I
AE
S
I
n
t
e
r
n
at
ion
al
Jou
r
n
al
of
Ar
t
if
icial
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
AI
)
Vol.
14
,
No.
2
,
Apr
il
2025
,
pp.
1183
~
119
1
I
S
S
N:
2252
-
8938
,
DO
I
:
10
.
11591/i
jai
.
v
14
.i
2
.
pp
11
83
-
119
1
1183
Jou
r
n
al
h
omepage
:
ht
tp:
//
ij
ai
.
iaes
c
or
e
.
c
om
Op
t
imiz
in
g t
h
e
gal
ls
t
on
e
d
e
t
e
c
t
io
n
p
r
o
c
e
ss wit
h
f
e
at
u
r
e
s
e
le
c
t
io
n
st
at
is
t
ic
al
an
al
ysi
s al
gor
ith
m
M
u
s
li
Yant
o
1
,
Y
u
h
an
d
r
i
1
,
M
u
h
am
m
ad
T
aj
u
d
d
i
n
2
,
Vin
a
T
r
i
S
e
p
t
ia
n
a
3
1
D
e
pa
r
tm
e
nt
of
I
nf
or
ma
ti
on T
e
c
hnol
ogy, F
a
c
ul
ty
of
C
omput
e
r
S
c
ie
nc
e
,
U
ni
ve
r
s
it
a
s
P
ut
r
a
I
ndone
s
ia
Y
P
T
K
, P
a
da
ng, I
ndone
s
ia
2
D
e
pa
r
tm
e
nt
of
C
omput
e
r
S
c
ie
nc
e
, F
a
c
ul
ty
of
E
ngi
ne
e
r
in
g, U
ni
ve
r
s
it
a
s
B
umi
G
or
a
, M
a
ta
r
a
m, I
ndone
s
ia
3
D
e
pa
r
tm
e
nt
of
R
a
di
ol
ogy, F
a
c
ul
ty
of
M
e
di
c
in
e
,
B
a
it
ur
r
a
hma
h U
ni
ve
r
s
it
y, P
a
da
ng, I
ndone
s
ia
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
M
a
r
16
,
2024
R
e
vis
e
d
Oc
t
31,
2024
Ac
c
e
pted
Nov
14
,
2024
E
arl
y
d
e
t
ect
i
o
n
i
s
o
n
e
fo
rm
o
f
ear
l
y
a
n
t
i
ci
p
at
i
o
n
i
n
t
reat
i
n
g
g
a
l
l
s
t
o
n
e
d
i
s
ea
s
e
p
at
i
en
t
s
u
s
i
n
g
me
d
i
ca
l
i
ma
g
es
.
H
o
w
ev
er,
t
h
e
p
r
o
b
l
em
t
h
at
ex
i
s
t
s
i
s
t
h
at
t
h
ere
are
s
t
i
l
l
man
y
s
h
o
rt
c
o
mi
n
g
s
i
n
med
i
ca
l
i
mag
e
s
,
s
u
ch
as
n
o
i
s
e
i
n
t
h
e
i
mag
e
t
h
at
cau
s
e
s
t
h
e
d
et
ec
t
i
o
n
p
r
o
ces
s
t
o
n
o
t
ru
n
o
p
t
i
mal
l
y
.
Bas
e
d
o
n
t
h
i
s
,
t
h
i
s
s
t
u
d
y
ai
ms
t
o
carry
o
u
t
t
h
e
p
ro
ce
s
s
o
f
d
et
ec
t
i
n
g
g
a
l
l
s
t
o
n
e
o
b
j
ect
s
i
n
mag
n
e
t
i
c
res
o
n
an
ce
c
h
o
l
an
g
i
o
p
a
n
creat
o
g
ra
p
h
y
(MRCP)
i
mag
e
s
b
y
o
p
t
i
m
i
zi
n
g
t
h
e
p
erfo
rma
n
ce
o
f
ex
t
ract
i
o
n
t
ec
h
n
i
q
u
es
f
o
r
feat
u
re
s
el
ect
i
o
n
.
O
p
t
i
m
i
zat
i
o
n
o
f
ex
t
rac
t
i
o
n
t
ec
h
n
i
q
u
es
i
n
feat
u
re
s
e
l
ect
i
o
n
i
s
carri
e
d
o
u
t
u
s
i
n
g
t
h
e
p
erf
o
rman
ce
o
f
t
h
e
fea
t
u
re
s
el
ec
t
i
o
n
s
t
at
i
s
t
i
c
s
an
a
l
y
s
i
s
(FSSA
)
al
g
o
r
i
t
h
m.
T
h
e
p
erf
o
rma
n
ce
o
f
t
h
e
FSSA
a
l
g
o
ri
t
h
m
can
p
ro
v
i
d
e
i
mp
r
o
v
eme
n
t
s
i
n
t
h
e
feat
u
re
s
e
l
ect
i
o
n
p
ro
ce
s
s
b
y
e
x
cel
l
i
n
g
i
n
t
h
e
p
erfo
rma
n
ce
o
f
c
l
as
s
i
f
i
cat
i
o
n
met
h
o
d
s
s
u
c
h
as
k
-
n
eares
t
n
ei
g
h
b
o
r
(K
N
N
),
s
u
p
p
o
rt
v
ec
t
o
r
mach
i
n
e
(SV
M)
,
an
d
art
i
fi
c
i
al
n
e
u
ral
n
et
w
o
r
k
(A
N
N
),
an
d
t
h
e
Pears
o
n
co
rre
l
at
i
o
n
(PC)
met
h
o
d
.
Bas
ed
o
n
t
h
e
t
e
s
t
s
t
h
a
t
h
av
e
b
een
carri
e
d
o
u
t
,
t
h
e
p
erfo
rma
n
ce
o
f
t
h
e
FSSA
al
g
o
r
i
t
h
m
i
n
t
h
e
d
et
ec
t
i
o
n
p
r
o
ces
s
p
ro
v
i
d
es
an
accu
rac
y
l
ev
e
l
o
f
9
5
.
6
9
%
,
a
s
en
s
i
t
i
v
i
t
y
o
f
8
9
.
6
5
%
,
an
d
a
s
p
eci
f
i
ci
t
y
o
f
9
8
.
4
3
%
.
O
v
era
l
l
,
t
h
i
s
s
t
u
d
y
can
co
n
t
ri
b
u
t
e
t
o
t
h
e
d
ev
e
l
o
p
men
t
o
f
e
x
t
rac
t
i
o
n
a
n
d
p
ro
v
i
d
e
a
s
i
g
n
i
f
i
can
t
t
ech
n
i
ca
l
i
m
p
act
o
n
o
p
t
i
mi
zi
n
g
t
h
e
g
a
l
l
s
t
o
n
e
d
e
t
ect
i
o
n
p
r
o
ces
s
.
K
e
y
w
o
r
d
s
:
De
tec
ti
on
FSSA
Ga
ll
s
tones
M
R
C
P
Optim
iza
ti
on
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
:
M
us
li
Ya
nto
D
e
p
a
r
t
men
t
o
f
I
n
f
o
r
mat
ics
E
ng
in
e
e
r
in
g
,
F
a
c
u
l
ty
of
C
om
pu
te
r
S
c
ien
c
e
,
U
n
ive
r
s
it
a
s
P
u
tr
a
I
n
do
ne
s
ia
Y
P
T
K
P
a
da
ng,
25145,
I
ndone
s
ia
E
mail:
mus
li
_ya
nto@upi
yptk
.
a
c
.
id
1.
I
NT
RODU
C
T
I
ON
Ga
ll
s
tones
a
r
e
a
dis
e
a
s
e
that
c
a
n
a
tt
a
c
k
the
bil
e
or
ga
ns
due
to
e
xc
e
s
s
ive
c
holes
ter
ol
c
ontent
[
1]
.
T
he
e
f
f
e
c
ts
o
f
e
xc
e
s
s
c
holes
ter
ol
c
a
n
ha
ve
a
ne
ga
ti
ve
im
pa
c
t
by
f
o
r
mi
ng
s
tone
pa
r
ti
c
les
in
the
ga
ll
b
ladde
r
a
r
e
a
[
2]
.
Not
only
that,
thi
s
dis
e
a
s
e
a
ls
o
c
a
us
e
s
pa
in
f
or
quit
e
a
long
ti
me
s
o
e
a
r
ly
p
r
e
ve
nti
on
ne
e
ds
to
be
done
[
3]
.
P
r
e
vious
r
e
s
e
a
r
c
h
s
tate
d
that
ga
ll
s
tones
a
r
e
a
di
s
e
a
s
e
that
c
a
n
thr
e
a
ten
human
he
a
lt
h
[
4]
.
B
a
s
e
d
on
thi
s
,
a
n
e
a
r
ly
de
tec
ti
on
pr
oc
e
s
s
is
r
e
a
ll
y
ne
e
de
d
by
uti
li
z
ing
de
ve
lopm
e
nts
in
medic
a
l
im
a
ge
tec
hnology
[
5]
.
M
e
dica
l
im
a
ge
s
a
r
e
a
tec
hnology
de
ve
loped
in
the
wor
ld
of
he
a
lt
h
that
p
lays
a
r
o
le
in
s
uppor
ti
ng
the
diagnos
is
pr
oc
e
s
s
[
6]
.
M
e
dica
l
im
a
ge
s
or
know
n
a
s
biom
e
dica
l
pr
oc
e
s
s
ing
tec
hnology
c
a
n
c
ontr
ibut
e
to
de
c
is
ion
making
[
7]
.
T
he
a
c
ti
ve
r
ole
of
medic
a
l
i
mage
s
in
s
e
ve
r
a
l
pr
e
vious
s
tudi
e
s
ha
s
a
ls
o
ha
d
a
s
igni
f
ica
nt
im
pa
c
t
on
p
r
ogr
e
s
s
in
the
he
a
lt
h
s
e
c
tor
[
8]
.
One
f
o
r
m
of
medic
a
l
im
a
ge
c
a
n
be
s
e
e
n
f
r
om
magne
ti
c
r
e
s
ona
nc
e
c
holangiopa
nc
r
e
a
togr
a
phy
(
M
R
C
P
)
.
T
he
im
pleme
ntation
of
M
R
C
P
c
a
n
be
us
e
d
to
a
s
s
is
t
medic
a
l
pe
r
s
onne
l
in
the
pr
oc
e
s
s
of
diagnos
ing
dis
e
a
s
e
[
9]
.
P
r
e
vious
r
e
s
e
a
r
c
h
e
xplains
that
M
R
C
P
im
a
ge
s
ha
ve
be
e
n
wide
l
y
us
e
d
in
the
pos
t
-
ope
r
a
ti
ve
de
tec
ti
on
pr
oc
e
s
s
[
10]
.
Othe
r
r
e
s
e
a
r
c
h
a
ls
o
s
tate
s
that
M
R
C
P
im
a
ge
s
a
r
e
a
tool
in
the
medic
a
l
wor
ld
that
plays
a
n
a
c
ti
ve
r
ole
in
s
e
ve
r
a
l
dis
e
a
s
e
de
tec
ti
on
pr
oc
e
s
s
e
s
[
11]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
Ar
ti
f
I
ntell
,
Vol.
14
,
No.
2
,
Apr
il
20
25
:
1183
-
119
1
1184
M
R
C
P
im
a
ge
s
in
de
tec
ti
on
c
a
s
e
s
ha
ve
a
ls
o
be
e
n
us
e
d
to
de
s
c
r
ibe
the
loca
ti
on
o
f
ga
ll
s
tone
objec
ts
[
12]
.
How
e
ve
r
,
the
f
a
c
t
is
that
the
r
e
a
r
e
s
ti
ll
vis
i
ble
s
hor
tcomings
in
M
R
C
P
im
a
ge
s
s
uc
h
a
s
s
pe
c
kle
nois
e
[
13]
,
[
14
]
.
B
a
s
e
d
on
th
is
,
the
im
a
ge
pr
oc
e
s
s
ing
pr
oc
e
s
s
is
ne
e
de
d
to
pr
ovide
a
n
opti
mal
r
ole
in
the
de
tec
ti
on
pr
oc
e
s
s
[
15]
.
T
he
c
onc
e
pt
of
im
a
ge
p
r
oc
e
s
s
ing
is
a
tec
hnique
f
or
a
na
lyzing
obj
e
c
ts
c
ontaine
d
in
a
n
im
a
ge
[
16]
.
I
mage
pr
oc
e
s
s
ing
c
a
n
a
ls
o
be
s
a
id
to
be
a
method
that
c
a
n
manipulate
i
mage
s
us
ing
c
olor
,
s
ha
pe
,
a
n
d
textur
e
[
17]
.
P
r
e
vious
r
e
s
e
a
r
c
h
e
xplains
that
i
mage
pr
o
c
e
s
s
ing
c
a
n
pr
ovide
quit
e
good
im
a
ge
outpu
t
in
s
e
ve
r
a
l
pr
oc
e
s
s
e
s
s
uc
h
a
s
de
tec
ti
on
[
18]
,
[
19]
.
T
he
im
a
ge
p
r
oc
e
s
s
ing
pe
r
f
or
manc
e
in
the
p
r
e
vious
c
a
s
e
ha
s
pr
o
vided
a
n
identif
ica
ti
on
a
c
c
ur
a
c
y
leve
l
of
[
20]
.
T
he
s
a
me
r
e
s
e
a
r
c
h
a
ls
o
r
e
por
ts
that
the
r
ole
of
im
a
ge
pr
oc
e
s
s
ing
c
a
n
c
ontr
ibut
e
e
f
f
e
c
ti
ve
ly
to
the
objec
t
identif
ica
ti
on
pr
oc
e
s
s
[
21]
.
S
im
il
a
r
r
e
s
e
a
r
c
h
ha
s
a
ls
o
p
r
ove
n
that
the
im
pleme
ntation
of
im
a
ge
pr
oc
e
s
s
ing
in
the
identif
i
c
a
ti
on
pr
oc
e
s
s
pr
ovides
quit
e
good
pe
r
f
or
manc
e
[
2
2]
.
O
ne
t
e
c
h
n
iq
ue
th
a
t
c
a
n
be
a
d
op
te
d
in
i
ma
ge
p
r
oc
e
s
s
i
ng
c
a
n
be
s
e
e
n
b
a
s
e
d
on
t
he
pe
r
f
o
r
ma
nc
e
o
f
s
e
gm
e
n
ta
ti
on
a
nd
e
x
t
r
a
c
ti
on
t
e
c
h
n
iq
ue
s
.
S
e
gm
e
n
tat
i
on
is
a
tec
hn
iq
ue
i
n
im
a
ge
p
r
o
c
e
s
s
in
g
th
a
t
is
c
a
p
a
b
le
o
f
s
ol
vi
ng
p
r
ob
le
ms
in
t
he
d
is
e
a
s
e
d
iag
nos
is
p
r
o
c
e
s
s
[
23
]
.
S
e
g
me
nt
a
t
io
n
te
c
h
ni
qu
e
s
a
r
e
a
b
le
t
o
c
a
r
r
y
ou
t
t
he
p
r
o
c
e
s
s
o
f
s
e
pa
r
a
ti
ng
o
bj
e
c
ts
in
a
n
im
a
ge
q
u
it
e
w
e
l
l
[
24
]
.
P
r
e
v
i
ous
r
e
s
e
a
r
c
h
r
e
p
o
r
t
e
d
t
ha
t
im
a
ge
s
e
gm
e
n
ta
ti
on
p
e
r
f
o
r
ma
nc
e
w
a
s
a
b
le
t
o
id
e
n
ti
f
y
ga
ll
s
to
n
e
i
ma
ge
ob
je
c
ts
w
it
h
a
n
a
c
c
u
r
a
c
y
r
a
te
o
f
9
1
%
[
25
]
.
F
ur
t
he
r
m
o
r
e
,
t
he
s
a
m
e
s
t
u
dy
ha
s
a
ls
o
r
e
po
r
te
d
t
ha
t
e
m
be
d
s
to
ne
ob
jec
ts
c
a
n
be
de
tec
te
d
we
ll
wi
th
a
ve
r
a
ge
pr
e
c
is
i
on
,
r
e
c
a
l
l
,
a
nd
d
e
v
ia
ti
o
n
r
a
t
io
v
a
l
ue
s
o
f
9
4
.
5
6
%
,
96
.
56
%
,
a
n
d
98
.
92
%
r
e
s
p
e
c
ti
ve
ly
[
2
6
]
.
S
e
gm
e
n
ta
ti
on
t
e
c
hn
iq
ue
s
ha
ve
a
ls
o
be
e
n
d
e
ve
lo
pe
d
in
a
c
c
o
r
da
nc
e
wi
th
th
e
n
e
e
d
f
o
r
s
o
lv
in
g
p
r
ob
le
ms
s
u
c
h
a
s
id
e
n
ti
f
ic
a
t
io
n
w
it
h
f
a
i
r
l
y
g
oo
d
r
e
s
u
l
t
s
[
27
]
–
[
2
9]
.
E
xtr
a
c
ti
on
tec
hniques
c
a
n
a
ls
o
play
a
n
im
por
tant
r
ole
in
digi
tal
im
a
ge
pr
oc
e
s
s
ing.
T
his
tec
hnique
is
a
ls
o
e
xpe
r
ienc
ing
de
ve
lopm
e
nt
a
long
with
incr
e
a
s
i
ng
pe
r
f
or
manc
e
in
the
identif
ica
ti
on
p
r
oc
e
s
s
[
30]
.
P
r
e
vious
r
e
s
e
a
r
c
h
e
xplains
that
e
xtr
a
c
ti
on
tec
hniques
c
a
n
pr
ovide
opti
mal
r
e
s
ult
s
in
the
objec
t
c
las
s
if
ica
ti
on
pr
oc
e
s
s
[
31]
.
T
he
de
ve
lopm
e
nt
of
e
xtr
a
c
ti
on
tec
hnique
s
is
a
ls
o
p
r
e
s
e
nted
in
model
f
or
m
to
pr
ovide
incr
e
a
s
e
d
a
c
c
ur
a
c
y
va
lues
in
the
identif
ica
ti
on
p
r
oc
e
s
s
[
32]
.
T
he
us
e
of
f
e
a
tur
e
s
e
lec
ti
on
a
lgor
it
hms
in
the
im
a
ge
e
xtr
a
c
ti
on
p
r
oc
e
s
s
is
a
ls
o
a
ble
to
make
a
n
a
c
ti
ve
c
ontr
ibut
ion
to
the
im
a
ge
pr
o
c
e
s
s
ing
pr
oc
e
s
s
[
33]
.
P
r
e
vious
r
e
s
e
a
r
c
h
r
e
po
r
ted
tha
t
f
e
a
tur
e
s
e
lec
ti
on
in
the
im
a
ge
e
xtr
a
c
ti
on
pr
oc
e
s
s
wa
s
a
b
le
to
pr
ovide
a
n
objec
t
s
e
lec
ti
on
pr
oc
e
s
s
by
uti
li
z
ing
the
a
tt
r
ibut
e
va
lues
of
a
n
im
a
ge
[
34]
.
B
a
s
e
d
on
pr
e
vious
r
e
s
e
a
r
c
h,
the
p
r
oc
e
s
s
of
de
tec
ti
ng
ga
ll
s
tone
objec
ts
in
M
R
C
P
im
a
ge
s
ne
e
ds
to
be
opti
mi
z
e
d
by
de
ve
lopi
ng
a
f
e
a
tur
e
s
e
lec
ti
on
pr
oc
e
s
s
in
e
xtr
a
c
ti
on
tec
hnique
s
.
T
his
opti
mi
z
a
ti
on
is
a
im
e
d
a
t
maximi
z
ing
the
de
tec
ti
on
pr
oc
e
s
s
a
nd
r
e
s
ult
s
that
will
be
c
a
r
r
ied
ou
t
in
the
de
ve
lopm
e
nt
of
the
f
e
a
tur
e
s
e
lec
ti
on
pr
oc
e
s
s
us
ing
the
f
e
a
tur
e
s
e
lec
ti
on
s
tati
s
ti
c
a
l
a
n
a
lys
is
(
F
S
S
A)
a
lgor
it
hm.
T
he
pe
r
f
o
r
manc
e
of
th
e
F
S
S
A
a
lgor
it
hm
p
r
ovides
a
n
a
na
lys
is
pr
oc
e
s
s
in
de
ter
mi
ning
opti
mal
c
ha
r
a
c
ter
is
ti
c
pa
tt
e
r
ns
invol
ving
c
l
a
s
s
if
ica
ti
on
methods
s
uc
h
a
s
k
-
ne
a
r
e
s
t
ne
ighbor
(
KN
N)
,
s
uppor
t
ve
c
tor
mac
hine
(
S
VM
)
,
a
nd
a
r
ti
f
icia
l
ne
ur
a
l
ne
twor
k
(
AN
N)
a
s
we
ll
a
s
the
P
e
a
r
s
on
c
or
r
e
lation
(
P
C
)
met
hod.
T
he
de
ve
lopm
e
nt
of
the
F
S
S
A
a
lgor
i
thm
in
th
e
f
e
a
tur
e
s
e
lec
ti
on
pr
oc
e
s
s
c
a
n
a
ls
o
pr
ovide
nove
lt
y
in
im
a
g
e
e
xtr
a
c
ti
on
tec
hniques
us
e
d
in
de
tec
ti
ng
ga
ll
s
tone
c
ontent.
T
he
pe
r
f
or
manc
e
of
F
S
S
A
is
a
ls
o
e
xpe
c
ted
to
be
a
b
le
to
pr
e
s
e
nt
opti
mal
c
ha
r
a
c
ter
is
ti
c
pa
tt
e
r
ns
f
r
om
pr
e
vious
ly
de
tec
ted
im
a
ge
objec
ts
.
Ove
r
a
ll
,
th
is
r
e
s
e
a
r
c
h
c
a
n
c
ontr
ibut
e
to
he
l
ping
medic
a
l
pa
r
t
ies
in
the
p
r
oc
e
s
s
of
diagnos
ing
pa
ti
e
nts
who
ha
ve
identif
ied
ga
ll
s
tone
dis
e
a
s
e
.
2.
RE
S
E
AR
CH
M
E
T
HO
D
T
he
pr
oc
e
s
s
of
de
tec
ti
ng
ga
ll
s
tone
objec
ts
in
M
R
C
P
im
a
ge
s
is
c
a
r
r
ied
out
by
de
ve
lopi
ng
e
xtr
a
c
ti
on
tec
hniques
in
the
f
e
a
tur
e
s
e
lec
ti
on
pr
oc
e
s
s
ba
s
e
d
o
n
the
pe
r
f
or
manc
e
of
the
F
S
S
A
a
lgor
it
hm.
T
he
pe
r
f
or
manc
e
of
the
de
ve
loped
F
S
S
A
is
a
im
e
d
a
t
p
r
ovidi
ng
a
c
c
ur
a
te
de
tec
ti
on
r
e
s
ult
s
f
or
ga
ll
s
tone
objec
ts
on
M
R
C
P
im
a
ge
s
.
T
he
pe
r
f
or
manc
e
o
f
the
F
S
S
A
a
lgor
it
hm
is
pr
e
s
e
nted
in
s
e
ve
r
a
l
s
tage
s
s
tar
ti
ng
f
r
o
m
the
p
r
e
pr
oc
e
s
s
ing,
s
e
gmenta
ti
on
a
nd
e
xtr
a
c
ti
on
s
tage
s
.
B
a
s
e
d
on
th
e
s
e
s
tage
s
,
the
pe
r
f
o
r
manc
e
of
F
S
S
A
in
the
p
r
oc
e
s
s
of
opti
mi
z
ing
the
ga
ll
s
tone
objec
t
de
tec
ti
on
pr
oc
e
s
s
in
f
e
a
tur
e
s
e
lec
ti
on
will
be
a
ble
to
pr
e
s
e
nt
ne
w
e
xtr
a
c
ti
on
tec
hniques
in
im
a
ge
pr
oc
e
s
s
ing.
T
he
pe
r
f
or
manc
e
of
the
F
S
S
A
a
lgo
r
it
hm
in
the
ga
ll
s
tone
objec
t
de
tec
ti
on
pr
oc
e
s
s
is
pr
e
s
e
nted
in
the
r
e
s
e
a
r
c
h
f
r
a
mew
or
k
in
F
igur
e
1
.
F
igur
e
1
is
a
n
il
lus
tr
a
ti
on
of
the
pe
r
f
o
r
manc
e
of
t
he
F
S
S
A
a
lgor
it
hm
in
de
tec
ti
ng
ga
ll
s
tone
objec
ts
.
T
he
de
tec
ti
on
p
r
oc
e
s
s
ba
s
e
d
on
the
pe
r
f
or
manc
e
of
the
F
S
S
A
a
lgor
it
hm
be
gins
with
the
pe
r
f
o
r
man
c
e
of
the
s
e
gmenta
ti
on
pr
oc
e
s
s
whi
c
h
is
ba
s
e
d
on
the
pe
r
f
or
manc
e
of
the
mul
ti
s
tage
s
e
gmenta
ti
on
a
lgor
it
hm
(
M
S
A)
in
F
igur
e
1
(
a
)
.
T
he
M
S
A
a
lgor
it
hm
invol
ve
s
k
-
m
e
a
ns
c
lus
ter
-
ba
s
e
d
s
e
gmenta
ti
on
(
C
B
S
)
c
ombi
n
e
d
with
mor
phologi
c
a
l
s
e
gmenta
ti
on.
T
he
output
o
f
the
s
e
g
menta
ti
on
r
e
s
ult
s
will
late
r
be
c
ome
inpu
t
f
o
r
the
e
x
tr
a
c
ti
on
pr
oc
e
s
s
us
ing
the
pe
r
f
or
manc
e
of
the
F
S
S
A
a
lg
or
it
hm
in
F
igur
e
1(
b
)
.
T
he
F
S
S
A
a
lgor
it
hm
p
r
e
s
e
nts
the
de
ve
lopm
e
nt
o
f
the
e
xtr
a
c
ti
on
pr
oc
e
s
s
in
f
e
a
tur
e
s
e
lec
ti
on
by
invol
ving
the
pe
r
f
or
manc
e
o
f
the
KN
N,
S
VM
,
a
nd
AN
N
c
las
s
if
ica
ti
on
methods
a
s
we
ll
a
s
the
P
C
method
in
mea
s
ur
ing
the
c
or
r
e
lation
of
e
a
c
h
f
e
a
tur
e
pa
tt
e
r
n
pr
oduc
e
d.
T
he
pe
r
f
o
r
manc
e
r
e
s
ult
s
of
the
F
S
S
A
a
lgor
it
hm
a
r
e
a
ble
to
pr
ovide
opti
mi
z
a
ti
on
of
the
d
e
tec
ti
on
pr
oc
e
s
s
f
or
ga
ll
s
tone
c
ontent
objec
ts
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
Ar
ti
f
I
ntell
I
S
S
N:
2252
-
8938
Optimiz
ing
the
gall
s
tone
de
tec
ti
on
pr
oc
e
s
s
w
it
h
featur
e
s
e
lec
ti
on
s
tat
is
ti
c
al
analys
is
algor
it
hm
(
M
us
li
Y
anto
)
1185
(
a
)
(
b)
F
igur
e
1.
R
e
s
e
a
r
c
h
f
r
a
mew
or
k
of
(
a
)
s
e
gmenta
ti
on
pr
oc
e
s
s
a
nd
(
b)
e
xtr
a
c
ti
on
with
f
e
a
tur
e
s
e
lec
ti
on
ba
s
e
d
on
F
S
S
A
a
lgor
it
h
m
pe
r
f
or
manc
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
Ar
ti
f
I
ntell
,
Vol.
14
,
No.
2
,
Apr
il
20
25
:
1183
-
119
1
1186
Ne
xt,
the
pe
r
f
or
manc
e
s
tage
s
of
the
F
S
S
A
a
lgor
it
h
m
c
a
n
be
e
xplaine
d
a
s
f
ol
lows
:
‒
I
mage
pr
e
pr
oc
e
s
s
ing
s
tage
:
t
his
s
tage
is
the
s
tage
u
s
e
d
f
or
the
pr
oc
e
s
s
of
im
pr
oving
the
input
im
a
ge
.
I
mage
im
pr
ove
ment
include
s
s
e
ve
r
a
l
pr
oc
e
s
s
e
s
includin
g
gr
a
y
im
a
ge
tr
a
ns
f
o
r
mation,
im
a
ge
a
djus
tm
e
nt,
a
nd
f
il
ter
ing.
T
he
outpu
t
o
f
the
p
r
e
pr
oc
e
s
s
ing
im
a
ge
will
late
r
be
c
ome
the
input
i
mage
a
t
the
s
e
gment
a
ti
on
pr
oc
e
s
s
s
tage
.
‒
I
mage
s
e
gmenta
ti
on
s
tage
:
t
he
s
e
gmenta
ti
on
pr
oc
e
s
s
is
a
n
a
dva
nc
e
d
s
tage
in
F
S
S
A
pe
r
f
or
manc
e
in
ga
ll
s
tone
objec
t
de
tec
ti
on.
T
he
s
e
gmenta
ti
on
pr
oc
e
s
s
a
dopts
the
pe
r
f
o
r
manc
e
of
the
M
S
A
a
lgor
it
hm
by
playin
g
the
r
ole
of
t
he
C
B
S
method
whic
h
is
opti
mi
z
e
d
with
th
e
e
lbow
method
in
s
e
pa
r
a
ti
ng
objec
ts
.
T
he
r
e
s
ult
s
of
the
s
e
gmenta
ti
on
pr
oc
e
s
s
will
late
r
be
c
ome
input
in
th
e
im
a
ge
e
xtr
a
c
ti
on
p
r
oc
e
s
s
.
‒
I
mage
e
xtr
a
c
ti
on
s
tage
:
t
he
e
xtr
a
c
ti
on
pr
oc
e
s
s
us
i
ng
f
e
a
tur
e
s
e
lec
ti
on
wa
s
de
ve
lo
pe
d
by
us
ing
s
tati
s
ti
c
a
l
a
na
lys
is
methods
on
the
pe
r
f
or
manc
e
of
P
C
.
T
he
pe
r
f
or
manc
e
of
the
P
C
method
c
a
n
p
r
ovide
a
n
op
ti
mal
r
ole
in
pr
e
s
e
nti
ng
opti
mal
c
ha
r
a
c
ter
is
ti
c
pa
tt
e
r
ns
of
ga
ll
s
tone
objec
ts
.
Ove
r
a
ll
,
the
im
p
r
ove
ment
i
n
the
f
e
a
tur
e
s
e
lec
ti
on
pr
oc
e
s
s
b
a
s
e
d
on
F
S
S
A
pe
r
f
or
manc
e
c
a
n
pr
ovide
nove
lt
y
in
the
pr
oc
e
s
s
of
de
t
e
c
ti
ng
ga
ll
s
tone
objec
ts
.
2.
1.
Re
s
e
ar
c
h
d
at
as
e
t
T
he
r
e
s
e
a
r
c
h
da
tas
e
t
us
e
s
M
R
C
P
im
a
ge
s
s
our
c
e
d
f
r
om
pa
ti
e
nts
a
t
S
a
nta
M
a
r
ia
Hos
pit
a
l
P
e
ka
nba
r
u
a
nd
S
it
i
R
a
hmah
Hos
pit
a
l
P
a
da
ng.
T
he
da
tas
e
t
c
ons
is
ts
of
2371
im
a
ge
s
f
r
om
32
pa
ti
e
nts
with
indi
c
a
ti
ons
of
ga
ll
s
tone
dis
e
a
s
e
.
T
he
s
a
mpl
e
r
e
s
e
a
r
c
h
da
tas
e
t
c
a
n
be
pr
e
s
e
nted
in
F
igur
e
2.
F
igur
e
2
is
a
s
a
mpl
e
of
M
R
C
P
im
a
ge
r
e
s
ult
s
us
e
d
a
s
a
r
e
s
e
a
r
c
h
da
tas
e
t
in
the
ga
ll
s
tone
objec
t
de
tec
ti
on
pr
oc
e
s
s
.
F
igur
e
2
(
a
)
de
picts
a
n
M
R
C
P
im
a
ge
of
a
ga
ll
s
tone
pa
ti
e
nt
on
im
a
ge
s
li
c
e
64
,
F
igur
e
2(
b)
i
s
a
ls
o
a
n
M
R
C
P
im
a
ge
of
a
ga
ll
s
tone
pa
ti
e
nt
on
im
a
ge
s
li
c
e
65,
a
nd
F
igur
e
2(
c
)
is
a
ls
o
one
of
the
M
R
C
P
im
a
ge
s
of
a
ga
ll
s
tone
pa
ti
e
nt
on
im
a
ge
s
li
c
e
66
.
T
he
da
tas
e
t
on
the
M
R
C
P
im
a
ge
(
.
*jpg
f
or
mat)
wi
th
a
n
im
a
ge
r
e
s
olut
ion
of
512×
512
pixels
.
T
he
da
tas
e
t
will
late
r
be
divi
de
d
int
o
1921
tr
a
ini
ng
da
ta
a
nd
450
da
t
a
f
o
r
tes
ti
ng.
(
a)
(
b)
(
c
)
F
igur
e
2.
M
R
C
P
im
a
ge
da
tas
e
t
:
(
a
)
S
li
c
e
64
,
(
b
)
S
l
ice
65
,
a
nd
(
c
)
S
li
c
e
66
2.
2.
M
u
lt
is
t
age
s
e
gm
e
n
t
at
ion
algorit
h
m
T
he
s
e
gmenta
ti
on
pr
oc
e
s
s
with
the
M
S
A
a
lgor
it
h
m
a
dopts
the
pe
r
f
or
manc
e
o
f
C
B
S
c
ombi
ne
d
with
mor
phologi
c
a
l
s
e
gmenta
ti
on.
C
B
S
s
e
gmenta
ti
on
i
nvolves
the
k
-
mea
ns
c
lus
ter
whic
h
is
opti
mi
z
e
d
u
s
ing
the
e
lbow
method.
T
he
pe
r
f
or
manc
e
of
the
e
lbow
m
e
thod
c
a
n
pr
e
s
e
nt
opti
mal
c
lus
ter
(
k)
va
lues
ba
s
e
d
on
the
c
a
lcula
ti
on
of
the
s
um
s
qua
r
e
e
r
r
or
(
S
S
E
)
[
35
]
.
T
h
e
e
qua
ti
ons
us
e
d
in
the
pe
r
f
or
manc
e
of
the
M
S
A
a
lgor
it
hm
a
r
e
pr
e
s
e
nted
in
(
1
)
-
(
3)
[
36]
,
[
37
]
.
In
(
1)
is
the
f
o
r
mu
la
us
e
d
to
mea
s
ur
e
dis
tanc
e
us
ing
E
uc
li
dian
dis
tanc
e
in
the
k
-
mea
ns
c
lu
s
ter
pr
oc
e
s
s
.
Dis
tanc
e
mea
s
ur
e
ments
a
r
e
c
a
lcula
ted
on
im
a
ge
int
e
ns
it
y
to
gr
oup
im
a
ge
objec
ts
.
In
(
2
)
is
the
f
or
mul
a
us
e
d
to
c
a
lcula
te
the
S
S
E
in
the
e
lbow
method
.
T
he
f
o
r
mul
a
is
us
e
d
a
s
a
f
or
m
of
opti
mi
z
ing
the
c
lus
ter
p
r
oc
e
s
s
in
C
B
S
in
s
e
pa
r
a
ti
ng
im
a
ge
objec
ts
.
T
he
c
ombi
na
ti
on
of
(
1)
a
nd
(
2)
p
r
oduc
e
s
(
3)
in
de
ter
m
ini
ng
the
opti
ma
l
k
va
lue
us
e
d
in
the
C
B
S
-
ba
s
e
d
c
lus
ter
pr
oc
e
s
s
.
(
,
)
=
√
∑
(
−
)
2
1
;
=
1
,
2
,
3
,
…
.
,
(
1)
=
∑
‖
−
‖
2
=
1
(
2)
K_E
lbow=
∑
‖
√
∑
(
x
i
−
y
i
)
2
k
1
−
d
(
x
,
y
)
∑
d
(
x
,
y
)
k
1
‖
n
k
=
1
(
3)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
Ar
ti
f
I
ntell
I
S
S
N:
2252
-
8938
Optimiz
ing
the
gall
s
tone
de
tec
ti
on
pr
oc
e
s
s
w
it
h
featur
e
s
e
lec
ti
on
s
tat
is
ti
c
al
analys
is
algor
it
hm
(
M
us
li
Y
anto
)
1187
2.
3.
F
e
a
t
u
r
e
s
e
lec
t
ion
F
e
a
tur
e
s
e
lec
ti
on
in
im
a
ge
e
xtr
a
c
ti
on
is
us
e
d
a
s
a
d
e
tec
ti
on
pr
oc
e
s
s
by
uti
li
z
ing
the
a
tt
r
ibut
e
va
lues
f
or
e
a
c
h
im
a
ge
f
e
a
tur
e
.
T
he
f
e
a
tur
e
s
e
lec
ti
on
pr
oc
e
s
s
a
dopts
the
p
e
r
f
or
manc
e
of
the
da
ta
c
las
s
if
ica
ti
on
pr
oc
e
s
s
by
taking
the
leve
l
o
f
a
c
c
ur
a
c
y
a
s
a
s
e
lec
ti
on
pa
r
a
met
e
r
[
38]
.
T
he
f
e
a
tu
r
e
s
e
lec
ti
on
pr
oc
e
s
s
us
e
s
s
e
ve
r
a
l
methods
s
uc
h
a
s
KN
N,
S
VM
,
a
nd
AN
N.
KN
N
is
a
n
a
lgor
it
hm
c
onc
e
pt
in
s
upe
r
vis
e
d
lea
r
ning
that
c
ons
ider
s
t
he
us
e
of
many
k
c
ompar
a
ble
pa
tt
e
r
ns
in
the
t
r
a
ini
ng
[
39]
.
S
VM
ha
s
c
ont
r
ibut
e
d
g
r
e
a
tl
y
to
ha
ndli
ng
c
las
s
if
ica
ti
on
pr
oblems
ba
s
e
d
on
hype
r
plane
s
[
40]
.
T
he
pe
r
f
or
manc
e
of
AN
N
ha
s
a
l
s
o
be
e
n
pr
ove
n
to
pr
ovide
maximum
r
e
s
ult
s
s
uc
h
a
s
identif
ica
ti
on,
c
las
s
if
ica
ti
on,
a
nd
pr
e
diction
pr
oblems
[
41
]
,
[
42]
.
2.
4.
P
e
r
s
on
co
r
r
e
lat
ion
P
C
is
a
s
tatis
ti
c
a
l
a
na
lys
is
c
onc
e
pt
that
is
c
a
pa
ble
of
mea
s
ur
ing
c
or
r
e
lation
[
43
]
.
P
C
methods
c
a
n
be
c
ombi
ne
d
to
im
pr
ove
a
na
lys
is
pe
r
f
or
manc
e
with
q
uit
e
good
output
[
44]
.
T
he
P
C
method
ha
s
a
ls
o
be
e
n
us
e
d
to
r
e
view
the
a
c
c
ur
a
c
y
of
a
model
a
na
lys
is
[
45
]
.
R
e
ga
r
ding
P
C
pe
r
f
or
manc
e
,
it
c
a
n
be
s
e
e
n
in
(
4
)
a
nd
(
5)
[
46]
.
In
(
4
)
a
nd
(
5
)
is
th
e
f
or
m
ula
us
e
d
to
c
a
l
c
u
la
te
t
he
va
lu
e
o
f
c
o
v
(
X
,
Y
)
wh
ic
h
is
t
he
c
ova
r
ia
nc
e
b
e
t
we
e
n
X
a
n
d
Y
.
T
he
va
lue
o
f
X
,
Y
c
a
n
b
e
i
nt
e
r
p
r
e
te
d
a
s
a
s
t
a
n
da
r
d
de
vi
a
t
io
n
va
l
ue
o
f
t
he
va
r
i
a
b
les
X
a
n
d
Y
[
4
7
]
.
P
C
p
e
r
f
o
r
m
a
n
c
e
h
a
s
c
o
n
tr
i
bu
ted
t
o
ma
x
im
iz
in
g
t
he
im
a
g
e
p
r
oc
e
s
s
in
g
p
r
oc
e
s
s
i
n
a
m
od
e
l
th
a
t
ha
s
be
e
n
de
s
i
gn
e
d
[
48
]
.
,
=
(
c
ov
(
X
,
Y
)
.
)
=
(
E
(
(
c
ov
(
X
−
µX
)
(
Y
−
µY
)
)
.
)
(
4)
,
=
(
c
ov
(
X
,
Y
)
.
)
=
E
(
XY
)
−
E
(
X
)
E
(
Y
)
√
E
(
X
.
X
)
−
.
(
)
√
(
.
)
−
.
(
)
(
5)
3.
RE
S
UL
T
S
AN
D
DI
S
CU
S
S
I
ON
O
p
ti
mi
z
i
ng
the
ga
ll
s
to
ne
o
bj
e
c
t
de
tec
t
io
n
p
r
o
c
e
s
s
b
y
d
e
ve
lo
p
in
g
f
e
a
t
u
r
e
s
e
lec
t
io
n
us
i
ng
F
S
S
A
is
t
he
m
a
in
t
op
ic
o
f
r
e
s
e
a
r
c
h
.
T
he
de
ve
lo
pm
e
n
t
o
f
f
e
a
t
u
r
e
s
e
le
c
t
io
n
ba
s
e
d
on
t
he
p
e
r
f
o
r
m
a
nc
e
o
f
the
F
S
S
A
a
lgo
r
i
th
m
in
t
he
e
x
t
r
a
c
ti
on
p
r
oc
e
s
s
is
a
no
ve
lt
y
t
o
p
r
ov
id
e
i
mp
r
o
ve
me
nts
in
th
e
de
te
c
t
io
n
o
f
g
a
l
ls
to
ne
o
bj
e
c
ts
.
T
he
pe
r
f
o
r
ma
nc
e
o
f
th
e
F
S
S
A
r
e
s
u
lt
s
th
a
t
ha
ve
b
e
e
n
de
ve
lo
pe
d
w
il
l
be
a
b
le
to
pr
ov
i
de
a
de
tec
ti
on
m
ode
l
tha
t
pr
ov
id
e
s
a
c
c
u
r
a
c
y
.
3.
1.
I
m
age
p
r
e
p
r
oc
e
s
s
in
g
T
he
pr
e
pr
oc
e
s
s
ing
s
tage
is
the
ini
ti
a
l
s
tage
of
the
ga
ll
s
tone
objec
t
de
tec
ti
on
pr
oc
e
s
s
.
T
his
pr
oc
e
s
s
plays
a
r
ole
in
pr
ovidi
ng
a
n
inc
r
e
a
s
e
in
the
qua
li
ty
of
t
he
i
nput
im
a
ge
us
e
d.
P
r
e
p
r
oc
e
s
s
ing
r
e
s
ult
s
invol
ve
s
e
ve
r
a
l
pr
oc
e
s
s
e
s
including
gr
a
y
im
a
ge
tr
a
ns
f
or
mation,
im
a
ge
a
djus
tm
e
nt,
a
nd
f
il
ter
ing.
T
he
pr
e
pr
oc
e
s
s
ing
r
e
s
ult
s
c
a
n
be
pr
e
s
e
nted
in
F
igur
e
3.
F
igur
e
3
is
the
output
of
the
pr
e
pr
oc
e
s
s
ing
s
tage
us
ing
the
gr
a
y
im
a
ge
tr
a
ns
f
or
mation
pr
oc
e
s
s
,
im
a
ge
a
djus
tm
e
nt,
a
nd
f
i
lt
e
r
ing.
F
igu
r
e
3(
a
)
is
the
input
i
mage
in
the
de
tec
ti
on
pr
oc
e
s
s
.
F
igur
e
3(
b
)
is
the
gr
a
y
im
a
ge
tr
a
ns
f
or
mation
pr
oc
e
s
s
whic
h
is
the
be
ginni
ng
of
pr
e
pr
oc
e
s
s
ing.
F
igur
e
3(
c
)
is
the
r
e
s
ult
of
c
onti
nue
d
pr
e
pr
oc
e
s
s
ing
invol
ving
the
im
a
ge
a
djus
tm
e
nt
pr
o
c
e
s
s
.
F
igur
e
3(
d)
is
the
f
inal
s
tage
of
p
r
e
pr
oc
e
s
s
ing,
whic
h
pr
e
s
e
nts
the
im
a
ge
output
f
r
om
the
f
il
ter
ing
pr
oc
e
s
s
.
I
n
the
p
r
e
pr
oc
e
s
s
ing
s
tage
c
a
n
be
s
e
e
n
that
th
e
r
e
is
a
n
im
pr
ove
ment
in
the
qua
li
ty
of
the
input
im
a
ge
us
e
d
pr
e
vious
ly.
T
he
output
o
f
the
pr
e
p
r
oc
e
s
s
ing
im
a
ge
will
be
the
input
in
the
de
tec
ti
on
p
r
oc
e
s
s
.
(
a
)
(
b)
(
c
)
(
d)
F
igur
e
3.
P
r
e
p
r
oc
e
s
s
ing
re
s
ult
(
a
)
inpu
t
im
a
ge
,
(
b
)
gr
a
y
tr
a
ns
f
or
mation
,
(
c
)
i
mage
a
djus
tm
e
nt,
a
nd
(
d)
r
e
s
ult
s
f
il
ter
ing
3.
2.
M
u
lt
is
t
age
s
e
gm
e
n
t
at
ion
alg
or
it
h
m
T
he
s
e
gmenta
ti
on
pr
oc
e
s
s
us
ing
the
M
S
A
a
lgor
it
h
m
is
pa
r
t
of
the
ga
ll
s
tone
objec
t
de
tec
ti
on
pr
oc
e
s
s
.
T
he
pe
r
f
or
manc
e
of
the
M
S
A
a
lgor
it
hm
a
dopts
a
C
B
S
a
ppr
oa
c
h
a
nd
ope
r
a
ti
ona
l
mor
phology
to
gua
r
a
ntee
the
a
c
c
ur
a
c
y
of
s
e
gmente
d
objec
ts
with
pr
e
c
is
e
a
nd
a
c
c
ur
a
te
r
e
s
ult
s
.
T
he
pe
r
f
or
manc
e
r
e
s
ult
s
of
the
M
S
A
a
lgor
it
hm
in
s
e
gmenta
ti
on
c
a
n
be
pr
e
s
e
nted
in
F
igur
e
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
Ar
ti
f
I
ntell
,
Vol.
14
,
No.
2
,
Apr
il
20
25
:
1183
-
119
1
1188
F
igur
e
4
is
the
r
e
s
u
lt
o
f
the
s
e
gmenta
ti
on
pr
oc
e
s
s
of
the
M
S
A
a
lgor
it
hm
pe
r
f
or
manc
e
in
de
tec
ti
ng
ga
ll
s
tone
dis
e
a
s
e
.
F
igur
e
4(
a
)
is
the
r
e
s
ult
of
the
pr
e
vious
pr
e
pr
oc
e
s
s
ing
im
a
ge
whic
h
is
us
e
d
a
s
the
input
im
a
ge
in
the
s
e
gmenta
ti
on
pr
oc
e
s
s
with
the
M
S
A
a
lgor
it
hm.
F
igur
e
4(
b
)
i
s
the
r
e
s
ult
of
C
B
S
s
e
gmenta
ti
on
whic
h
is
one
pa
r
t
of
the
s
e
gmenta
ti
on
pr
oc
e
s
s
in
the
M
S
A
a
lgor
it
hm.
F
igur
e
4
(
c
)
is
the
r
e
s
ult
of
mor
ph
ologi
c
a
l
s
e
gmenta
ti
on
whic
h
is
the
ne
xt
s
tage
in
the
s
e
gmen
tation
pr
oc
e
s
s
in
the
M
S
A
a
lgor
it
hm.
F
igur
e
4
(
d)
is
the
f
inal
output
of
the
M
S
A
a
lgor
it
hm
s
e
gmenta
ti
on
pr
oc
e
s
s
in
de
tec
ti
ng
ga
ll
s
tone
objec
ts
.
T
he
pe
r
f
or
manc
e
of
t
he
M
S
A
a
lgor
it
hm
ha
s
be
e
n
a
ble
to
de
s
c
r
ibe
the
objec
t
of
ga
ll
s
tone
dis
e
a
s
e
quit
e
we
ll
.
T
he
r
e
s
ult
s
of
the
M
S
A
a
lgor
it
hm
s
e
gmenta
ti
on
will
be
us
e
d
a
s
input
in
the
i
mage
e
xt
r
a
c
ti
on
pr
oc
e
s
s
.
(
a
)
(
b)
(
c
)
(
d)
F
igur
e
4.
R
e
s
ult
s
of
the
M
S
A
a
lgor
it
hm
s
e
gmenta
ti
on
pr
oc
e
s
s
(
a
)
p
r
e
pr
oc
e
s
s
ing
r
e
s
ult
,
(
b
)
C
B
S
r
e
s
ult
s
,
(
c
)
m
or
pho
logy
r
e
s
ult
s
,
a
nd
(
d
)
M
S
A
r
e
s
ult
s
3.
3
.
F
e
a
t
u
r
e
s
e
lec
t
ion
s
t
at
is
t
ical
an
alys
is
algorit
h
m
e
xt
r
ac
t
ion
De
ve
lopm
e
nt
of
f
e
a
tur
e
s
e
lec
ti
on
in
the
de
lay
im
a
g
e
e
xtr
a
c
ti
on
p
r
oc
e
s
s
us
ing
F
S
S
A
f
or
the
p
r
oc
e
s
s
of
incr
e
a
s
ing
de
tec
ti
on.
T
he
F
S
S
A
pe
r
f
o
r
manc
e
pr
oc
e
s
s
plays
t
he
r
ole
of
the
c
las
s
if
ica
ti
on
method
a
nd
c
o
r
r
e
lation
e
xa
mi
ne
r
s
f
ound
in
the
im
a
ge
c
ha
r
a
c
ter
is
ti
c
pa
tt
e
r
n
.
C
las
s
if
ica
ti
on
methods
s
uc
h
a
s
KN
N,
S
VM
,
a
nd
AN
N
a
r
e
invol
ve
d
in
f
indi
ng
opti
mal
f
e
a
tur
e
pa
t
ter
ns
.
T
he
pe
r
f
or
manc
e
r
e
s
ult
s
o
f
the
F
S
S
A
a
lgor
it
hm
in
th
e
f
e
a
tur
e
s
e
lec
ti
on
pr
oc
e
s
s
c
a
n
be
pr
e
s
e
nted
in
T
a
ble
1.
B
a
s
e
d
on
T
a
ble
1
,
i
t
c
a
n
be
s
e
e
n
that
the
f
e
a
tur
e
s
e
lec
ti
on
pr
oc
e
s
s
by
a
dopti
ng
the
pe
r
f
or
manc
e
of
the
F
S
S
A
a
lgor
it
h
m
ha
s
gone
we
ll
in
de
ter
mi
ning
im
a
ge
c
ha
r
a
c
ter
is
ti
c
pa
tt
e
r
ns
.
T
he
r
e
s
ult
s
of
the
f
e
a
tur
e
s
e
lec
ti
on
pr
oc
e
s
s
whic
h
invol
ve
s
s
e
ve
r
a
l
c
las
s
if
ica
ti
on
methods
with
P
C
s
t
a
ti
s
ti
c
a
l
a
na
lys
is
methods
ha
ve
be
e
n
a
ble
to
il
lus
tr
a
te
a
mo
r
e
e
f
f
e
c
ti
ve
f
e
a
tur
e
s
e
lec
ti
on
pr
oc
e
s
s
.
T
he
pe
r
f
o
r
m
a
nc
e
tes
t
r
e
s
ult
s
of
the
F
S
S
A
a
lgor
it
hm
c
a
n
be
pr
e
s
e
nted
in
T
a
ble
2.
T
a
bl
e
1
is
th
e
r
e
s
u
lt
of
mea
s
u
r
i
ng
th
e
pe
r
f
o
r
ma
nc
e
of
t
he
F
S
S
A
a
lg
o
r
i
t
hm
in
d
e
t
e
c
t
i
ng
g
a
l
ls
to
ne
o
bj
e
c
ts
.
B
a
s
e
d
on
th
is
t
a
b
le
,
it
c
a
n
be
s
e
e
n
th
a
t
th
e
te
x
tu
r
e
f
e
a
tu
r
e
o
ut
pe
r
f
o
r
ms
o
the
r
f
e
a
t
u
r
e
s
in
pr
e
s
e
nt
in
g
o
pt
i
ma
l
i
ma
ge
c
ha
r
a
c
te
r
is
ti
c
p
a
t
te
r
ns
.
T
h
is
c
ha
r
a
c
te
r
is
t
ic
pa
tt
e
r
n
c
a
n
b
e
s
e
e
n
ba
s
e
d
on
t
he
t
ot
a
l
pe
r
f
or
ma
nc
e
a
c
c
u
r
a
c
y
o
f
t
he
K
NN
c
las
s
i
f
i
c
a
ti
on
me
th
od
o
f
84
.
61
%
,
S
VM
90
.
38
%
,
a
n
d
A
NN
99
.
95
%
.
T
he
r
e
s
u
lt
s
o
f
c
o
r
r
e
l
a
t
io
n
mea
s
u
r
e
m
e
n
ts
us
i
ng
t
he
P
C
me
th
od
ha
v
e
p
r
e
s
e
nt
e
d
a
c
or
r
e
la
t
io
n
le
ve
l
o
f
9
5
.
90
%
.
B
a
s
e
d
on
thes
e
r
e
s
u
lt
s
,
t
he
p
e
r
f
o
r
m
a
nc
e
o
f
th
e
F
S
S
A
m
e
th
od
is
q
ui
te
go
od
,
p
r
o
vi
di
ng
a
n
im
p
r
o
ve
men
t
in
t
he
de
tec
t
io
n
p
r
oc
e
s
s
f
or
ga
ll
s
t
one
ob
je
c
ts
.
I
mpr
oving
the
ga
ll
s
tone
objec
t
de
tec
ti
on
pr
oc
e
s
s
us
ing
the
F
S
S
A
a
lgor
it
hm
c
a
n
pr
ovide
nove
lt
y
in
the
f
e
a
tur
e
s
e
le
c
ti
on
pr
oc
e
s
s
in
e
xtr
a
c
ti
on
tec
hniques
.
T
e
s
ti
ng
the
pe
r
f
o
r
manc
e
of
F
S
S
A
in
incr
e
a
s
ing
de
tec
ti
on
ha
s
be
e
n
pr
ove
n
to
be
quit
e
good
with
a
n
a
c
c
ur
a
c
y
of
95.
83
%
,
s
e
ns
it
ivi
ty
of
96
.
96%
,
a
nd
s
pe
c
if
icity
of
95.
23%
.
T
e
s
ti
ng
ba
s
e
d
on
s
e
ve
r
a
l
pr
e
vious
s
tudi
e
s
is
a
l
s
o
a
pr
oc
e
s
s
of
pr
oving
the
pe
r
f
o
r
manc
e
of
the
F
S
S
A
a
l
gor
it
hm
in
im
pr
oving
the
p
r
oc
e
s
s
of
de
tec
ti
ng
ga
ll
s
tone
obj
e
c
ts
pr
e
s
e
nted
in
T
a
ble
3.
T
a
ble
3
is
a
f
or
m
of
tes
ti
ng
the
pe
r
f
or
manc
e
r
e
s
ult
s
of
the
F
S
S
A
a
lgor
it
hm
with
p
r
e
vious
r
e
s
e
a
r
c
h
in
the
de
tec
ti
on
pr
o
c
e
s
s
.
B
a
s
e
d
on
thes
e
r
e
s
ult
s
,
it
c
a
n
be
s
tate
d
that
the
F
S
S
A
a
lgor
it
hm
c
a
n
p
r
ovide
a
c
c
ur
a
te
im
a
ge
c
ha
r
a
c
ter
is
ti
c
pa
tt
e
r
ns
in
the
ga
ll
s
tone
obje
c
t
de
tec
ti
on
p
r
oc
e
s
s
.
Ove
r
a
ll
,
th
is
r
e
s
e
a
r
c
h
ha
s
be
e
n
quit
e
s
uc
c
e
s
s
f
ul
a
nd
c
a
n
c
ontr
ibut
e
to
opti
mi
z
ing
the
pr
o
c
e
s
s
of
diagnos
ing
ga
ll
s
tone
dis
e
a
s
e
.
T
a
ble
1.
T
he
pe
r
f
or
manc
e
r
e
s
ult
s
of
the
F
S
S
A
a
lgor
it
hm
in
the
f
e
a
tur
e
s
e
lec
ti
on
pr
oc
e
s
s
S
ha
pe
f
e
a
tu
r
e
s
T
e
xt
ur
e
f
e
a
tu
r
e
s
C
ombi
na
ti
on of
s
ha
pe
s
a
nd t
e
xt
ur
e
s
A
r
e
a
100
119
418
-
100
119
21
P
e
r
im
e
te
r
38.117
40.466
86.540
-
38.117
40.466
12.756
M
e
tr
ic
0.864913
0.913222
0.701378
-
0.864913
0.913222
1.6261
E
c
c
e
nt
r
ic
it
y
0.833286
0.847956
0.693967
-
0.833286
0.847956
N
/A
Y
1
1
1
1
1
1
1
1
1
C
ont
r
a
s
t
-
0.003256
0.002554
0.002527
0.003256
0.002554
0.0009
C
or
r
e
la
ti
on
-
0.627183
0.704434
0.849754
0.627183
0.704434
0.569403
E
ne
r
gy
-
0.99835
0.998167
0.995909
0.99835
0.998167
0.9199
H
omoge
ne
it
y
-
0.999722
0.999732
0.999661
0.999722
0.999732
0.9999
K
-
N
N
(
%
)
80.77
84.62
82.69
S
V
M
(
%
)
88.40
88.46
86.54
A
N
N
(
%
)
99.98
99.98
99.97
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
Ar
ti
f
I
ntell
I
S
S
N:
2252
-
8938
Optimiz
ing
the
gall
s
tone
de
tec
ti
on
pr
oc
e
s
s
w
it
h
featur
e
s
e
lec
ti
on
s
tat
is
ti
c
al
analys
is
algor
it
hm
(
M
us
li
Y
anto
)
1189
T
a
ble
2
.
T
he
pe
r
f
or
manc
e
tes
t
r
e
s
ult
s
of
the
F
S
S
A
a
lgor
it
hm
(
%
)
F
e
a
tu
r
e
s
K
-
N
N
a
c
c
ur
a
c
y
S
V
M
a
c
c
ur
a
c
y
A
N
N
a
c
c
ur
a
c
y
A
na
ly
s
is
(
P
C
)
T
r
a
in
in
g
T
e
s
ti
ng
T
ot
a
l
a
c
c
ur
a
c
y
T
r
a
in
in
g
T
e
s
ti
ng
T
ot
a
l
a
c
c
ur
a
c
y
T
r
a
in
in
g
T
e
s
ti
ng
T
ot
a
l
a
c
c
ur
a
c
y
S
ha
pe
f
e
a
tu
r
e
s
61.53
100
80.76
69.23
100
84.61
99.46
99.99
99.73
80.04
T
e
xt
ur
e
f
e
a
tu
r
e
s
69.23
100
84.61
80.76
100
90.38
99.93
99.97
99.95
95.90
C
o
mb
in
a
t
io
n
o
f
s
ha
p
e
s
a
n
d
te
xt
ur
e
s
65.38
100
82.69
69.23
100
84.62
99.90
99.94
99.92
87.96
T
a
ble
3
.
C
ompar
is
on
of
the
pe
r
f
or
manc
e
of
the
F
S
S
A
a
lgor
it
hm
in
im
p
r
oving
the
pr
oc
e
s
s
of
de
tec
ti
n
g
ga
ll
s
tone
objec
ts
with
pr
e
vious
r
e
s
e
a
r
c
h
No
P
r
e
vi
ous
r
e
s
e
a
r
c
h r
e
s
ul
ts
P
e
r
f
or
ma
nc
e
r
e
s
ul
ts
of
t
he
F
S
S
A
a
lg
or
it
hm
1
T
he
r
e
s
ul
ts
of
th
e
te
s
ts
th
a
t
ha
ve
be
e
n
c
a
r
r
ie
d
out
s
how
th
a
t
th
e
pe
r
f
or
ma
nc
e
a
c
c
ur
a
c
y
le
v
e
l
of
th
e
mul
ti
c
la
s
s
c
l
a
s
s
if
ic
a
ti
on
m
e
th
od
c
om
bi
ne
d
w
it
h
de
e
p
le
a
r
ni
ng r
e
a
c
he
s
93.4%
a
nd 94.36%
[
48]
.
P
r
e
s
e
nt
s
th
e
de
ve
lo
pme
nt
of
th
e
im
a
ge
e
xt
r
a
c
ti
on pr
oc
e
s
s
w
it
h t
he
F
S
S
A
f
e
a
tu
r
e
i
n
de
te
c
ti
ng
ga
ll
s
to
ne
obj
e
c
ts
.
P
e
r
f
or
ma
nc
e
te
s
ti
ng
of
F
S
S
A
pe
r
f
or
ma
nc
e
in
in
c
r
e
a
s
in
g
de
te
c
ti
on
pr
ovi
de
d
a
n
a
c
c
ur
a
c
y
r
a
te
of
95.83%
,
s
e
ns
it
iv
it
y
of
96.96
%
,
a
nd
s
pe
c
if
ic
it
y
of
95.23%
.
B
a
s
e
d
on
th
e
s
e
r
e
s
ul
ts
,
th
e
F
S
S
A
a
lg
or
it
hm
c
a
n
pr
e
s
e
nt
a
nove
lt
y
in
th
e
f
or
m
of
a
n
e
f
f
e
c
ti
ve
a
nd
e
f
f
ic
ie
nt
a
lg
or
it
hm
f
or
im
p
r
ovi
ng
th
e
ga
ll
s
to
ne
obj
e
c
t
de
te
c
ti
on pr
oc
e
s
s
.
2
T
he
C
N
N
c
la
s
s
if
ic
a
ti
on
me
th
od
w
it
h
f
e
a
tu
r
e
e
xt
r
a
c
ti
on
a
nd
f
e
a
tu
r
e
s
e
le
c
ti
on
in
th
e
de
te
c
ti
on
pr
oc
e
s
s
pr
ovi
de
s
out
put
w
it
h
a
n
a
c
c
ur
a
c
y
le
ve
l
of
0.6536%
a
nd
0.8942%
f
or
t
he
t
r
a
in
in
g a
nd t
e
s
ti
ng [
38]
.
3
T
he
M
S
A
a
lg
or
it
hm
us
e
s
c
o
a
r
s
e
n
e
twor
k
s
e
gm
e
nt
a
ti
on
w
it
h
s
h
a
pe
ope
r
a
ti
ons
in
obj
e
c
t
de
te
c
ti
on pr
ovi
di
ng a
de
te
c
ti
on a
c
c
ur
a
c
y r
a
te
of
90%
[
49]
.
4
M
S
A
a
lg
or
it
hm
de
ve
lo
pme
nt
w
a
s
c
a
r
r
ie
d
out
u
s
in
g
th
e
E
nha
nc
e
me
nt
te
c
hni
que
us
in
g
th
e
f
e
a
tu
r
e
im
pr
ove
me
nt
pyr
a
mi
d
ne
t
w
or
k
(
mul
ti
-
s
ta
ge
F
E
P
N
)
pr
ovi
di
ng
a
n a
c
c
ur
a
c
y r
a
te
of
92.1%
[
50]
.
4.
CONC
L
USI
ON
I
mpr
oving
the
ga
ll
s
tone
objec
t
de
tec
ti
on
pr
oc
e
s
s
by
de
ve
lopi
ng
f
e
a
tur
e
s
e
lec
ti
on
with
the
F
S
S
A
a
lgor
it
hm
ha
s
pr
ovided
maximu
m
output
r
e
s
ult
s
.
T
he
s
e
r
e
s
ult
s
a
r
e
ba
s
e
d
on
the
r
e
s
ult
s
of
F
S
S
A
pe
r
f
or
manc
e
tes
ti
ng
with
a
n
a
c
c
ur
a
c
y
r
a
te
of
95.
83
%
,
s
e
ns
it
ivi
ty
of
96
.
96%
,
a
nd
s
pe
c
if
icity
of
95
.
23%
.
B
a
s
e
d
on
thes
e
r
e
s
ult
s
,
it
c
a
n
be
p
r
ove
n
that
the
F
S
S
A
a
lgor
it
h
m
c
a
n
pr
ovide
opti
mal
r
e
s
ult
s
in
the
pr
oc
e
s
s
o
f
de
tec
ti
ng
ga
ll
s
tone
objec
ts
.
T
he
ove
r
a
ll
pe
r
f
o
r
manc
e
of
th
e
F
S
S
A
a
lgor
it
hm
c
a
n
be
us
e
d
a
s
a
nove
lt
y
in
e
xtr
a
c
ti
on
tec
hniques
,
e
s
pe
c
ially
in
the
f
e
a
tur
e
s
e
lec
ti
on
pr
oc
e
s
s
.
AC
KNOWL
E
DGE
M
E
NT
S
T
he
a
uthor
would
li
ke
to
e
xpr
e
s
s
his
g
r
a
ti
tude
to
Dr
.
Hj.
Z
e
r
ni
M
e
lm
us
i,
S
E
,
M
M
,
Ak
,
C
A,
a
s
the
C
ha
ir
pe
r
s
on
of
the
Y
P
T
K
P
a
da
ng
C
omput
e
r
C
oll
e
ge
F
ounda
ti
on
who
ha
s
p
r
ovided
s
uppor
t
in
thi
s
r
e
s
e
a
r
c
h
ba
s
e
d
on
letter
no.
004/P
S
D
T
I
/UP
I
YPT
K/S
B
/V/20
23,
r
e
ga
r
ding
the
a
ppoint
ment
of
e
xpe
r
ts
in
thi
s
r
e
s
e
a
r
c
h.
RE
F
E
RE
NC
E
S
[
1]
D
.
A
.
G
ol
d
ma
n,
“
G
a
ll
bl
a
dde
r
,
ga
ll
s
to
ne
s
,
a
nd
di
s
e
a
s
e
s
of
th
e
ga
ll
bl
a
dde
r
in
c
hi
ld
r
e
n,”
P
e
di
at
r
ic
s
in
R
e
v
ie
w
,
vol
.
41,
no.
12,
pp. 623
–
629, De
c
. 2020
, doi
:
10.1542/pi
r
.2019
-
0077.
[
2]
A
.
M
.
M
a
dde
n,
D
.
T
r
iv
e
di
,
N
.
C
.
S
me
e
to
n,
a
nd
A
.
C
ul
ki
n,
“
M
od
if
ie
d
di
e
ta
r
y
f
a
t
in
ta
ke
f
o
r
tr
e
a
tm
e
nt
of
ga
ll
s
to
ne
di
s
e
a
s
e
,”
C
oc
h
r
ane
D
at
abas
e
of
Sy
s
te
m
at
i
c
R
e
v
ie
w
s
, vol
. 2021, no. 6,
J
un. 2021, do
i:
10.1002/14651858.
C
D
012608
.pub2.
[
3]
P
.
C
ia
nc
i
a
nd
E
.
R
e
s
ti
ni
,
“
M
a
na
ge
me
nt
of
c
hol
e
li
th
ia
s
i
s
w
it
h
c
hol
e
doc
hol
it
hi
a
s
i
s
:
e
ndo
s
c
opi
c
a
nd
s
ur
gi
c
a
l
a
ppr
oa
c
he
s
,”
W
or
ld
J
our
nal
of
G
as
tr
o
e
nt
e
r
ol
ogy
, vol
. 27, no. 28, pp. 4536
–
4554,
J
ul
. 2021, doi:
10.3748/wjg.v27.i
28.4536.
[
4]
L
.
P
a
n
e
t
al
.
,
“
T
he
tr
e
a
tm
e
nt
of
c
hol
e
c
ys
ti
ti
s
a
nd
c
hol
e
li
th
ia
s
is
by
ti
be
ta
n
me
di
c
in
e
,”
E
v
id
e
nc
e
-
bas
e
d
C
om
pl
e
m
e
nt
a
r
y
and
A
lt
e
r
nat
iv
e
M
e
di
c
in
e
, vol
. 2021, pp. 1
–
21, S
e
p. 2021, doi:
10.1155/2021/
9502609.
[
5]
J
.
W
a
ng,
H
.
Z
hu,
S
.
H
.
W
a
ng,
a
nd
Y
.
D
.
Z
ha
ng,
“
A
r
e
vi
e
w
o
f
de
e
p
le
a
r
ni
ng
on
me
di
c
a
l
im
a
ge
a
n
a
ly
s
is
,
”
M
obi
le
N
e
tw
o
r
k
s
and
A
ppl
ic
at
io
ns
, vol
. 26, no. 1, pp. 351
–
380, 2021, doi:
10.1007/s
11036
-
020
-
01672
-
7.
[
6]
S
.
S
uga
nya
de
vi
,
V
.
S
e
e
th
a
la
k
s
hmi
,
a
nd
K
.
B
a
la
s
a
my
,
“
A
r
e
vi
e
w
on
de
e
p
le
a
r
ni
ng
in
me
di
c
a
l
im
a
g
e
a
n
a
ly
s
is
,”
I
nt
e
r
nat
io
nal
J
our
nal
of
M
ul
ti
m
e
di
a I
nf
or
m
at
io
n R
e
tr
ie
v
al
, vol
. 11, no. 1, pp. 19
–
38, M
a
r
. 2022, doi:
10.1007/s
13735
-
021
-
00218
-
1.
[
7]
X
.
L
iu
,
L
.
S
ong,
S
.
L
iu
,
a
nd
Y
.
Z
ha
ng,
“
A
r
e
vi
e
w
of
de
e
p
-
le
a
r
ni
ng
-
ba
s
e
d
me
di
c
a
l
im
a
ge
s
e
gme
nt
a
ti
on
me
th
ods
,”
Su
s
ta
in
abi
li
ty
,
vol
. 13, no. 3, J
a
n. 2021, doi:
10.3390/s
u13031224.
[
8]
W
.
Z
hou,
H
.
W
a
ng,
a
nd
Z
.
W
a
n,
“
O
r
e
im
a
ge
c
la
s
s
if
ic
a
ti
on
ba
s
e
d
on
im
pr
ove
d
C
N
N
,”
C
om
put
e
r
s
and
E
le
c
tr
ic
al
E
ngi
ne
e
r
in
g
,
vol
. 99, Apr
. 2022, d
oi
:
10.1016/j
.c
ompe
le
c
e
ng.2022.107819.
[
9]
M
.
A
.
S
.
A
i
e
t
al
.
,
“
R
e
a
l
-
ti
me
f
a
c
e
ma
s
k
de
t
e
c
ti
on
f
or
pr
e
ve
nt
in
g
c
ovi
d
-
19
s
pr
e
a
d
us
in
g
tr
a
ns
f
e
r
le
a
r
ni
ng
ba
s
e
d
de
e
p
ne
ur
a
l
n
e
two
r
k,”
E
le
c
tr
oni
c
s
, vol
. 11, no. 14, J
ul
. 2022, doi:
10.3390/ele
c
tr
oni
c
s
11142250.
[
10]
L
.
W
e
i,
S
.
L
.
Z
ha
ng,
N
.
X
ie
,
C
.
M
.
L
i,
a
nd
B
.
M
.
F
u,
“
T
he
va
lu
e
of
M
R
C
P
in
c
hi
ld
r
e
n
w
it
h
bi
li
a
r
y
s
ympt
oma
to
lo
gy
–
a
n
e
s
s
e
nt
ia
l
a
dj
unc
t
f
or
s
a
f
e
c
hol
e
c
y
s
te
c
to
my,”
Sout
h
A
fr
ic
an
J
ou
r
nal
of
Sur
ge
r
y
,
vol
.
60,
no.
1,
pp.
67
–
69,
2022,
doi
:
10.17159/20
78
-
5151/2022/
v60n1a
3430.
[
11]
A
.
J
a
va
id
,
R
.
M
a
hmood,
H
.
U
la
h,
M
.
S
ha
f
iq
,
N
.
D
il
da
r
,
a
nd
G
.
A
bba
s
,
“
D
ia
gnos
ti
c
a
c
c
ur
a
c
y
of
ma
gne
ti
c
r
e
s
ona
nc
e
c
hol
a
ngi
opa
nc
r
e
a
to
gr
a
phy in t
he
de
te
c
ti
on of
c
hol
e
doc
hol
it
h, t
a
ki
ng pos
t
-
ope
r
a
ti
ve
f
in
di
ngs
a
s
t
he
gol
d
s
ta
nda
r
d,”
P
ak
is
ta
n A
r
m
e
d
F
or
c
e
s
M
e
di
c
al
J
ou
r
nal
, vol
. 73, no. 2, pp. 394
–
397, Apr
. 2023, doi:
10.51253/paf
mj
.v73i2.7015.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
Ar
ti
f
I
ntell
,
Vol.
14
,
No.
2
,
Apr
il
20
25
:
1183
-
119
1
1190
[
12]
M
.
H
ig
a
s
hi
,
M
.
T
a
na
b
e
,
K
.
I
ha
r
a
,
E
.
I
id
a
,
M
.
F
ur
uka
w
a
,
a
nd
K
.
I
to
,
“
B
il
e
f
lo
w
dyna
mi
c
s
in
pa
ti
e
nt
s
w
it
h
c
hol
e
li
th
ia
s
is
:
a
n
e
v
a
lu
a
ti
on
w
it
h
c
in
e
-
dyna
mi
c
m
a
gne
ti
c
r
e
s
ona
nc
e
c
hol
a
ngi
op
a
nc
r
e
a
to
gr
a
phy
us
in
g
a
s
pa
ti
a
ll
y
s
e
le
c
ti
ve
in
ve
r
s
io
n
-
r
e
c
ove
r
y
pul
s
e
,”
T
om
ogr
aphy
, vol
. 8, no. 2, pp. 815
–
823, M
a
r
. 2022, doi:
10.33
90/
to
mogr
a
phy8020067.
[
13]
S
.
R
a
jp
ut
,
A
.
S
in
gh,
a
nd
R
.
G
upt
a
,
“
A
ut
oma
te
d
ki
dne
y
s
t
one
de
te
c
ti
on
u
s
in
g
im
a
ge
pr
oc
e
s
s
in
g
te
c
hni
que
s
,”
in
2021
9t
h
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
R
e
li
abi
li
ty
,
I
nf
oc
om
T
e
c
hnol
ogi
e
s
and
O
pt
imi
z
at
io
n
(
T
r
e
nds
and
F
ut
ur
e
D
ir
e
c
ti
ons
)
,
I
C
R
I
T
O
20
21
,
I
E
E
E
, S
e
p. 2021, pp. 1
–
5
, doi
:
10.1109/I
C
R
I
T
O
51393.2021.9596175.
[
14]
E
.
C
.
Y
e
pe
z
,
L
.
M
.
L
.
C
a
r
r
il
lo
,
G
.
B
ot
e
ll
a
,
a
nd
I
.
A
me
r
in
i,
“
I
nt
r
oduc
ti
on
to
th
e
s
pe
c
ia
l
s
e
c
ti
on
on
im
a
ge
pr
oc
e
s
s
in
g
in
s
e
c
ur
it
y
a
ppl
ic
a
ti
ons
(
V
S
I
-
I
P
S
A
)
,”
C
om
put
e
r
s
and E
le
c
tr
ic
al
E
ngi
ne
e
r
i
ng
, vol
. 89, 2021, doi:
10.1016/j
.c
ompe
le
c
e
ng.2020.106935.
[
15]
T
.
D
.
S
ha
s
hi
ka
la
,
B
.
L
.
S
uni
th
a
,
S
.
B
a
s
a
va
r
a
ja
ppa
,
a
nd
J
.
P
.
D
a
vi
m,
“
S
ome
s
tu
di
e
s
on
me
a
s
ur
e
me
nt
of
w
or
n
s
ur
f
a
c
e
by
di
gi
ta
l
im
a
ge
pr
oc
e
s
s
in
g,”
I
nt
e
r
nat
io
nal
J
ou
r
nal
of
I
m
age
and Gr
aphi
c
s
, vol
.
23, no. 2, M
a
r
. 2023, doi:
10.1142/S
021946782350016X
.
[
16]
J
. G
r
ühn, M
. V
oge
l,
a
n
d N
.
K
oc
kma
nn, “
D
ig
it
a
l
im
a
ge
pr
oc
e
s
s
i
ng of
ga
s
‐
li
qui
d r
e
a
c
ti
ons
i
n
c
oi
le
d c
a
pi
ll
a
r
ie
s
,
”
C
he
m
ie
-
I
nge
ni
e
ur
-
T
e
c
hni
k
, vol
. 93, no. 5, pp. 825
–
829, M
a
y 2021, doi:
10.1002/ci
te
.202000240.
[
17]
M
.
B
a
kt
ha
va
tc
h
a
la
m
e
t
al
.
,
“
M
or
phol
ogi
c
a
l
a
nd
e
le
me
nt
a
l
ma
ppi
ng
of
ga
ll
s
to
ne
s
us
in
g
s
ync
hr
ot
r
on
mi
c
r
ot
omogr
a
phy
a
nd
s
ync
hr
ot
r
on x‐
r
a
y f
lu
or
e
s
c
e
nc
e
s
pe
c
tr
os
c
opy,”
J
G
H
O
p
e
n
, vol
.
3, no. 5, pp. 381
–
387, 2019, doi:
10.1002/j
gh3.12171.
[
18]
I
.
R
.
I
.
H
a
que
a
nd
J
.
N
e
ube
r
t,
“
D
e
e
p
le
a
r
ni
ng
a
ppr
oa
c
he
s
to
bi
ome
di
c
a
l
im
a
ge
s
e
gme
nt
a
ti
on,”
I
nf
or
m
at
ic
s
in
M
e
di
c
in
e
U
nl
oc
k
e
d
,
vol
. 18, 2020, doi:
10.1016/j
.i
mu.2020.100297.
[
19]
M
.
J
i
a
nd
Z
.
W
u,
“
A
ut
oma
ti
c
de
t
e
c
ti
on
a
nd
s
e
ve
r
it
y
a
na
ly
s
is
of
gr
a
pe
bl
a
c
k
me
a
s
le
s
di
s
e
a
s
e
ba
s
e
d
on
de
e
p
le
a
r
ni
ng
a
nd
f
u
z
z
y
lo
gi
c
,”
C
om
put
e
r
s
and E
le
c
tr
oni
c
s
i
n A
gr
ic
ul
tu
r
e
, vol
. 193, F
e
b
. 2022, doi:
10.1016/j
.c
ompa
g.2022.106718.
[
20]
K
.
B
us
c
hb
a
c
he
r
,
D
.
A
hr
e
ns
, M
.
E
s
pe
la
nd,
a
nd
V
.
S
te
in
ha
ge
, “
I
ma
ge
-
ba
s
e
d s
pe
c
ie
s
id
e
nt
if
ic
a
ti
on
of
w
i
ld
b
e
e
s
us
in
g
c
onvolut
i
ona
l
ne
ur
a
l
ne
twor
ks
,”
E
c
ol
ogi
c
al
I
nf
or
m
at
ic
s
, vol
. 55, J
a
n. 2020, d
oi
:
10.1016/j
.e
c
oi
nf
.2019.101017.
[
21]
A
.
R
.
J
a
v
e
d
a
nd
Z
.
J
a
li
l,
“
B
yt
e
-
le
ve
l
obj
e
c
t
id
e
nt
if
ic
a
ti
on
f
or
f
or
e
ns
ic
in
ve
s
ti
ga
ti
on
of
di
gi
ta
l
im
a
ge
s
,
”
in
1s
t
A
nnual
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
C
y
be
r
W
ar
fa
r
e
and
Se
c
u
r
it
y
,
I
C
C
W
S
2020
-
P
r
oc
e
e
di
ngs
,
I
E
E
E
,
O
c
t.
2020,
pp.
1
–
4
,
doi
:
10.1109/I
C
C
W
S
48432.2020.9292387.
[
22]
S
.
T
a
va
kol
i,
A
.
G
ha
f
f
a
r
i,
Z
.
M
.
K
ouz
e
hka
n
a
n,
a
nd
R
.
H
o
s
s
e
in
i,
“
N
e
w
s
e
gme
nt
a
ti
on
a
nd
f
e
a
tu
r
e
e
xt
r
a
c
ti
on
a
lg
or
it
hm
f
or
c
la
s
s
if
ic
a
ti
on
of
w
hi
te
bl
ood
c
e
ll
s
in
pe
r
ip
he
r
a
l
s
me
a
r
im
a
ge
s
,”
Sc
ie
nt
if
ic
R
e
por
ts
,
vol
.
11,
no.
1,
S
e
p.
2021,
doi
:
10.1038/
s
415
98
-
021
-
98599
-
0.
[
23]
X
.
C
he
n
e
t
al
.
,
“
D
iv
e
r
s
e
da
t
a
a
ugme
nt
a
ti
on
f
or
le
a
r
ni
ng
im
a
g
e
s
e
gme
nt
a
ti
on
w
it
h
c
r
os
s
-
moda
li
ty
a
nnot
a
ti
ons
,”
M
e
di
c
al
I
m
age
A
nal
y
s
is
, vol
. 71, J
ul
. 2021, doi:
10.1016/j
.me
di
a
.2021.102060.
[
24]
X
. L
e
i,
X
. Y
u, J
. C
hi
, Y
.
W
a
ng,
J
. Z
ha
ng, a
nd C
.
W
u, “
B
r
a
in
t
u
mor
s
e
gme
nt
a
ti
on i
n mr
i
ma
ge
s
us
in
g
a
s
pa
r
s
e
c
on
s
tr
a
in
e
d l
e
ve
l
s
e
t
a
lg
or
it
hm,”
E
x
pe
r
t
Sy
s
te
m
s
w
it
h A
ppl
ic
at
io
ns
, vol
. 168, Apr
. 2021, doi:
10.1016/j
.e
s
w
a
.2020.114262.
[
25]
S
.
P
a
ng
e
t
al
.
,
“
A
nove
l
yol
ov3
-
a
r
c
h
mode
l
f
o
r
id
e
nt
if
yi
ng
c
h
ol
e
li
th
ia
s
is
a
nd
c
la
s
s
if
yi
ng
ga
ll
s
to
ne
s
on
C
T
im
a
ge
s
,
”
P
L
oS
O
N
E
,
vol
. 14, no. 6, J
un. 2019, doi:
10.1371/j
our
na
l.
pone
.0217647
.
[
26]
D
.
L
i,
B
.
D
u,
Y
.
S
he
n,
a
nd
L
.
G
e
,
“
A
r
ti
f
ic
ia
l
in
te
ll
ig
e
nc
e
‐
a
s
s
i
s
te
d
vi
s
ua
l
s
e
n
s
in
g
te
c
hnol
ogy
unde
r
duode
no
s
c
opy
of
ga
ll
bl
a
dd
e
r
s
to
ne
s
,”
J
ou
r
nal
of
Se
ns
o
r
s
, vol
. 2021, no. 1, 2021, doi
:
10.115
5/
2021/5158577.
[
27]
A
.
G
.
M
ol
s
os
a
,
H
.
A
.
O
r
e
ngo,
D
.
L
a
w
r
e
nc
e
,
G
.
P
hi
li
p,
K
.
H
op
pe
r
,
a
nd
C
.
A
.
P
e
tr
ie
,
“
P
ot
e
nt
ia
l
of
de
e
p
le
a
r
ni
ng
s
e
gme
nt
a
ti
on
f
or
th
e
e
xt
r
a
c
ti
on
of
a
r
c
ha
e
ol
ogi
c
a
l
f
e
a
tu
r
e
s
f
r
om
h
is
to
r
ic
a
l
ma
p
s
e
r
ie
s
,”
A
r
c
hae
ol
ogi
c
al
P
r
os
pe
c
ti
on
,
vol
.
28,
no.
2,
pp.
187
–
199, Apr
.
2021, doi:
10.1002/ar
p.1807.
[
28
]
H
.
W
a
ng,
W
.
Z
ha
ng,
S
.
L
i,
Y
.
F
a
n,
M
.
F
e
ng,
a
nd
R
.
W
a
ng,
“
D
e
ve
lo
pme
nt
a
nd
e
va
lu
a
ti
on
of
d
e
e
p
le
a
r
ni
ng
-
ba
s
e
d
a
ut
oma
te
d
s
e
gme
nt
a
ti
on
of
pi
tu
it
a
r
y
a
de
nom
a
in
c
li
ni
c
a
l
ta
s
k,
”
T
he
J
o
ur
nal
of
C
li
ni
c
al
E
ndoc
r
in
ol
ogy
&
M
e
ta
bol
is
m
,
vol
.
106,
n
o.
9,
pp. 2535
–
2546,
A
ug. 2021, doi:
10.1210/clinem/dga
b371.
[
29]
C
.
P
.
J
a
ya
pa
ndi
a
n
e
t
al
.
,
“
D
e
ve
lo
pme
nt
a
nd
e
va
lu
a
ti
on
of
de
e
p l
e
a
r
ni
ng
–
ba
s
e
d
s
e
gme
nt
a
ti
on
of
hi
s
to
lo
gi
c
s
tr
uc
tu
r
e
s
in
th
e
ki
d
ne
y
c
or
te
x
w
it
h
mul
ti
pl
e
hi
s
to
lo
gi
c
s
ta
in
s
,”
K
id
ne
y
I
nt
e
r
nat
io
nal
,
vo
l.
99,
no.
1,
pp
.
86
–
101,
J
a
n.
2021,
doi
:
10.1016/j
.ki
nt
.2020.07.044.
[
30]
S
.
L
a
th
a
,
P
.
M
ut
hu,
S
.
D
ha
na
la
ks
hmi
,
R
.
K
uma
r
,
K
.
W
.
L
a
i,
a
nd
X
.
W
u,
“
E
me
r
gi
ng
f
e
a
tu
r
e
e
xt
r
a
c
ti
on
te
c
hni
qu
e
s
f
or
ma
c
hi
ne
le
a
r
ni
ng
-
ba
s
e
d c
la
s
s
if
ic
a
ti
on of
c
a
r
ot
id
a
r
te
r
y ul
tr
a
s
ound im
a
ge
s
,”
C
om
put
at
io
nal
I
nt
e
ll
ig
e
nc
e
and
N
e
ur
o
s
c
ie
nc
e
, vol
. 2022, pp
. 1
–
14, M
a
y 2022, doi:
10.1155/2022/
1847981.
[
31]
G
.
L
a
ti
f
,
“
D
e
e
pT
umor
:
f
r
a
me
w
or
k
f
or
br
a
in
mr
im
a
ge
c
la
s
s
if
ic
a
ti
on,
s
e
gme
nt
a
ti
on
a
nd
tu
mor
de
te
c
ti
on,”
D
ia
gnos
ti
c
s
,
vol
.
12,
no. 11, Nov. 2022, d
oi
:
10.3390
/d
ia
gnos
ti
c
s
12112888.
[
32]
R
.
B
yg
a
r
i
,
K
.
R
it
h
e
s
h
,
S
.
A
m
b
e
s
a
ng
e
,
a
nd
S
.
G
.
K
oo
l
a
g
udi
,
“
P
r
o
s
t
a
t
e
c
a
n
c
e
r
gr
a
di
n
g
u
s
in
g
m
ul
t
i
s
t
a
g
e
d
e
e
p
n
e
ur
a
l
n
e
t
w
or
k
s
,”
i
n
M
ac
hi
n
e
L
e
ar
ni
ng
,
I
m
ag
e
P
r
o
c
e
s
s
in
g,
N
e
tw
o
r
k
Se
c
u
r
it
y
a
nd
D
at
a
Sc
ie
n
c
e
s
,
pp
.
27
1
–
28
3,
2
02
3,
d
oi
:
10
.1
00
7/
9
78
-
9
81
-
19
-
5
86
8
-
7
_2
1
.
[
33]
T
.
S
.
P
r
i
y
a
,
“
R
e
s
n
e
t
ba
s
e
d f
e
a
tu
r
e
e
xt
r
a
c
ti
on
w
it
h
d
e
c
i
s
i
on
tr
e
e
c
la
s
s
if
ie
r
f
or
c
la
s
s
if
i
c
a
t
on
of
m
a
mm
o
gr
a
m
im
a
g
e
s
,”
T
u
r
k
i
s
h
J
o
ur
na
l
o
f
C
o
m
p
ut
e
r
a
nd
M
a
th
e
m
at
ic
s
E
d
u
c
a
ti
o
n
(
T
U
R
C
O
M
A
T
)
,
v
ol
.
12
,
no
.
2
,
pp
.
11
47
–
1
15
3,
A
pr
.
2
02
1,
d
oi
:
10
.1
77
62/
tu
r
c
om
a
t
.v
12
i2
.1
1
36.
[
34]
Y
.
K
.
G
a
ne
e
va
a
nd
E
.
V
.
M
ya
s
ni
kov,
“
I
de
nt
if
yi
ng
pe
r
s
ons
f
r
om
ir
is
im
a
ge
s
us
in
g
ne
ur
a
l
n
e
twor
ks
f
or
im
a
ge
s
e
gm
e
nt
a
ti
on
a
n
d
f
e
a
tu
r
e
e
xt
r
a
c
ti
on,”
C
om
put
e
r
O
pt
ic
s
, vol
. 46, no. 2, pp. 308
–
3
16, 2022, doi:
10.18287/2412
-
6179
-
CO
-
1023.
[
35]
A
.
B
.
H
a
bi
b,
“
E
lb
ow
me
th
od
vs
s
il
houe
tt
e
c
o
-
e
f
f
ic
ie
nt
in
de
te
r
mi
ni
ng
th
e
numbe
r
of
c
lu
s
te
r
s
,”
R
e
s
e
ar
c
h
G
at
e
,
pp.
1
-
7,
2021
,
doi
:
10.13140/R
G
.2.2.27982.79688
.
[
36]
S
.
S
.
Y
u,
S
.
W
.
C
hu,
C
.
M
.
W
a
ng,
Y
.
K
.
C
ha
n,
a
nd
T
.
C
.
C
h
a
ng,
“
T
w
o
im
pr
ove
d
k
-
me
a
ns
a
lg
or
it
hms
,”
A
ppl
ie
d
So
ft
C
om
pu
ti
ng
J
our
nal
, vol
. 68, pp. 747
–
755, J
ul
. 2018, doi:
10.1016/j
.a
s
oc
.20
17.08.032.
[
37]
M
. C
ui
, “
I
nt
r
oduc
ti
on t
o t
he
k
-
me
a
ns
c
lu
s
te
r
in
g a
lg
or
it
hm ba
s
e
d on the
e
lb
ow
me
th
od,”
A
c
c
ount
in
g,
A
udi
ti
ng and
F
in
anc
e
, vo
l.
1,
pp. 5
–
8, 2020.
[
38]
B
.
S
.
R
iz
a
,
J
.
N
a
’
a
m
,
a
nd
S
.
S
u
mi
j
a
n,
“
C
on
vo
lu
ti
on
a
l
ne
ur
a
l
n
e
t
w
or
k
a
s
a
n
im
a
g
e
pr
o
c
e
s
s
in
g
t
e
c
hn
iq
ue
f
or
c
l
a
s
s
if
ic
a
t
io
n
of
b
a
c
i
ll
i
t
ub
e
r
c
ul
o
s
i
s
e
x
tr
a
pu
lm
on
a
r
y
(
T
B
E
P
)
di
s
e
a
s
e
,”
T
E
M
J
o
ur
na
l
,
v
ol
.
11
,
no
.
3,
p
p.
1
33
1
–
13
40
,
A
u
g.
2
02
2,
d
oi
:
1
0.
18
42
1/
T
E
M
11
3
-
43
.
[
39]
H
.
S
a
le
m,
M
.
Y
.
S
ha
m
s
,
O
.
M
.
E
lz
e
ki
,
M
.
A
.
E
lf
a
tt
a
h,
J
.
F
.
A
l‐
a
mr
i,
a
nd
S
.
E
ln
a
z
e
r
,
“
F
in
e
-
tu
ni
ng
f
uz
z
y
knn
c
la
s
s
if
ie
r
ba
s
e
d
on
unc
e
r
ta
in
ty
me
mbe
r
s
hi
p
f
or
th
e
me
di
c
a
l
di
a
gnos
is
of
di
a
be
te
s
,”
A
ppl
ie
d
Sc
ie
nc
e
s
,
vol
.
12,
no.
3,
J
a
n.
20
22,
doi
:
10.3390/a
pp12030950.
[
40]
T
.
S
.
R
e
ddy
a
nd
J
.
H
a
r
ik
ir
a
n,
“
H
yp
e
r
s
pe
c
tr
a
l
im
a
ge
c
la
s
s
if
ic
a
ti
on
us
in
g
s
uppor
t
ve
c
to
r
ma
c
hi
ne
s
,
”
I
A
E
S
I
nt
e
r
nat
io
nal
J
our
na
l
of
A
r
ti
fi
c
ia
l
I
nt
e
ll
ig
e
nc
e
, vol
. 9, no. 4, pp. 684
–
690, 2020, doi:
10.
11591/i
ja
i.
v9.i
4.pp684
-
690
.
[
41]
G
.
S
hoba
na
a
nd
N
.
P
r
iy
a
,
“
C
a
nc
e
r
dr
ug
c
la
s
s
if
ic
a
ti
on
us
in
g
a
r
ti
f
ic
ia
l
ne
ur
a
l
ne
twor
k
w
it
h
f
e
a
tu
r
e
s
e
le
c
ti
on,”
in
P
r
o
c
e
e
di
ngs
of
t
he
3r
d
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
I
nt
e
ll
ig
e
nt
C
om
m
uni
c
at
io
n
T
e
c
hnol
ogi
e
s
and
V
ir
tu
al
M
obi
le
N
e
tw
or
k
s
,
I
C
I
C
V
2021
,
I
E
E
E
,
F
e
b.
2021, pp. 1250
–
1255
, doi
:
10.1109/I
C
I
C
V
50876.2021.9388542.
[
42]
M
.
Y
a
nt
o,
S
.
S
a
nj
a
ya
,
Y
ul
a
s
mi
,
D
.
G
us
w
a
ndi
,
a
nd
S
.
A
r
li
s
,
“
I
mpl
e
me
nt
a
ti
on
mul
ti
pl
e
li
ne
a
r
r
e
gr
e
s
io
n
in
ne
ur
a
l
ne
twor
k
pr
e
di
c
t
gol
d
pr
ic
e
,”
I
ndone
s
ia
n
J
our
nal
of
E
le
c
tr
ic
al
E
ngi
ne
e
r
in
g
an
d
C
om
put
e
r
Sc
ie
nc
e
,
vol
.
22,
no.
3,
pp.
1635
–
1642,
J
un.
2021,
doi
:
10.11591/i
je
e
c
s
.v22.i3.pp1635
-
1642.
[
43]
H
.
P
a
n,
X
.
Y
ou,
S
.
L
iu
,
a
nd
D
.
Z
ha
ng,
“
P
e
a
r
s
on
c
or
r
e
la
ti
on
c
o
e
f
f
ic
ie
nt
-
ba
s
e
d
phe
r
omone
r
e
f
a
c
to
r
in
g
me
c
ha
ni
s
m
f
or
mul
ti
-
c
ol
ony
a
nt
c
ol
ony opti
mi
z
a
ti
o
n,”
A
ppl
ie
d I
nt
e
ll
ig
e
nc
e
, vol
. 51, no. 2, pp. 752
–
774, F
e
b. 2021, doi:
10.1007/s
10489
-
020
-
01841
-
x.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
Ar
ti
f
I
ntell
I
S
S
N:
2252
-
8938
Optimiz
ing
the
gall
s
tone
de
tec
ti
on
pr
oc
e
s
s
w
it
h
featur
e
s
e
lec
ti
on
s
tat
is
ti
c
al
analys
is
algor
it
hm
(
M
us
li
Y
anto
)
1191
[
44]
P
.
W
a
ld
ma
nn,
“
O
n
th
e
u
s
e
of
th
e
pe
a
r
s
on
c
or
r
e
la
ti
on
c
oe
f
f
ic
ie
nt
f
or
mode
l
e
va
lu
a
ti
on
in
ge
nome
-
w
id
e
pr
e
di
c
ti
on,”
F
r
ont
ie
r
s
in
G
e
ne
ti
c
s
, vol
. 10, S
e
p. 2019, doi:
10.3389/f
ge
ne
.2019.00899.
[
45]
J
.
C
a
i,
M
.
H
.
Z
ha
ng,
Y
.
T
.
Z
hu,
a
nd
Y
.
H
.
L
iu
,
“
M
od
e
l
of
f
r
e
ig
ht
ve
hi
c
le
e
ne
r
gy
c
ons
umpt
io
n
ba
s
e
d
on
p
e
a
r
s
on
c
or
r
e
la
ti
on
c
oe
f
f
ic
ie
nt
,”
J
ia
ot
ong
Y
uns
hu
X
it
ong
G
ongc
he
ng
Y
u
X
in
x
i/
J
our
nal
of
T
r
ans
por
ta
ti
on
Sy
s
te
m
s
E
ngi
ne
e
r
in
g
and
I
nf
or
m
at
io
n
T
e
c
hnol
ogy
, vol
. 18, no. 5, pp. 241
–
246, 2018, doi:
10.16097/j
.c
nki
.1009
-
6744.201
8.05.035.
[
46]
H
.
Z
hu,
X
.
Y
ou,
a
nd
S
.
L
iu
,
“
M
ul
ti
pl
e
a
nt
c
ol
ony
opt
im
iz
a
ti
o
n
ba
s
e
d
on
pe
a
r
s
on
c
or
r
e
la
ti
on
c
oe
f
f
ic
ie
nt
,”
I
E
E
E
A
c
c
e
s
s
,
vol
.
7,
pp. 61628
–
61638, 2019, doi:
10.1109/AC
C
E
S
S
.2019.2915673.
[
47]
A
.
B
a
hr
a
mi
,
A
.
K
a
r
im
ia
n,
E
.
F
a
te
mi
z
a
de
h,
H
.
A
r
a
bi
,
a
nd
H
. Z
a
i
di
,
“
A
ne
w
de
e
p
c
onvolut
io
na
l
ne
ur
a
l
ne
twor
k
de
s
ig
n
w
it
h
e
f
f
ic
ie
nt
le
a
r
ni
ng
c
a
pa
bi
li
ty
:
a
ppl
ic
a
ti
on
to
c
t
im
a
ge
s
ynt
he
s
is
f
r
om
mr
i,
”
M
e
di
c
al
P
hy
s
ic
s
,
vol
.
47,
no.
10,
pp.
5158
–
5171,
O
c
t.
2020,
doi
:
10.1002/m
p.14418.
[
48]
Z
.
W
a
ng,
E
.
W
a
ng,
a
nd
Y
.
Z
hu,
“
I
ma
ge
s
e
gme
nt
a
ti
on
e
va
lu
a
ti
on:
a
s
ur
ve
y
of
me
th
ods
,”
A
r
ti
fi
c
ia
l
I
nt
e
ll
ig
e
nc
e
R
e
v
ie
w
,
vol
.
53,
no. 8, pp. 5637
–
5674, De
c
. 2020, doi:
10.1007/s
10462
-
020
-
09830
-
9.
[
49]
T
.
A
li
e
t
al
.
,
“
M
ul
ti
s
ta
ge
s
e
gm
e
nt
a
ti
on
of
pr
os
ta
te
c
a
nc
e
r
ti
s
s
u
e
s
us
in
g
s
a
mpl
e
e
nt
r
opy
te
xt
ur
e
a
na
ly
s
i
s
,”
E
nt
r
opy
,
vol
.
22,
no.
12,
2020, doi:
10.3390/e22121370
.
[
50]
K
.
Z
ha
ng
a
nd
H
.
S
he
n,
“
M
ul
ti
-
s
ta
ge
f
e
a
tu
r
e
e
nha
nc
e
me
nt
pyr
a
mi
d
ne
twor
k
f
or
de
te
c
ti
ng
obj
e
c
ts
in
opt
ic
a
l
r
e
mot
e
s
e
ns
in
g
im
a
ge
s
,”
R
e
m
ot
e
Se
ns
in
g
, vol
. 14, no. 3, 2022, doi
:
10.3390/r
s
14030579.
B
I
OG
RA
P
HI
E
S
OF
AU
T
HO
RS
M
us
l
i
Y
a
nto
w
as
b
o
rn
i
n
J
ak
ar
t
a
o
n
J
u
l
y
7
,
1
9
8
9
.
Cu
r
ren
t
l
y
,
h
e
i
s
a
l
e
ct
u
rer
at
t
h
e
Pu
t
ra
In
d
o
n
e
s
i
a
U
n
i
v
ers
i
t
y
Y
PT
K
Pad
an
g
.
T
h
e
ed
u
cat
i
o
n
a
l
h
i
s
t
o
r
y
o
f
t
h
e
In
fo
rma
t
i
c
s
E
n
g
i
n
eeri
n
g
U
n
d
erg
ra
d
u
a
t
e
Pr
o
g
ram
w
as
c
o
mp
l
et
ed
i
n
2
0
1
2
a
n
d
t
h
e
In
f
o
rmat
i
cs
E
n
g
i
n
eeri
n
g
Mas
t
er
s
P
ro
g
ram
i
n
2
0
1
4
.
H
i
s
areas
o
f
e
x
p
er
t
i
s
e
i
n
cl
u
d
e
art
i
fi
c
i
al
i
n
t
el
l
i
g
en
ce
(
A
I),
ex
p
ert
s
y
s
t
em
s
(
E
S),
d
a
t
a
mi
n
i
n
g
(D
M),
a
n
d
d
eci
s
i
o
n
s
u
p
p
o
r
t
s
y
s
t
ems
(D
SS).
H
e
can
b
e
c
o
n
t
act
e
d
a
t
emai
l
:
mu
s
l
i
_
y
a
n
t
o
@
u
p
i
y
p
t
k
.
ac.
i
d
.
P
ro
f.
D
r.
Y
uha
ndri
,
S.
Ko
m
.
,
M
.
Ko
m
.
w
as
b
o
rn
i
n
T
an
j
u
n
g
A
l
am
o
n
May
1
5
,
1
9
7
3
.
H
e
i
s
an
A
s
s
i
s
t
a
n
t
Pro
fe
s
s
o
r
i
n
Facu
l
t
y
o
f
Co
mp
u
t
e
r
Sci
en
ce,
U
n
i
v
ers
i
t
a
s
Pu
t
ra
In
d
o
n
e
s
i
a
Y
PT
K
.
H
e
recei
v
ed
t
h
e
b
ach
e
l
o
r’
s
d
eg
r
ee
i
n
i
n
fo
rma
t
i
c
s
man
ag
eme
n
t
an
d
mas
t
er’
s
d
eg
ree
i
n
i
n
f
o
rmat
i
o
n
t
ecn
o
l
o
g
y
i
n
1
9
9
2
an
d
2
0
0
6
fro
m
U
n
i
v
ers
i
t
as
Pu
t
ra
In
d
o
n
e
s
i
a
Y
P
T
K
.
Mo
reo
v
er,
h
e
co
mp
l
e
t
ed
h
i
s
D
o
c
t
o
r
o
f
In
f
o
rmat
i
o
n
T
ech
n
o
l
o
g
y
as
In
fo
rma
t
i
c
s
Med
i
cal
Imag
e
E
x
p
er
t
i
s
e
fro
m
G
u
n
a
d
arma
U
n
i
v
ers
i
t
y
i
n
A
p
r
i
l
2
0
1
7
.
H
e
i
s
a
l
ec
t
u
r
er
at
t
h
e
Fac
u
l
t
y
o
f
Co
m
p
u
t
er
Sci
e
n
ce,
U
n
i
v
er
s
i
t
as
Pu
t
ra
In
d
o
n
es
i
a
Y
PT
K
.
H
e
can
b
e
c
o
n
t
act
e
d
at
emai
l
:
y
u
y
u
@
u
p
i
y
p
t
k
.
ac.
i
d
.
M
uha
m
m
a
d
Ta
juddi
n
i
s
a
l
ect
u
rer
at
Bu
mi
g
o
ra
U
n
i
v
er
s
i
t
y
.
H
e
s
erv
e
s
as
a
Pro
fes
s
o
r
at
t
h
e
Fac
u
l
t
y
o
f
Co
m
p
u
t
er
Sci
e
n
ce.
H
e
o
b
t
a
i
n
e
d
a
Bach
el
o
r's
d
eg
ree
a
t
Mat
ram
a
U
n
i
v
er
s
i
t
y
i
n
1
9
8
6
.
T
h
en
a
Mas
t
er
o
f
Sci
en
ce
d
e
g
ree
i
n
2
0
0
0
at
Braw
i
j
a
y
a
U
n
i
v
er
s
i
t
y
.
H
e
a
l
s
o
ach
i
e
v
ed
a
d
o
c
t
o
ra
t
e
d
eg
ree
i
n
2
0
1
2
at
Braw
i
j
a
y
a
U
n
i
v
ers
i
t
y
.
T
each
i
n
g
h
i
s
t
o
r
y
i
n
cl
u
d
e
s
res
earch
met
h
o
d
s
,
man
ag
eme
n
t
i
n
fo
rma
t
i
o
n
s
y
s
t
ems
,
d
eci
s
i
o
n
s
u
p
p
o
r
t
s
y
s
t
em
s
,
s
o
ft
w
are
en
g
i
n
eer
i
n
g
,
an
d
i
n
fo
rma
t
i
o
n
t
ec
h
n
o
l
o
g
y
res
earc
h
.
H
e
can
b
e
co
n
t
act
e
d
at
emai
l
:
t
aj
u
d
d
i
n
@
u
n
i
v
er
s
i
t
as
b
u
m
i
g
o
ra.
ac.
i
d
.
V
i
na
Tr
i
Septi
a
na
,
Sp.
R
a
d.
,
dr.
S.
Ked.
i
s
a
ra
d
i
o
l
o
g
y
d
o
ct
o
r
a
n
d
a
l
s
o
w
o
r
k
s
a
s
a
l
ect
u
rer
at
Ba
i
t
u
rrah
ma
h
U
n
i
v
er
s
i
t
y
,
Pad
a
n
g
C
i
t
y
,
W
es
t
Su
mat
ra
Pro
v
i
n
ce
.
H
i
s
t
o
r
y
o
f
med
i
ca
l
s
t
u
d
i
es
g
ra
d
u
a
t
e
d
i
n
2
0
0
6
t
h
e
n
co
m
p
l
e
t
ed
p
ro
fe
s
s
i
o
n
al
s
t
u
d
i
e
s
i
n
2
0
0
8
a
t
A
n
d
a
l
as
U
n
i
v
er
s
i
t
y
.
In
2
0
1
8
h
e
al
s
o
c
o
mp
l
et
e
d
Sp
-
1
s
t
u
d
i
e
s
at
t
h
e
U
n
i
v
er
s
i
t
y
o
f
In
d
o
n
e
s
i
a.
Sk
i
l
l
s
an
d
ex
p
er
t
i
s
e
i
n
c
l
u
d
e
d
i
ag
n
o
s
t
i
c
i
mag
i
n
g
,
d
i
ag
n
o
s
t
i
c
rad
i
o
l
o
g
y
,
co
mp
u
t
e
d
t
o
mo
g
rap
h
y
,
mag
n
et
i
c
res
o
n
a
n
ce,
i
mag
i
n
g
,
med
i
cal
i
mag
i
n
g
,
rad
i
o
g
ra
p
h
y
,
cl
i
n
i
ca
l
i
mag
i
n
g
,
u
l
t
ra
s
o
u
n
d
i
ma
g
i
n
g
,
an
d
n
eu
r
o
i
ma
g
i
n
g
.
Sh
e
can
b
e
co
n
t
act
e
d
at
ema
i
l
:
s
3
p
t
a
v
i
n
a
@
fk
.
u
n
b
ra
h
.
ac.
i
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.