
s
(
r
an
g
e
o
f
SP
F
f
u
n
ctio
n
is
[
-
1
1
]
)
ar
o
u
n
d
th
e
b
o
u
n
d
ar
ies
o
f
th
e
o
b
j
e
ct,
s
o
th
e
co
n
to
u
r
ca
n
ex
p
a
n
d
w
h
e
n
it
is
in
s
id
e
t
h
e
b
o
u
n
d
ar
y
an
d
ca
n
s
h
r
i
n
k
s
w
h
e
n
it
i
s
o
u
t
s
id
e
th
e
b
o
u
n
d
ar
y
[
7
]
.
T
h
e
SP
F f
u
n
c
tio
n
i
n
th
i
s
m
o
d
e
l is d
ef
i
n
ed
b
y
t
h
e
f
o
r
m
u
la
[
14
]
:
2
2
2
1
2
,
m
a
x
,
,
1
c
c
y
x
I
c
c
y
x
I
y
x
I
s
p
f
(
8
)
w
it
h
c
1
a
n
d
c
2
ar
e
ex
p
r
ess
ed
in
E
q
u
atio
n
5
an
d
E
q
u
atio
n
6
,
r
esp
ec
tiv
el
y
.
T
h
en
b
y
s
u
b
s
tit
u
tin
g
SP
F
f
u
n
ctio
n
i
n
to
t
h
e
E
q
u
at
io
n
5
,
w
e
o
b
tain
ed
th
e
lev
e
l
s
e
t
f
o
r
m
u
l
atio
n
t
h
at
d
ef
in
ed
as
[
14
]
:
y
x
I
s
pf
di
v
y
x
I
s
pf
t
,
,
(
9
)
T
h
e
r
eg
u
lar
ter
m
div
is
u
n
n
ec
e
s
s
ar
y
b
ec
a
u
s
e
Ga
u
s
s
ian
f
il
ter
ca
n
b
e
u
tili
ze
d
to
s
m
o
o
th
t
h
e
le
v
el
s
et
f
u
n
ctio
n
to
k
ee
p
th
e
in
ter
f
ac
e
r
eg
u
lar
.
B
esid
e
th
at,
th
e
ter
m
s
p
f
ca
n
b
e
r
em
o
v
ed
b
ec
au
s
e
th
e
m
et
h
o
d
u
tili
ze
s
th
e
s
tati
s
tical
i
n
f
o
r
m
at
io
n
o
f
th
e
r
e
g
io
n
s
.
T
h
u
s
t
h
e
f
i
n
al
lev
e
l
s
et
m
o
d
el
i
s
g
i
v
e
n
b
y
th
e
f
o
r
m
u
la
[
14
]
:
(
1
0
)
2.
1.
3
.
Dis
t
a
nce
Reg
ula
rize
d L
ev
el
Set
E
v
o
lutio
n M
o
del
T
h
e
p
r
o
b
lem
o
f
tr
ad
itio
n
al
lev
el
s
et
m
et
h
o
d
is
h
ig
h
co
m
p
u
tatio
n
al
co
s
t
w
h
ic
h
ca
u
s
ed
u
s
e
r
ein
itializat
io
n
.
L
i
et
al.
p
r
o
p
o
s
ed
a
lev
el
s
et
m
et
h
o
d
w
it
h
o
u
t
r
ein
itializatio
n
b
y
in
teg
r
ati
n
g
a
p
en
alt
y
ter
m
in
to
th
e
e
n
er
g
y
f
u
n
ct
io
n
al
[
2
]
.
Mo
r
e
r
ec
en
tl
y
,
b
y
ad
d
in
g
a
d
is
ta
n
ce
r
eg
u
lar
izatio
n
ter
m
in
to
e
n
er
g
y
f
u
n
c
tio
n
al,
L
i
et
al.
p
r
esen
ted
a
n
o
v
el
d
is
ta
n
ce
r
eg
u
lar
ized
le
v
el
s
e
t
ev
o
l
u
t
io
n
m
e
th
o
d
,
th
at
is
,
Di
s
ta
n
ce
R
eg
u
lar
ized
L
ev
el
Set
E
v
o
l
u
tio
n
(
DR
L
SE)
m
o
d
e
l
[
2
]
.
T
h
is
m
o
d
el
ca
n
n
o
t
o
n
l
y
co
m
p
lete
l
y
e
li
m
in
a
te
t
h
e
n
ee
d
f
o
r
r
ein
it
ializatio
n
b
u
t a
ls
o
av
o
id
th
e
u
n
d
es
ir
ab
le
s
id
e
ef
f
ec
t [
2
].
L
et
Ω
⊂
b
e
a
b
o
u
n
d
ed
L
ip
s
ch
itz
i
m
a
g
e
d
o
m
ain
.
T
h
e
en
e
r
g
y
f
u
n
ct
io
n
al
(
⋅
)
is
d
ef
in
ed
b
y
th
e
f
o
llo
w
in
g
f
u
n
ctio
n
[
2
]
:
e
x
t
p
E
R
E
(
1
1
)
with
p
R
is
th
e d
is
tan
ce re
g
u
lar
izatio
n
ter
m
d
ef
in
ed
i
n
E
q
u
atio
n
(
1
2
)
an
d
e
x
t
E
is
th
e ex
ter
n
al en
er
g
y
is
w
r
itten
in
E
q
u
atio
n
(
1
3
)
[
2
].
dx
p
R
p
(
1
2
)
dx
gH
dx
g
E
ext
(
1
3
)
p
is
th
e
p
o
te
n
tial
f
u
n
ctio
n
f
o
r
th
at
d
ef
in
e
d
in
E
q
u
ati
o
n
14
an
d
g
is
an
E
S
F
s
u
ch
th
at
0
lim
t
g
t
,
th
at
d
ef
in
e
d
in
E
q
u
atio
n
15
[
2
].
y
x
u
s
p
f
t
,
0
p
R
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
7
:
3
8
3
–
391
387
1
,
1
2
1
1
,
2
c
o
s
1
2
1
2
2
s
s
s
s
s
p
(
1
4
)
2
1
1
I
G
g
(
1
5
)
w
h
er
e
G
is
a
Ga
u
s
s
ia
n
f
u
n
ct
io
n
w
it
h
t
h
e
s
ta
n
d
ar
d
d
ev
iatio
n
,
I
is
a
g
i
v
en
2
D
i
m
a
g
e,
s
y
m
b
o
l
*
r
ep
r
esen
t
co
n
v
o
lu
tio
n
,
an
d
|
⋅
|
is
th
e
m
o
d
u
l
u
s
o
f
t
h
e
s
m
o
o
th
ed
i
m
ag
e
g
r
a
d
ien
ts
[
2
].
T
h
e
ass
o
ciate
d
E
u
ler
-
L
a
g
r
an
g
e
eq
u
at
io
n
,
w
e
o
b
tai
n
ed
b
y
m
i
n
i
m
izi
n
g
f
u
n
ctio
n
E
q
u
ati
o
n
10
w
it
h
r
esp
ec
t to
is
d
ef
i
n
ed
as f
o
llo
ws
[
2
]
:
(
1
6
)
wh
er
e
(
⋅
)
is
th
e d
iv
er
g
en
ce o
p
er
ato
r
an
d
p
d
is
a f
u
n
ctio
n
g
iv
en
b
y
th
e f
o
r
m
u
la
[
2
]
:
s
s
p
s
d
p
(
1
7
)
2
.
1
.
4
.
M
o
rpho
lo
g
ica
l O
pera
t
io
ns
Mo
r
p
h
o
lo
g
ical
o
p
er
atio
n
d
escr
ib
es a
r
an
g
e
o
f
i
m
a
g
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
e
th
at
d
ea
l
w
i
th
t
h
e
s
h
ap
e
o
f
f
ea
t
u
r
es
in
an
i
m
a
g
e
[
4
]
.
M
o
r
p
h
o
lo
g
ical
o
p
er
atio
n
s
ar
e
a
p
p
lied
to
r
em
o
v
e
i
m
p
er
f
ec
tio
n
s
t
h
at
in
tr
o
d
u
ce
d
d
u
r
in
g
s
eg
m
e
n
tatio
n
.
T
w
o
b
as
ic
m
o
r
p
h
o
lo
g
ical
o
p
er
atio
n
s
ar
e
d
ilatio
n
an
d
er
o
s
io
n
.
T
h
e
y
ar
e
d
ef
in
ed
in
ter
m
s
o
f
m
o
r
e
ele
m
e
n
tar
y
s
et
o
p
er
atio
n
s
,
b
u
t a
r
e
e
m
p
lo
y
ed
as t
h
e
b
asic
ele
m
e
n
t
s
o
f
m
an
y
al
g
o
r
ith
m
s
[
1
5
].
Dilatio
n
an
d
er
o
s
io
n
ar
e
p
r
o
d
u
c
ed
b
y
th
e
in
ter
actio
n
o
f
a
s
et
ca
lled
a
s
tr
u
ctu
r
in
g
elem
en
t
(
SE)
with
a
s
et
o
f
p
ix
els
o
f
in
ter
est
in
th
e
im
ag
e
[
1
5
].
L
et
b
e
a
s
et
o
f
p
ix
els
an
d
let
a
SE,
th
en
th
e
m
o
r
p
h
o
l
o
g
ical
d
ilatio
n
o
f
im
ag
e
b
y
th
e
s
t
r
u
ctu
r
in
g
elem
en
t
is
d
ef
in
ed
in
E
q
u
atio
n
(
1
8
)
an
d
th
e
m
o
r
p
h
o
l
o
g
ical
er
o
s
io
n
o
f
im
ag
e
b
y
th
e
s
tr
u
ctu
r
in
g
elem
en
t
B
is
d
e
f
in
ed
in
E
q
u
atio
n
(
1
9
)
.
T
h
e
m
o
r
p
h
o
lo
g
ical
d
ilatio
n
w
ill
ex
p
an
d
t
h
e
co
m
p
o
n
en
ts
o
f
an
im
ag
e b
u
t
an
d
th
e m
o
r
p
h
o
lo
g
ical
er
o
s
io
n
wil
l
s
h
r
in
k
th
em
[
4
]
.
(
1
8
)
(
1
9
)
Fu
r
t
h
er
m
o
r
e,
d
ilatio
n
an
d
er
o
s
io
n
o
p
er
atio
n
s
ca
n
b
e
co
m
b
i
n
ed
w
it
h
ea
ch
o
th
er
.
T
h
e
co
m
b
in
atio
n
o
f
m
o
r
p
h
o
lo
g
ical
o
p
er
atio
n
s
th
at
o
f
ten
u
s
ed
ar
e
o
p
en
in
g
an
d
cl
o
s
in
g
.
T
h
e
o
p
en
in
g
o
p
er
atio
n
s
o
f
im
a
g
e
b
y
SE
,
is
d
en
o
ted
,
is
d
ef
in
ed
in
E
q
u
atio
n
20
.
W
h
ich
s
a
y
s
th
e
o
p
en
in
g
o
f
i
m
ag
e
b
y
SE
is
th
e
er
o
s
i
o
n
o
f
i
m
a
g
e
b
y
SE
,
f
o
llo
w
ed
b
y
a
d
ilatio
n
o
f
th
e
r
es
u
lt
b
y
S
E
[1
2
]
.
T
h
e
o
p
en
in
g
o
p
er
atio
n
s
g
en
er
al
l
y
s
m
o
o
th
e
s
t
h
e
co
n
to
u
r
o
f
a
n
o
b
j
ec
t,
b
r
ea
k
s
n
ar
r
o
w
i
s
t
h
m
u
te
s
,
an
d
eli
m
i
n
ate
s
th
i
n
p
r
o
tr
u
s
io
n
s
[
4
]
.
T
h
e
clo
s
in
g
o
p
er
atio
n
s
o
f
im
a
g
e
b
y
SE
,
d
en
o
ted
,
is
d
ef
in
ed
in
E
q
u
atio
n
21
.
W
h
ich
s
a
y
s
th
e
clo
s
i
n
g
o
f
i
m
a
g
e
b
y
SE
is
s
i
m
p
l
y
t
h
e
d
ilatio
n
o
f
i
m
ag
e
b
y
SE
,
f
o
llo
w
ed
b
y
th
e
er
o
s
io
n
o
f
t
h
e
r
esu
lt
b
y
SE
[1
2
]
.
T
h
e
clo
s
in
g
o
p
er
atio
n
s
also
ten
d
s
to
s
m
o
o
t
h
s
ec
tio
n
s
o
f
co
n
to
u
r
s
b
u
t,
as
o
p
p
o
s
ed
t
o
o
p
en
in
g
,
i
t
g
e
n
er
all
y
f
u
s
e
s
n
a
r
r
o
w
b
r
ea
k
s
an
d
lo
n
g
t
h
i
n
g
u
lf
s
,
eli
m
i
n
ate
s
s
m
all
h
o
les,
an
d
f
ills
g
ap
s
i
n
th
e
co
n
to
u
r
[
4
]
.
B
B
A
B
A
(
2
0
)
(
2
1
)
2
.
2
.
Va
lid
a
t
i
o
n
M
et
ho
ds
T
h
e
p
r
o
p
o
s
ed
m
e
th
o
d
s
i
n
th
i
s
p
ap
er
h
av
e
b
ee
n
e
v
alu
a
ted
u
s
in
g
ar
ea
-
b
ased
s
eg
m
e
n
tatio
n
a
s
s
es
s
m
en
t
m
etr
ics.
A
r
ea
-
b
ased
s
eg
m
e
n
t
atio
n
ass
e
s
s
m
e
n
t
m
e
tr
ics
m
e
asu
r
e
th
e
a
m
o
u
n
t
o
f
ar
ea
o
v
er
lap
b
etw
ee
n
t
h
e
g
g
d
i
v
d
d
i
v
t
p
y
x
A
,
t
s
B
,
A
B
A
t
s
B
t
y
s
x
A
t
s
B
y
x
A
,
,
m
a
x
,
,
t
s
B
t
y
s
x
A
t
s
B
y
x
A
,
,
m
i
n
,
,
A
B
B
A
A
B
A
B
B
A
B
B
A
A
B
A
B
B
B
B
A
B
A
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
C
o
mp
a
r
is
o
n
o
f
Th
r
ee
S
eg
men
t
a
tio
n
Meth
o
d
s
fo
r
B
r
ea
s
t U
ltr
a
s
o
u
n
d
I
ma
g
es B
a
s
ed
o
n
…
(
Dewi P
.
L
.
)
388
o
b
tain
ed
s
e
g
m
e
n
tatio
n
a
n
d
t
h
e
Gr
o
u
n
d
T
r
u
th
.
I
n
t
h
e
il
lu
s
t
r
ate
d
Fig
u
r
e
2
,
let
A
r
ep
r
ese
n
ted
t
h
e
s
eg
m
e
n
ted
r
eg
io
n
to
b
e
ass
e
s
s
ed
a
n
d
M
d
escr
ib
e
th
e
m
a
n
u
a
ll
y
d
elin
ea
ted
o
r
GT
r
ef
er
en
ce
r
eg
io
n
to
b
e
ass
e
s
s
ed
,
it
w
il
l
f
o
r
m
4
r
eg
io
n
s
:
T
r
u
e
-
P
o
s
itiv
e
(
T
P
)
,
T
r
u
e
Neg
ativ
e
(
T
N)
,
Fals
e
-
P
o
s
iti
v
e
(
FP
)
,
an
d
Fal
s
e
Neg
ati
v
e
(
FN)
.
T
h
e
T
P
ar
ea
co
r
r
esp
o
n
d
s
to
th
e
co
r
r
ec
tly
s
e
g
m
e
n
ted
ar
ea
s
b
elo
n
g
in
g
to
t
h
e
le
s
io
n
,
t
h
e
T
N
ar
ea
co
r
r
esp
o
n
d
s
to
th
e
co
r
r
ec
tly
s
e
g
m
en
ted
ar
ea
s
b
elo
n
g
i
n
g
to
th
e
b
ac
k
g
r
o
u
n
d
o
f
th
e
i
m
a
g
e,
t
h
e
FP
ar
ea
co
r
r
esp
o
n
d
s
to
th
e
ar
ea
s
w
r
o
n
g
l
y
lab
eled
as
a
lesi
o
n
s
in
ce
t
h
is
a
r
ea
s
d
o
n
’
t
b
elo
n
g
to
th
e
r
ef
er
en
ce
d
elin
ea
tio
n
,
an
d
t
h
e
FN
ar
ea
co
r
r
esp
o
n
d
s
to
th
e
ar
ea
s
o
f
th
e
tr
u
e
s
e
g
m
e
n
tat
io
n
th
a
t
h
av
e
b
ee
n
m
is
s
ed
b
y
t
h
e
s
eg
m
e
n
tatio
n
u
n
d
er
ass
es
s
m
en
t [
4
]
.
Fig
u
r
e
2
.
A
r
ea
r
ep
r
esen
tatio
n
o
f
th
e
p
er
f
o
r
m
a
n
ce
i
n
ter
m
s
o
f
ar
ea
-
b
ased
s
eg
m
en
tatio
n
a
s
s
e
s
s
m
en
t
m
etr
ic
s
[
4
]
Dice
Si
m
i
lar
it
y
C
o
ef
f
icie
n
t
(
DSC
)
is
a
co
m
m
o
n
s
i
m
i
lar
it
y
m
etr
ic
f
o
r
r
ep
r
esen
ti
n
g
t
h
e
p
er
ce
n
tag
e
o
r
a
m
o
u
n
t
o
f
ar
ea
co
m
m
o
n
to
th
e
ass
e
s
s
ed
d
elin
ea
tio
n
A
an
d
th
e
r
ef
er
en
ce
d
eli
n
ea
tio
n
M
.
I
ts
v
al
u
e
r
an
g
e
b
et
w
ee
n
0
(
n
o
o
v
er
lap
)
an
d
1
(
p
er
f
ec
t a
g
r
ee
m
e
n
t)
[
4
]
.
T
h
e
D
SC
m
etr
ic
is
e
x
p
r
ess
ed
b
y
f
o
r
m
u
la:
FN
FP
TP
TP
M
A
M
A
D
S
C
.
2
.
2
2
(
2
2
)
w
h
er
e
is
t
h
e
in
ter
s
ec
tio
n
o
p
er
ato
r
an
d
|
⋅
|
r
ep
r
esen
ts
th
e
n
u
m
b
e
r
o
f
p
ix
els i
n
th
e
co
r
r
esp
o
n
d
in
g
p
ix
el
s
e
t.
T
r
u
e
-
P
o
s
itiv
e
R
a
tio
(
T
P
R
)
als
o
k
n
o
w
n
as
s
en
s
iti
v
it
y
,
co
r
r
es
p
o
n
d
s
to
th
e
a
m
o
u
n
t
o
f
p
r
o
p
er
l
y
lab
eled
p
ix
els
a
s
le
s
io
n
w
i
th
r
esp
ec
t
to
th
e
a
m
o
u
n
t
o
f
les
io
n
p
i
x
el
s
f
r
o
m
t
h
e
r
ef
er
e
n
ce
d
eli
n
ea
t
io
n
[
4
]
.
T
h
e
T
P
R
m
etr
ics i
s
ex
p
r
ess
ed
b
y
f
o
r
m
u
l
a:
(
2
3
)
T
r
u
e
-
Ne
g
ati
v
e
R
at
io
(
T
NR
)
also
k
n
o
w
n
a
s
s
p
ec
if
icit
y
,
co
r
r
esp
o
n
d
s
to
th
e
a
m
o
u
n
t
o
f
b
ac
k
g
r
o
u
n
d
co
r
r
ec
tly
lab
eled
.
T
h
e
T
NR
m
etr
ics
is
ex
p
r
es
s
ed
by
f
o
r
m
u
la:
(
2
4
)
A
cc
u
r
ac
y
(
AC
C
)
q
u
a
n
ti
f
ie
s
th
e
a
m
o
u
n
t
o
f
p
r
o
p
er
l
y
lab
eled
p
ix
els
as
le
s
io
n
w
i
th
r
e
s
p
ec
t
to
th
e
a
m
o
u
n
t p
ix
el
s
b
o
th
o
f
as
s
es
s
e
d
d
elin
ea
tio
n
an
d
d
elin
ea
tio
n
.
T
h
e
A
C
C
m
e
tr
ics
is
e
x
p
r
ess
ed
b
y
f
o
r
m
u
la:
(
2
5
)
3.
RE
SU
L
T
S
A
ND
AN
AL
Y
SI
S
T
h
is
s
ec
tio
n
d
escr
ib
es
t
h
e
r
e
s
u
lt
s
o
f
t
h
e
i
m
p
le
m
en
ta
tio
n
o
f
th
e
m
et
h
o
d
p
r
o
p
o
s
ed
.
T
h
e
r
esu
lt
s
o
f
i
m
p
le
m
en
ta
tio
n
co
n
s
is
t
o
f
t
h
e
s
e
g
m
en
ta
tio
n
r
es
u
lt
s
u
s
i
n
g
t
h
e
lev
e
l
s
et
al
g
o
r
ith
m
t
h
at
co
m
b
in
ed
w
it
h
m
o
r
p
h
o
lo
g
ical
o
p
er
atio
n
s
a
n
d
th
e
v
al
id
atio
n
r
es
u
lt
s
u
s
i
n
g
ar
ea
-
b
ased
s
eg
m
e
n
tatio
n
ass
e
s
s
m
en
t
m
e
tr
ics.
FP
TP
TP
M
M
A
T
P
R
FP
TN
TN
M
M
A
T
N
R
A
M
FN
FP
TN
TP
TN
TP
A
CC
TP
FP
FN
TN
A
M
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
7
:
3
8
3
–
391
389
3
.
1
.
S
eg
m
ent
a
t
io
n
T
h
e
test
im
a
g
e
u
s
ed
in
th
i
s
r
esear
ch
co
m
es
f
r
o
m
o
u
r
b
r
ea
s
t
u
ltra
s
o
u
n
d
i
m
a
g
e
d
atab
as
e
an
d
th
e
lesi
o
n
’
s
co
n
to
u
r
is
m
a
n
u
al
l
y
d
elin
ea
ted
b
y
r
ad
io
lo
g
is
t
s
.
T
h
is
d
elin
ea
ted
ar
ea
,
d
en
o
m
i
n
ated
Gr
o
u
n
d
T
r
u
th
(
GT
)
th
at
u
s
ed
o
n
v
a
lid
atio
n
p
r
o
ce
s
s
to
ev
al
u
ate
t
h
e
ac
c
u
r
a
c
y
o
f
s
e
g
m
en
ta
tio
n
m
et
h
o
d
s
p
r
o
p
o
s
ed
.
T
w
o
o
f
3
0
b
r
ea
s
t u
ltra
s
o
u
n
d
i
m
a
g
e
s
u
s
ed
in
o
u
r
r
esear
ch
ca
n
b
e
s
ee
n
i
n
T
ab
le
1
.
I
n
T
ab
le
1
it
ca
n
b
e
s
ee
n
t
wo
ex
a
m
p
les
o
f
u
ltra
s
o
u
n
d
i
m
ag
es
t
h
at
i
n
d
icate
d
th
er
e
ar
e
th
e
b
r
ea
s
t
tu
m
o
r
.
O
n
t
h
e
le
f
t
co
l
u
m
n
co
n
tain
t
h
e
u
ltra
s
o
u
n
d
i
m
a
g
es
w
it
h
g
r
a
y
-
s
ca
le
i
m
a
g
e
t
y
p
e,
w
h
er
e
th
e
b
u
b
b
le
b
lack
is
a
b
r
ea
s
t
tu
m
o
r
.
T
h
e
i
m
ag
e
s
o
n
th
e
r
ig
h
t
co
lu
m
n
ar
e
th
e
r
esu
lt
s
o
f
t
h
e
b
r
ea
s
t
tu
m
o
r
s
e
g
m
en
tatio
n
o
f
th
e
le
f
t
i
m
a
g
es
w
h
o
d
o
n
e
m
a
n
u
a
ll
y
b
y
th
e
r
ad
io
lo
g
is
t
s
.
T
h
e
r
esu
lts
o
f
th
is
m
a
n
u
a
l
s
e
g
m
en
tat
io
n
ar
e
u
s
ed
as
a
co
m
p
ar
is
o
n
to
th
e
r
es
u
lt
s
o
f
th
e
s
eg
m
en
tatio
n
o
f
m
et
h
o
d
s
p
r
o
p
o
s
ed
in
th
is
s
t
u
d
y
.
T
h
e
s
eg
m
en
tat
io
n
p
r
o
ce
s
s
i
n
t
h
is
r
esear
c
h
u
s
i
n
g
th
r
ee
m
et
h
o
d
s
b
ased
o
n
lev
el
s
e
t
th
a
t
ea
ch
m
eth
o
d
co
m
b
i
n
ed
w
i
th
m
o
r
p
h
o
lo
g
ica
l
o
p
er
atio
n
s
.
T
h
r
ee
lev
e
l
s
et
m
o
d
el
s
t
h
at
co
m
p
ar
ed
in
th
is
p
ap
er
ar
e
th
e
C
-
V
m
o
d
el,
t
h
e
SB
GFR
L
S
m
o
d
el
an
d
th
e
DR
L
SE
m
o
d
el.
T
h
en
t
h
e
u
s
e
o
f
m
o
r
p
h
o
lo
g
ical
o
p
er
atio
n
s
i
n
th
e
th
r
ee
lev
el
s
et
m
o
d
els ai
m
ed
to
o
b
tain
m
o
r
e
ac
cu
r
ate
s
eg
m
e
n
t
atio
n
r
esu
lts
.
T
h
e
ex
a
m
p
le
s
o
f
s
e
g
m
en
tatio
n
r
es
u
lt
s
of
th
e
th
r
ee
m
et
h
o
d
s
th
at
p
r
o
p
o
s
ed
in
th
i
s
p
ap
er
ca
n
b
e
s
ee
n
in
T
ab
le
2
.
T
ab
le
1
.
T
w
o
s
a
m
p
les o
f
B
r
ea
s
t U
ltra
s
o
u
n
d
I
m
a
g
es
i
n
Da
ta
s
et
O
r
i
g
i
n
a
l
i
mag
e
G
r
o
u
n
d
t
r
u
t
h
T
a
b
le
2.
Seg
m
en
ta
tio
n
R
e
s
u
lt
s
C
-
V
mo
d
e
l
+
mo
r
p
h
o
l
o
g
i
c
a
l
o
p
e
r
a
t
i
o
n
s
S
B
G
F
R
L
S
mo
d
e
l
+
mo
r
p
h
o
l
o
g
i
c
a
l
o
p
e
r
a
t
i
o
n
s
D
R
L
S
E
mo
d
e
l
+
mo
r
p
h
o
l
o
g
i
c
a
l
o
p
e
r
a
t
i
o
n
s
W
e
ca
n
s
ee
i
n
T
ab
le
2
th
at
t
h
e
r
esu
lt
s
o
f
t
h
is
s
eg
m
e
n
tatio
n
p
r
o
ce
s
s
in
t
h
e
f
o
r
m
o
f
b
in
ar
y
i
m
ag
e
th
at
d
iv
id
ed
in
to
t
w
o
,
n
a
m
el
y
a
w
h
ite
ar
ea
is
th
e
b
r
ea
s
t
lesi
o
n
a
n
d
b
lack
ar
ea
is
th
e
b
ac
k
g
r
o
u
n
d
.
T
h
is
b
in
ar
y
i
m
a
g
e
is
ca
lled
t
h
e
R
e
g
io
n
o
f
I
n
ter
est
(
R
OI
)
i
m
ag
e
.
I
n
T
ab
le
2
it
also
ca
n
b
e
s
ee
n
th
at
th
e
r
esu
lt
s
o
f
th
e
th
r
e
e
m
et
h
o
d
s
p
r
o
p
o
s
ed
ar
e
s
ig
n
i
f
ic
an
tl
y
q
u
ite
d
if
f
er
en
t
.
T
h
e
n
o
n
th
e
i
m
a
g
es
o
f
th
e
s
e
g
m
e
n
tat
io
n
r
es
u
lts
w
er
e
d
o
n
e
th
e
v
al
id
atio
n
p
r
o
ce
s
s
to
s
ee
wh
ich
m
et
h
o
d
is
m
o
r
e
ac
cu
r
a
te.
3
.
2
.
Va
lid
a
t
i
o
n
Fu
r
th
er
m
o
r
e,
th
e
r
esu
lts
o
f
t
h
e
s
eg
m
e
n
tatio
n
p
r
o
ce
s
s
n
ee
d
s
to
b
e
p
er
f
o
r
m
ed
a
v
alid
atio
n
p
r
o
ce
s
s
to
en
s
u
r
e
th
e
ac
cu
r
ac
y
o
f
t
h
e
s
e
g
m
e
n
tat
io
n
m
et
h
o
d
p
r
o
p
o
s
ed
.
T
h
e
v
alid
atio
n
p
r
o
ce
s
s
is
d
o
n
e
b
y
co
m
p
ar
in
g
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
C
o
mp
a
r
is
o
n
o
f
Th
r
ee
S
eg
men
t
a
tio
n
Meth
o
d
s
fo
r
B
r
ea
s
t U
ltr
a
s
o
u
n
d
I
ma
g
es B
a
s
ed
o
n
…
(
Dewi P
.
L
.
)
390
R
OI
i
m
a
g
e
w
ith
t
h
e
GT
im
a
g
e
b
y
u
s
i
n
g
ar
ea
-
b
ased
s
eg
m
en
tatio
n
as
s
es
s
m
e
n
t
m
etr
ic
s
.
I
n
th
is
v
a
lid
atio
n
p
r
o
ce
s
s
h
as
b
ee
n
g
e
n
er
ated
th
e
v
al
u
e
o
f
s
i
m
i
lar
it
y
b
et
w
ee
n
th
e
R
OI
i
m
a
g
e
w
it
h
th
e
GT
im
ag
e.
T
h
e
s
u
m
m
ar
y
r
esu
lt
s
o
f
t
h
e
v
alid
atio
n
v
al
u
e
f
r
o
m
t
h
e
3
0
test
i
m
a
g
es i
n
t
h
is
r
esear
ch
ca
n
b
e
s
ee
n
i
n
T
ab
le
3
.
B
ased
o
n
th
e
r
esu
lts
o
b
tai
n
ed
i
n
T
ab
le
3
,
w
e
ca
n
s
u
m
m
ar
ize
t
h
at
t
h
e
co
m
b
i
n
atio
n
o
f
th
e
C
-
V
m
o
d
el
an
d
m
o
r
p
h
o
lo
g
ical
o
p
er
atio
n
s
h
as
a
v
er
y
g
o
o
d
p
er
f
o
r
m
a
n
ce
in
ter
m
o
f
s
e
g
m
en
tatio
n
b
r
ea
s
t
lesi
o
n
w
it
h
Dice
Si
m
i
lar
it
y
C
o
e
f
f
icien
t
(
D
SC
)
is
9
4
%
an
d
T
r
u
e
-
P
o
s
itiv
e
R
at
io
(
T
P
R
)
is
9
4
%,
g
iv
e
th
e
b
e
s
t
th
a
n
th
e
o
t
h
er
s
.
Ho
w
e
v
er
T
r
u
e
-
Neg
ati
v
e
R
ati
o
(
T
NR
)
is
w
o
r
s
e
t
h
an
D
R
L
SE
m
o
d
el
th
at
co
m
b
i
n
ed
w
it
h
m
o
r
p
h
o
lo
g
ica
l
o
p
er
atio
n
s
an
d
ac
cu
r
ac
y
h
as a
s
a
m
e
w
i
th
D
R
L
SE
m
o
d
el
t
h
at
co
m
b
i
n
ed
w
it
h
m
o
r
p
h
o
lo
g
ica
l
o
p
er
atio
n
s
.
T
ab
le
3
.
E
v
alu
atio
n
Me
tr
ics
M
e
t
h
o
d
s
M
e
a
n
D
S
C
(
%)
M
e
a
n
T
P
R
(
%)
M
e
a
n
T
N
R
(
%)
M
e
a
n
A
C
C
(
%)
C
-
V
mo
d
e
l
+
mo
r
p
h
o
l
o
g
i
c
a
l
o
p
e
r
a
t
i
o
n
s
94
94
99
94
S
B
G
F
R
L
S
mo
d
e
l
+
mo
r
p
h
o
l
o
g
i
c
a
l
o
p
e
r
a
t
i
o
n
s
90
86
99
93
D
R
L
S
E
mo
d
e
l
+
mo
r
p
h
o
l
o
g
i
c
a
l
o
p
e
r
a
t
i
o
n
s
90
87
1
0
0
94
4.
CO
NCLU
SI
O
N
I
n
th
is
p
ap
er
,
w
e
h
a
v
e
p
r
es
en
ted
a
co
m
p
r
eh
e
n
s
iv
e
co
m
p
ar
is
o
n
o
f
th
r
ee
l
e
v
el
s
e
t
m
e
t
h
o
d
s
th
at
c
o
m
b
in
e
d
w
ith
m
o
r
p
h
o
l
o
g
i
c
al
o
p
e
r
a
t
i
o
n
s
f
o
r
s
eg
m
en
t
a
ti
o
n
o
f
b
r
e
a
s
t
u
l
t
r
as
o
u
n
d
im
ag
es
.
T
h
r
e
e
l
e
v
e
l
s
et
m
o
d
el
s
u
s
e
d
in
th
i
s
p
a
p
e
r
a
r
e
th
e
C
h
a
n
V
es
e
(
C
-
V
)
m
o
d
e
l
,
th
e
S
e
le
c
t
iv
e
B
in
a
r
y
an
d
G
au
s
s
i
an
Fi
l
t
e
r
in
g
R
eg
u
l
a
r
i
z
e
d
L
ev
el
S
et
(
S
B
GFR
L
S
)
m
o
d
el
a
n
d
th
e
Di
s
ta
n
c
e
R
eg
u
l
a
r
i
ze
d
L
ev
e
l
Se
t
E
v
o
lu
ti
o
n
(
D
R
L
S
E
)
m
o
d
e
l
.
T
h
e
p
er
f
o
r
m
a
n
ce
ev
al
u
atio
n
o
f
o
u
r
m
et
h
o
d
s
w
er
e
b
ee
n
d
e
m
o
n
s
tr
ated
b
y
th
e
ex
p
er
i
m
e
n
tal
r
e
s
u
lt
s
w
ith
u
s
ed
3
0
b
r
ea
s
t
u
ltra
s
o
u
n
d
i
m
ag
e
s
,
w
h
i
ch
th
e
lesi
o
n
h
ad
b
ee
n
d
eli
n
ea
ted
m
an
u
all
y
b
y
r
ad
io
lo
g
i
s
ts
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
C
-
V
m
o
d
el
th
a
t c
o
m
b
i
n
ed
w
it
h
m
o
r
p
h
o
lo
g
ical
o
p
er
atio
n
s
was si
g
n
if
ican
t
l
y
b
etter
t
h
a
n
t
h
e
o
th
er
t
w
o
m
e
th
o
d
s
.
I
t
w
a
s
in
d
icate
d
f
r
o
m
th
e
av
e
r
ag
e
v
alu
e
o
f
p
r
ec
is
io
n
,
w
h
ic
h
is
9
4
%
f
o
r
Dice
Si
m
ilar
it
y
C
o
ef
f
ic
ien
t
(
DS
C
)
m
etr
ic
t
h
at
m
u
c
h
h
i
g
h
er
th
a
n
th
e
SB
GF
R
L
S
m
o
d
el
t
h
at
co
m
b
in
ed
w
it
h
m
o
r
p
h
o
lo
g
ical
o
p
er
atio
n
s
(
9
0
%)
an
d
th
e
DR
L
SE
m
o
d
el
th
a
t c
o
m
b
i
n
ed
w
ith
m
o
r
p
h
o
lo
g
ical
o
p
er
atio
n
s
(
9
0
%).
RE
F
E
R
E
NC
E
S
[1
]
A
.
Je
m
a
l,
e
t
a
l
.
,
"
G
lo
b
a
l
Ca
n
c
e
r
S
tatisti
c
s"
,
CA:
A
Ca
n
c
e
r
J
o
u
rn
a
l
f
o
r Cli
n
icia
n
s
,
v
o
l.
6
1
,
p
p
.
6
9
-
9
0
,
M
a
r
2
0
1
1
.
[2
]
L
.
Ga
o
,
e
t
a
l
.
,
"
P
h
a
se
-
a
n
d
G
V
F
-
Ba
se
d
L
e
v
e
l
S
e
t
S
e
g
m
e
n
tatio
n
o
f
Ultras
o
n
ic
Bre
a
st
T
u
m
o
rs"
,
J
o
u
rn
a
l
o
f
A
p
p
li
e
d
M
a
th
e
ma
ti
c
s
,
v
o
l.
2
0
1
2
,
2
0
1
2
.
[3
]
H.D.
Ch
e
n
g
,
e
t
a
l
.
,
"
A
u
to
m
a
ted
Bre
a
st
Ca
n
c
e
r
De
te
c
ti
o
n
a
n
d
Clas
sif
ic
a
ti
o
n
Us
in
g
Ultras
o
u
n
d
Im
a
g
e
s:
A
S
u
rv
e
y
"
,
Pa
tt
e
rn
Rec
o
g
n
it
io
n
,
v
o
l
.
4
3
,
p
p
.
2
9
9
-
3
1
7
,
2
0
1
0
.
[4
]
D.P
.
L
e
sta
ri,
e
t
a
l
.
,
"
A
S
e
g
m
e
n
tatio
n
A
lg
o
rit
h
m
f
o
r
Bre
a
st
L
e
sio
n
Ba
se
d
o
n
A
c
ti
v
e
Co
n
to
u
r
M
o
d
e
l
a
n
d
M
o
rp
h
o
l
o
g
ica
l
Op
e
ra
ti
o
n
s"
,
Ad
v
a
n
c
e
d
S
c
ien
c
e
E
n
g
i
n
e
e
rin
g
a
n
d
M
e
d
icin
e
,
v
o
l.
7
,
p
p
.
9
2
0
-
9
2
4
,
Oc
t
2
0
1
5
.
[5
]
H.
Ku
m
a
r,
e
t
a
l
.
,
"
S
e
g
m
e
n
tatio
n
a
n
d
F
e
a
tu
re
Ex
trac
ti
o
n
o
f
Ultras
o
u
n
d
Im
a
g
e
s
b
y
M
o
d
if
ied
L
e
v
e
l
S
e
t
M
e
th
o
d
a
n
d
Ch
a
in
-
V
e
se
M
e
th
o
d
s
Us
in
g
S
R
A
D
F
il
ter"
,
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
Ad
v
a
n
c
e
d
Res
e
a
rc
h
in
Co
mp
u
ter
a
n
d
Co
mm
u
n
ica
ti
o
n
E
n
g
i
n
e
e
rin
g
,
v
o
l.
1
,
p
p
.
1
7
1
-
1
7
7
,
M
a
y
2
0
1
2
.
[6
]
G
.
Zh
u
,
e
t
a
l
.
,
"
Bo
u
n
d
a
ry
-
Ba
s
e
d
Im
a
g
e
S
e
g
m
e
n
tatio
n
Us
in
g
Bin
a
r
y
L
e
v
e
l
S
e
t
M
e
th
o
d
"
,
Op
ti
c
a
l
E
n
g
in
e
e
rin
g
,
v
o
l.
4
6
,
n
o
.
5
,
M
a
y
2
0
0
7
.
[7
]
K.
Zh
a
n
g
,
e
t
a
l
., "
A
c
ti
v
e
Co
n
to
u
r
s
w
it
h
S
e
lec
ti
v
e
L
o
c
a
l
o
r
G
lo
b
a
l
S
e
g
m
e
n
tatio
n
:
A
Ne
w
F
o
r
m
u
latio
n
a
n
d
L
e
v
e
l
S
e
t
M
e
th
o
d
"
,
Ima
g
e
a
n
d
Vi
si
o
n
Co
m
p
u
ti
n
g
,
v
o
l.
2
8
,
p
p
.
6
6
8
-
6
7
6
,
A
p
r
2
0
1
0
.
[8
]
C.
X
u
,
e
t
a
l
.
,
Ha
n
d
b
o
o
k
o
f
M
e
d
ica
l
Ima
g
in
g
Vo
l
u
me
2
:
M
e
d
ica
l
Im
a
g
e
Pro
c
e
ss
in
g
a
n
d
An
a
lys
is
.
Be
ll
in
g
h
a
m
,
US
A
:
S
P
IE
P
re
ss
,
2
0
0
0
.
[9
]
X
.
C
h
a
n
g
,
e
t
a
l
.
,
"
A
S
tu
d
y
o
f
I
m
a
g
e
S
e
g
m
e
n
tatio
n
Ba
se
d
o
n
L
e
v
e
l
S
e
t
M
e
th
o
d
"
,
in
I
n
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Ad
v
a
n
c
e
d
Co
mp
u
ter
S
c
ien
c
e
a
n
d
El
e
c
tro
n
ics
In
f
o
rm
a
ti
o
n
, p
p
.
3
6
0
-
3
6
3
,
2
0
1
3
.
[1
0
]
C.
L
i,
e
t
a
l
.,
"
Dista
n
c
e
Re
g
u
lar
ize
d
L
e
v
e
l
S
e
t
Ev
o
lu
ti
o
n
a
n
d
I
ts
A
p
p
li
c
a
ti
o
n
t
o
Im
a
g
e
S
e
g
m
e
n
tatio
n
"
,
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Ima
g
e
Pro
c
e
ss
in
g
,
v
o
l.
1
9
,
p
p
.
3
2
4
3
-
3
2
5
4
,
De
c
2
0
1
0
.
[1
1
]
T
.
F
.
Ch
a
n
a
n
d
L
.
A
.
V
e
se
,
"
A
c
ti
v
e
Co
n
t
o
u
r
w
it
h
o
u
t
E
d
g
e
s"
,
IEE
E
T
r
a
n
sa
c
ti
o
n
s
o
n
Ima
g
e
Pro
c
e
ss
in
g
,
v
o
l
.
1
0
,
p
p
.
266
-
2
7
7
,
F
e
b
2
0
0
1
.
[1
2
]
R.
C.
G
o
n
z
a
lez
a
n
d
R.
E.
W
o
o
d
s,
Dig
it
a
l
Ima
g
e
Pro
c
e
ss
in
g
.
S
i
n
g
a
p
o
re
:
P
e
a
rso
n
P
re
n
ti
c
e
Ha
ll
,
2
0
0
2
.
[1
3
]
K.O.
Ba
b
a
lo
la,
e
t
a
l
.
,
"
Co
m
p
a
riso
n
a
n
d
Ev
a
lu
a
ti
o
n
o
f
S
e
g
m
e
n
tatio
n
T
e
c
h
n
iq
u
e
s
f
o
r
S
u
b
c
o
rt
ica
l
S
tr
u
c
tu
re
s
in
Bra
in
M
RI"
,
in
M
ICCAI
2
0
0
8
P
a
rt I
L
N
CS
,
v
o
l.
5
2
4
1
,
p
p
.
4
0
9
-
4
1
6
,
2
0
0
8
.
[1
4
]
K.
S
a
in
i,
e
t
a
l
.
,
"
L
e
v
e
l
S
e
t
Ba
se
d
o
n
Ne
w
S
ig
n
e
d
P
re
ss
u
re
F
o
rc
e
F
u
n
c
t
i
o
n
f
o
r
Ech
o
c
a
rd
io
g
r
a
p
h
ic
Im
a
g
e
"
,
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
I
n
n
o
v
a
ti
o
n
a
n
d
A
p
p
li
e
d
S
tu
d
ies
,
v
o
l.
3
,
p
p
.
5
6
0
-
5
6
9
,
Ju
n
e
2
0
1
3
.
[1
5
]
M
.
G
o
y
a
l,
"
M
o
rp
h
o
lo
g
ica
l
Im
a
g
e
P
ro
c
e
ss
in
g
"
,
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
Co
mp
u
ter
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
2
,
p
p
.
1
6
1
-
1
6
5
,
De
c
2
0
1
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
7
:
3
8
3
–
391
391
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
De
w
i
Pu
tr
ie
Le
sta
r
i
re
c
e
i
v
e
d
th
e
B.
S
.
d
e
g
re
e
in
M
a
th
e
m
a
ti
c
s
f
r
o
m
Un
iv
e
rsit
y
o
f
In
d
o
n
e
sia
i
n
2
0
0
8
,
th
e
M
.
S
.
in
M
a
th
e
m
a
ti
c
s
fro
m
Un
iv
e
rsit
y
o
f
In
d
o
n
e
sia
in
2
0
1
2
a
n
d
th
e
P
h
.
D.
d
e
g
re
e
in
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
f
ro
m
U
n
iv
e
rsity
o
f
G
u
n
a
d
a
rm
a
in
2
0
1
5
.
S
h
e
h
a
s
p
u
b
l
ish
e
d
e
x
ten
siv
e
l
y
i
n
t
h
e
a
re
a
o
f
m
e
d
ica
l
ima
g
in
g
.
He
r
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
d
y
n
a
m
ic
m
o
d
e
ls,
sta
ti
stics
,
im
a
g
e
p
ro
c
e
ss
in
g
,
a
n
d
m
e
d
ica
l
i
m
a
g
in
g
.
Em
a
il
:
d
e
w
ip
u
tri
e
2
6
@g
m
a
il
.
c
o
m
Pro
f.
S
a
r
if
u
d
d
i
n
M
a
d
e
n
d
a
-
Cu
rre
n
tl
y
Dire
c
to
r
o
f
Hig
h
S
c
h
o
o
l
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
M
a
n
a
g
e
m
e
n
t
Ja
k
a
rta
(S
T
M
IK
Ja
k
a
rta
S
T
I&
K).
H
e
re
c
e
iv
e
d
th
e
B
.
S
.
d
e
g
re
e
f
ro
m
Un
iv
e
rsit
y
o
f
In
d
o
n
e
sia
,
in
1
9
8
9
,
th
e
M
.
S
.
d
e
g
re
e
f
ro
m
In
stit
u
te
Na
ti
o
n
a
l
d
e
s
S
c
ien
c
e
s
A
p
p
li
q
u
é
e
s
d
e
Ly
o
n
(INSA
d
e
Ly
o
n
)
–
F
re
n
c
h
,
in
1
9
9
2
,
a
n
d
th
e
P
h
.
D.
d
e
g
re
e
f
ro
m
Un
iv
e
rsit
y
o
f
Bu
rg
u
n
d
y
-
F
re
n
c
h
,
in
1
9
9
5
.
F
r
o
m
1
9
9
5
to
1
9
9
6
,
h
e
w
a
s
a
Re
s
e
a
rc
h
A
s
so
c
iate
a
t
L
IE
S
IB
L
a
b
o
ra
to
ry
-
Un
iv
e
r
sit
y
o
f
Bu
rg
u
n
d
y
.
In
1
9
9
7
,
h
e
b
e
c
a
m
e
a
L
e
c
tu
re
r
in
Co
m
p
u
ter
S
c
ie
n
c
e
De
p
a
rtm
e
n
t,
G
u
n
a
d
a
r
m
a
Un
iv
e
rsit
y
–
In
d
o
n
e
sia
.
F
ro
m
2
0
0
2
t
o
2
0
0
7
,
h
e
w
a
s
a
re
se
a
rc
h
e
r
a
t
A
c
a
d
e
m
i
c
Re
se
a
r
c
h
Co
n
so
rt
iu
m
o
n
d
ig
it
a
l
im
a
g
in
g
,
v
i
d
e
o
,
a
u
d
i
o
a
n
d
m
u
lt
im
e
d
ia (Co
RI
M
e
d
ia),
Ca
n
a
d
a
.
His res
e
a
r
c
h
in
tere
sts
a
re
im
a
g
e
p
ro
c
e
ss
in
g
:
im
a
g
e
c
o
m
p
re
s
sio
n
,
c
o
lo
r
im
a
g
in
g
,
im
a
g
e
d
a
tab
a
se
a
n
d
se
a
rc
h
in
g
,
m
e
d
ica
l
i
m
a
g
e
a
n
a
l
y
sis,
im
p
le
m
e
n
ti
n
g
o
f
i
m
a
g
e
p
ro
c
e
ss
in
g
a
lg
o
rit
h
m
s
o
n
F
P
GA
f
o
r
re
a
l
ti
m
e
i
m
a
g
e
a
n
a
l
y
sis.
E
m
a
il
:
sa
rif
@s
ta
ff
.
g
u
n
a
d
a
rm
a
.
a
c
.
id
Dr
.
Er
n
a
st
u
ti
-
h
a
s
re
c
e
iv
e
d
B.
S
.
d
e
g
re
e
in
M
a
th
e
m
a
ti
c
s
f
ro
m
Un
iv
e
rsit
y
o
f
In
d
o
n
e
sia
in
De
c
e
m
b
e
r
1
9
8
5
,
a
n
d
th
e
M
.
S
.
in
Co
m
p
u
ter
S
c
ien
c
e
f
ro
m
Un
iv
e
rsit
y
o
f
In
d
o
n
e
sia
,
in
Ju
ly
1
9
9
4
,
a
n
d
P
h
D
d
e
g
re
e
s
in
Co
m
p
u
ter
S
c
ien
c
e
f
ro
m
G
u
n
a
d
a
rm
a
Un
iv
e
rsi
ty
,
In
d
o
n
e
sia
,
in
A
p
ril
2
0
0
8
.
S
h
e
is
c
u
rre
n
tl
y
a
n
a
ss
o
c
iate
p
ro
f
e
ss
o
r
in
th
e
f
a
c
u
lt
y
o
f
c
o
m
p
u
ter
sc
ien
c
e
a
n
d
In
f
o
rm
a
ti
o
n
En
g
in
e
e
rin
g
,
G
u
n
a
d
a
rm
a
Un
iv
e
r
sity
.
He
r
c
u
rre
n
t
re
se
a
rc
h
in
tere
st
in
c
lu
d
e
g
ra
p
h
th
e
o
ry
a
n
d
c
o
m
b
in
a
to
rial
o
p
t
im
iza
ti
o
n
,
g
ra
p
h
-
th
e
o
re
ti
c
i
n
terc
o
n
n
e
c
ti
o
n
n
e
tw
o
rk
s,
p
a
ra
ll
e
l
a
n
d
d
istri
b
u
ted
c
o
m
p
u
ti
n
g
,
a
n
d
d
e
sig
n
a
n
d
a
n
a
ly
s
is
o
f
a
lg
o
rit
h
m
s.
Ema
il
:
e
rn
a
s_
tu
ti
@
y
a
h
o
o
.
c
o
.
i
d
Er
i
Pra
se
ty
o
W
ib
o
w
o
b
o
rn
in
Ke
n
d
a
l,
I
n
d
o
n
e
sia
,
i
n
1
9
6
6
.
H
e
re
c
e
iv
e
d
th
e
B.
S
d
e
g
re
e
in
El
e
c
tro
n
ics
a
n
d
In
stru
m
e
n
tatio
n
f
ro
m
Ga
d
jah
M
a
d
a
Un
iv
e
rsit
y
,
I
n
d
o
n
e
sia
,
in
1
9
9
2
,
th
e
M
.
S
.
d
e
g
re
e
in
In
f
o
r
m
a
ti
o
n
S
y
st
e
m
f
r
o
m
G
u
n
a
d
a
r
m
a
Un
iv
e
r
sit
y
,
In
d
o
n
e
sia
,
in
1
9
9
4
a
n
d
T
h
e
P
h
.
D.
d
e
g
re
e
in
El
e
c
tro
n
ics
In
f
o
rm
a
ti
c
s
f
ro
m
th
e
Bu
rg
u
n
d
y
Un
iv
e
rsit
y
,
F
ra
n
c
e
,
in
2
0
0
5
.
He
is
m
e
m
b
e
r
o
f
IEE
E,
A
CM
,
P
r
o
f
e
ss
io
n
a
l
Org
a
n
iza
ti
o
n
i
n
t
h
e
f
ield
o
f
In
f
o
r
m
a
ti
o
n
a
n
d
Co
m
p
u
ter
T
e
c
h
n
o
l
o
g
y
(IP
KIN
),
A
P
T
IKO
M
.
He
h
a
s
b
e
e
n
w
o
rk
e
d
a
s
S
e
c
r
e
tar
y
o
f
Do
c
to
ra
l
P
ro
g
ra
m
in
In
f
o
r
m
a
ti
o
n
T
e
c
h
n
o
lo
g
y
a
t
G
u
n
a
d
a
rm
a
Un
iv
e
rsit
y
.
He
w
a
s
m
e
m
b
e
r
o
f
E
A
C
OV
IROE
P
ro
jec
t
to
p
r
o
m
o
ted
M
a
ste
r
V
IBOT
w
h
ich
f
o
u
n
d
e
d
b
y
Eras
m
u
s
M
u
n
d
u
s.
His
c
u
rre
n
t
in
t
e
re
sts
a
re
in
Re
a
l
ti
m
e
I
m
a
g
e
p
ro
c
e
ss
in
g
a
p
p
li
c
a
ti
o
n
s an
d
S
y
ste
m
On
CHIP
s De
sig
n
.
Em
a
il
:
e
ri
@s
taff
.
g
u
n
a
d
a
r
m
a
.
a
c
.
id
Evaluation Warning : The document was created with Spire.PDF for Python.