I
nte
rna
t
io
na
l J
o
urna
l o
f
Adv
a
nces in Applie
d Science
s
(
I
J
AAS)
Vo
l.
6
,
No
.
4
,
Dec
em
b
er
2
0
1
7
,
p
p
.
3
6
8
~3
7
4
I
SS
N:
2252
-
8814
368
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
J
AAS
Cla
ss
ificatio
n of
Co
ntent
Ba
sed M
edica
l I
m
a
g
e
Ret
r
iev
a
l Using
Tex
ture
a
nd
Sha
p
e F
ea
tu
re w
ith
Ne
ura
l Net
w
o
rk
Sw
ee
t
y
M
a
nia
r,
J
a
g
dis
h S.
Sh
a
h
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
/I
T
E
n
g
i
n
ee
r
i
n
g
,
G
u
j
ar
at
T
ec
h
n
o
lo
g
ical
U
n
iv
er
s
it
y
,
G
u
j
ar
at,
I
n
d
ia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Sep
1
7
,
2
0
1
7
R
ev
i
s
ed
No
v
1
8
,
2
0
1
7
A
cc
ep
ted
No
v
2
4
,
2
0
1
7
M
e
d
ica
l
ima
g
e
c
la
ss
i
f
ica
ti
o
n
a
n
d
re
tri
e
v
a
l
s
y
ste
m
s
h
a
v
e
b
e
e
n
f
in
d
in
g
e
x
ten
siv
e
u
se
in
th
e
a
re
a
s
o
f
i
m
a
g
e
c
las
sif
i
c
a
ti
o
n
a
c
c
o
rd
in
g
to
im
a
g
in
g
m
o
d
a
li
ti
e
s,
b
o
d
y
p
a
rt
a
n
d
d
ise
a
se
s.
On
e
o
f
th
e
m
a
jo
r
c
h
a
ll
e
n
g
e
s
in
th
e
m
e
d
ica
l
c
las
si
f
ica
ti
o
n
is
th
e
larg
e
siz
e
i
m
a
g
e
s
l
e
a
d
in
g
to
a
larg
e
n
u
m
b
e
r
o
f
e
x
trac
ted
f
e
a
tu
re
s
w
h
ich
is
a
b
u
rd
e
n
f
o
r
th
e
c
las
si
f
ica
ti
o
n
a
lg
o
rit
h
m
a
n
d
th
e
re
so
u
rc
e
s.
In
t
h
is
p
a
p
e
r,
a
n
o
v
e
l
a
p
p
r
o
a
c
h
f
o
r
a
u
to
m
a
ti
c
c
las
si
fica
ti
o
n
o
f
f
u
n
d
u
s
im
a
g
e
s
is
p
ro
p
o
se
d
.
T
h
e
m
e
th
o
d
u
se
s
im
a
g
e
a
n
d
d
a
ta
p
re
-
p
ro
c
e
ss
in
g
tec
h
n
iq
u
e
s
to
im
p
ro
v
e
th
e
p
e
rf
o
rm
a
n
c
e
o
f
m
a
c
h
in
e
lea
rn
in
g
c
las
si
f
iers
.
S
o
m
e
p
re
d
o
m
in
a
n
t
im
a
g
e
m
in
in
g
a
lg
o
rit
h
m
s
su
c
h
a
s
Clas
sif
i
c
a
ti
o
n
,
Re
g
re
ss
io
n
T
re
e
(C
A
R
T
),
Ne
u
ra
l
Ne
t
w
o
rk
,
Na
iv
e
B
a
y
e
s
(NB),
De
c
isio
n
T
re
e
(DT
)
K
-
Ne
a
re
st
Ne
i
g
h
b
o
r
T
h
e
p
e
rf
o
r
m
a
n
c
e
o
f
M
CBIR
sy
ste
m
s
u
sin
g
tex
tu
re
a
n
d
sh
a
p
e
f
e
a
tu
re
s
e
ff
icie
n
t.
T
h
e
p
o
ss
ib
le
o
u
tc
o
m
e
s
o
f
a
t
w
o
c
la
ss
p
re
d
ictio
n
b
e
re
p
re
se
n
ted
a
s
T
ru
e
p
o
siti
v
e
(T
P
),
T
ru
e
n
e
g
a
ti
v
e
(TN),
F
a
lse
P
o
s
it
iv
e
(F
P
)
a
n
d
F
a
lse
Ne
g
a
ti
v
e
(F
N).
K
ey
w
o
r
d
:
C
las
s
i
fi
ca
tio
n
Me
d
ical
i
m
a
g
in
g
Neu
r
al
n
et
w
o
r
k
s
T
ex
tu
r
e
an
d
s
h
ap
e
f
ea
tu
r
es
Co
p
y
rig
h
t
©
201
7
In
s
t
it
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
:
S
w
ee
t
y
Ma
n
iar
,
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
/I
T
E
n
g
i
n
ee
r
i
n
g
,
Gu
j
ar
at
T
ec
h
n
o
lo
g
ical
Un
i
v
er
s
it
y
,
G
u
j
ar
at,
I
n
d
ia
.
1.
I
NT
RO
D
UCT
I
O
N
I
m
ag
e
m
i
n
in
g
i
s
th
e
p
r
o
ce
s
s
u
s
ed
to
ex
tr
ac
t
m
ea
n
i
n
g
f
u
l
i
n
f
o
r
m
atio
n
f
r
o
m
i
m
a
g
es.
I
t
d
ea
ls
w
it
h
th
e
e
m
b
ed
d
ed
k
n
o
w
led
g
e,
ex
tr
ac
t
in
g
i
n
h
er
en
t
d
ata,
i
m
a
g
e
d
ata
r
elatio
n
s
h
ip
an
d
o
th
er
p
atter
n
s
th
at
ar
e
n
o
t
clea
r
l
y
f
o
u
n
d
in
th
e
i
m
a
g
es
[
1
]
.
T
h
er
e
is
a
s
y
s
te
m
w
h
ic
h
is
C
o
n
te
n
t
B
ased
I
m
a
g
e
R
e
tr
iev
al
(
C
B
I
R
)
w
h
ic
h
ai
m
s
at
s
ea
r
ch
i
n
g
o
f
i
m
a
g
es
av
a
il
ab
l
e
in
d
atab
ase
s
f
o
r
an
y
p
ar
tic
u
lar
i
m
a
g
es
s
o
as
to
g
et
a
r
elate
d
i
m
a
g
e.
T
h
e
ex
tr
ac
ti
n
g
i
m
a
g
e
s
b
ased
o
n
s
o
m
e
f
ea
t
u
r
es
s
u
c
h
a
s
s
h
ap
e
,
tex
t
u
r
e,
r
eg
io
n
an
d
s
o
o
n
.
On
t
h
e
o
t
h
er
en
d
,
R
etr
iev
al
o
f
i
m
a
g
e
is
t
h
e
f
ast
d
ev
elo
p
in
g
a
n
d
ch
al
len
g
i
n
g
r
e
s
ea
r
ch
p
ar
t in
b
o
th
u
n
m
o
v
i
n
g
an
d
m
o
v
in
g
i
m
a
g
es
.
E
s
p
ec
iall
y
,
t
h
e
m
ed
ical
i
m
a
g
e
class
i
f
icat
io
n
p
la
y
s
a
n
i
m
p
o
r
tan
t
r
o
le
i
n
h
u
m
a
n
d
iag
n
o
s
i
s
an
d
tr
ea
t
m
e
n
t.
I
t
is
also
u
s
ed
f
o
r
h
ea
lth
ca
r
e
s
t
u
d
en
ts
in
th
e
ed
u
ca
tio
n
a
l
d
o
m
a
in
a
n
d
s
t
u
d
ies
b
y
e
x
p
lain
i
n
g
w
it
h
th
e
s
e
i
m
ag
e
s
.
Me
d
ical
i
m
a
g
es a
r
e
m
ai
n
l
y
u
s
ed
to
d
etec
t sp
ec
if
ic
d
is
ea
s
es o
cc
u
r
in
t
h
e
h
u
m
an
b
o
d
y
.
I
m
ag
e
m
atc
h
i
n
g
i
s
m
o
r
e
i
m
p
o
r
tan
t
in
t
h
e
f
ield
o
f
m
i
n
in
g
i
m
a
g
es.
Fre
q
u
e
n
tl
y
u
s
ed
tec
h
n
iq
u
e
is
n
ea
r
est
n
eig
h
b
o
r
h
o
o
d
in
w
h
ic
h
o
b
j
ec
ts
ar
e
r
ep
r
esen
ted
as
n
d
i
m
en
s
io
n
al
v
ec
to
r
s
.
I
n
[
6
]
th
e
v
is
u
al
q
u
er
ie
s
ar
e
r
ep
r
esen
ted
in
t
h
e
r
etr
iev
al
p
r
o
ce
s
s
.
So
th
at
t
h
e
i
m
a
g
es
m
ai
n
l
y
b
a
s
ed
o
n
th
e
u
s
er
r
eq
u
e
s
t
an
d
th
e
m
ec
h
an
i
s
m
is
co
n
s
id
er
ed
as
q
u
er
y
-
by
-
ex
a
m
p
le
u
s
ed
to
co
m
p
ar
e
th
e
tar
g
et
i
m
ag
e
s
to
f
i
n
d
th
e
i
m
a
g
e
in
d
ices
p
r
esen
t
in
t
h
e
i
m
a
g
e
d
atab
ase.
Fo
r
ea
s
e
o
f
a
cc
ess
d
ig
ital
m
ed
ical
i
m
a
g
es
s
to
r
ed
in
h
u
g
e
d
atab
ases
as
w
e
ll
as
C
o
n
te
n
t
b
ased
i
m
a
g
e
r
etr
iev
a
l (
C
B
I
R
)
w
h
ic
h
is
m
ai
n
l
y
u
s
ed
i
n
d
iag
n
o
s
tic
c
ases
li
k
e
q
u
er
y
m
ed
ical
i
m
ag
e
.
T
h
e
C
B
I
R
i
m
ag
e
s
is
b
ased
o
n
s
o
m
e
f
ea
tu
r
e
s
s
u
c
h
as
ed
g
e,
s
h
ap
e
a
n
d
te
x
t
u
r
e
w
h
ic
h
ar
e
e
x
tr
ac
ted
au
to
m
atic
all
y
[
7
]
.
I
f
th
er
e
i
s
e
m
p
t
y
i
n
th
e
i
m
a
g
e
s
et
o
r
less
th
an
t
h
e
to
tal
i
m
a
g
es
t
h
en
t
h
e
s
y
s
te
m
r
an
d
o
m
l
y
c
h
o
s
en
t
h
e
i
m
a
g
e
f
o
r
cr
ea
tin
g
th
e
as
s
o
ciatio
n
r
u
les.
T
h
is
p
a
p
er
g
iv
es
a
s
u
r
v
e
y
o
n
s
e
v
er
al
tech
n
iq
u
es
in
i
m
a
g
e
m
in
in
g
w
h
ic
h
w
as
al
r
ea
d
y
p
r
o
p
o
s
ed
m
et
h
o
d
th
e
y
ar
e
N
eu
r
al
Net
w
o
r
k
,
C
AR
T
,
Naiv
e
B
a
y
es,
K
NN
a
n
d
Dec
i
s
io
n
T
r
ee
.
T
h
is
p
ap
er
p
r
o
v
id
es
b
est
m
et
h
o
d
in
m
ed
ical
i
m
ag
e
clas
s
i
f
icatio
n
b
ased
o
n
t
h
e
c
lass
if
icatio
n
ac
cu
r
ac
y
,
p
r
o
ce
s
s
i
n
g
ti
m
e
an
d
er
r
o
r
r
ates.
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
AA
S
I
SS
N:
2252
-
8814
C
la
s
s
i
fica
tio
n
o
f Co
n
ten
t B
a
s
e
d
Med
ica
l I
ma
g
e
R
etri
ev
a
l U
s
in
g
…
(
S
w
ee
ty
Ma
n
ia
r
)
369
2.
RE
L
AT
E
D
WO
RK
Desh
p
a
n
d
e
et
al
[
8
]
p
r
o
v
id
es
d
ata
m
i
n
i
n
g
ap
p
r
o
ac
h
w
h
ic
h
is
u
s
ed
to
id
en
tify
t
h
e
i
m
a
g
e
co
n
ten
t
p
r
esen
t
i
n
t
h
e
as
s
o
ciatio
n
r
u
le
s
.
T
h
e
ass
o
ciatio
n
r
u
le
alg
o
r
it
h
m
h
e
lp
s
to
d
etec
t
t
h
e
r
e
g
u
la
r
ite
m
s
et
w
i
th
th
e
h
elp
o
f
s
o
m
e
iter
ati
v
e
m
eth
o
d
s
.
T
h
is
alg
o
r
it
h
m
h
elp
s
to
m
i
n
i
m
ize
t
h
e
n
u
m
b
er
o
f
s
ca
n
s
in
A
p
r
io
r
i a
lg
o
r
it
h
m
.
I
t
is
v
er
y
es
s
en
tial
to
ad
v
an
ce
th
e
i
m
ag
e
q
u
ali
t
y
a
n
d
m
ak
e
t
h
e
ex
tr
ac
tio
n
p
h
a
s
e
as
s
i
m
p
le
a
n
d
r
eliab
le.
Li
-
Ho
n
g
J
u
an
g
et
al
[
9
]
f
o
cu
s
ed
o
n
tr
ac
k
in
g
t
u
m
o
r
o
b
j
ec
ts
o
f
(
MRI)
b
r
ain
i
m
ag
e
s
b
y
u
s
in
g
K
-
m
e
an
s
al
g
o
r
ith
m
.
T
h
e
p
r
o
ce
s
s
w
h
ich
i
s
also
u
s
e
f
u
l
f
o
r
d
etec
tin
g
ex
ac
t
le
s
io
n
o
b
j
ec
ts
in
i
m
a
g
es.
T
h
e
m
ai
n
p
u
r
p
o
s
e
o
f
th
i
s
alg
o
r
it
h
m
is
to
r
eso
lv
e
th
e
M
R
I
i
m
a
g
e
b
y
ch
a
n
g
in
g
t
h
e
g
r
a
y
-
le
v
el
i
m
a
g
e
in
to
co
lo
u
r
i
m
ag
e.
S.L
.
A
.
L
ee
et
al
[
1
0
]
co
n
ce
n
tr
ated
o
n
lu
n
g
n
o
d
u
le
d
ete
ctio
n
w
h
ich
i
s
u
s
ed
to
s
p
o
t
th
e
lu
n
g
ab
n
o
r
m
alitie
s
i
n
C
T
lu
n
g
i
m
a
g
es
w
it
h
t
h
e
h
elp
o
f
R
an
d
o
m
f
o
r
est
al
g
o
r
ith
m
.
T
h
i
s
al
g
o
r
ith
m
p
r
o
v
id
es
h
y
b
r
id
r
an
d
o
m
f
o
r
est
b
ased
n
o
d
u
le
class
i
f
icatio
n
.
I
t
is
also
u
s
ed
to
d
etec
t
3
2
p
atien
ts
w
i
th
5
7
2
1
im
ag
e
s
.
T
h
e
ac
cu
r
a
c
y
in
p
r
o
p
o
s
ed
s
y
s
te
m
is
n
o
ted
as
9
7
.
1
1
w
h
er
ea
s
in
th
e
d
ev
elo
p
ed
s
y
s
te
m
th
e
h
ig
h
r
ec
eiv
er
o
p
er
ato
r
ch
ar
ac
ter
is
tic
i
s
g
iv
e
n
9
7
.
8
6
%
ac
cu
r
ac
y
.
Ma
h
n
az
E
te
h
ad
T
av
ak
o
l
et
al
[
1
1
]
p
r
o
v
id
e
th
e
h
i
g
h
i
n
f
r
ar
ed
ca
m
er
as
to
d
iag
n
o
s
e
t
h
e
v
asc
u
lar
c
h
an
g
es
o
f
b
r
ea
s
t
s
b
y
u
s
i
n
g
t
h
e
ad
a
b
o
o
s
t a
lg
o
r
ith
m
.
T
h
e
al
g
o
r
ith
m
i
s
u
s
ed
to
clas
s
i
f
y
th
e
i
n
v
is
ib
le
i
m
a
g
es
i
n
to
b
en
i
g
n
,
m
al
ig
n
a
n
t
a
n
d
n
o
r
m
a
l.
I
n
th
is
s
y
s
te
m
th
e
ac
cu
r
ac
y
o
f
8
3
%
is
g
i
v
en
w
h
ic
h
g
iv
e
s
b
etter
p
er
f
o
r
m
an
ce
t
h
a
n
th
e
p
r
o
p
o
s
ed
s
y
s
te
m
o
f
6
6
%.
Min
g
-
Yih
L
ee
e
t
al
[
1
2
]
p
r
o
p
o
s
ed
an
en
tr
o
p
y
b
ased
f
ea
t
u
r
e
ex
tr
ac
tio
n
a
n
d
s
o
m
e
o
th
er
p
r
o
to
co
ls
f
o
r
th
e
b
r
ea
s
t
ca
n
ce
r
d
iag
n
o
s
is
u
s
in
g
d
ec
is
io
n
tr
ee
alg
o
r
ith
m
.
T
h
e
Mo
r
p
h
o
lo
g
ical
o
p
er
atio
n
s
u
s
ed
in
t
h
is
s
y
s
te
m
to
d
etec
t
th
e
u
n
if
ied
ab
n
o
r
m
a
l
r
eg
io
n
s
.
T
h
is
m
et
h
o
d
g
iv
e
s
8
6
%
ac
cu
r
ac
y
w
h
ic
h
is
b
etter
th
an
th
e
p
r
o
p
o
s
ed
s
y
s
te
m
o
f
5
9
%.
Ye
C
h
e
n
et
al
[
1
3
]
f
o
cu
s
ed
o
n
th
e
d
etec
tio
n
o
f
b
r
ain
s
tr
u
ctu
r
al
ch
an
g
es
f
r
o
m
t
h
e
Ma
g
n
etic
r
eso
n
an
ce
i
m
a
g
es
w
h
ich
h
elp
s
to
aid
th
e
tr
ea
t
m
en
t
o
f
n
eu
r
o
lo
g
ical
d
is
ea
s
es
w
it
h
t
h
e
h
elp
o
f
S
u
p
p
o
r
t
Vec
to
r
Ma
ch
in
e
alg
o
r
ith
m
.
I
n
ad
d
itio
n
th
e
al
g
o
r
ith
m
w
h
ic
h
h
elp
s
to
an
al
y
z
e
th
e
MR
i
m
a
g
e
s
f
r
o
m
t
h
e
v
a
r
io
u
s
d
atasets
.
T
h
e
ac
cu
r
ac
y
r
an
g
e
b
et
w
ee
n
7
0
%
an
d
8
7
% a
r
e
n
o
ted
.
W
en
-
J
ie
W
u
et
al
[
1
4
]
s
u
g
g
ested
b
o
th
t
h
e
clas
s
i
f
icatio
n
ac
cu
r
ac
y
an
d
t
h
e
o
p
ti
m
al
cla
s
s
i
f
icat
io
n
m
o
d
el
w
h
ic
h
h
elp
s
to
d
etec
t
t
h
e
u
ltra
s
o
u
n
d
b
r
ea
s
t
tu
m
o
r
i
m
ag
es
b
y
u
s
i
n
g
g
e
n
etic
alg
o
r
it
h
m
.
T
h
e
alg
o
r
ith
m
is
to
ca
lcu
late
t
h
e
n
ea
r
o
p
ti
m
al
p
ar
am
eter
s
to
d
if
f
er
en
tia
te
th
e
tu
m
o
r
as
b
e
n
i
g
n
o
r
m
ali
g
n
a
n
t.
T
h
e
ac
cu
r
ac
y
o
f
p
r
o
p
o
s
ed
s
y
s
te
m
i
s
9
5
%
w
h
i
ch
is
i
m
p
r
o
v
ed
b
etter
in
t
h
e
d
ev
elo
p
in
g
s
y
s
te
m
b
y
r
ed
u
ci
n
g
th
e
b
io
p
s
ies
o
f
b
en
ig
n
lesi
o
n
s
.
Dan
iel
J
.
E
v
er
s
et
al
[
1
5
]
h
as
g
iv
e
n
t
h
e
s
t
u
d
y
to
ev
al
u
ate
wh
eth
er
t
h
e
o
p
ti
m
al
s
p
ec
tr
o
s
co
p
y
i
m
p
r
o
v
e
th
e
ac
c
u
r
ac
y
o
f
tr
a
n
s
t
h
o
r
ac
ic
l
u
n
g
b
io
p
s
ies
u
s
i
n
g
C
la
s
s
i
f
ica
ti
o
n
an
d
r
e
g
r
e
s
s
io
n
tr
ee
(
C
A
R
T
)
alg
o
r
ith
m
.
B
ased
o
n
th
e
d
er
iv
ed
p
ar
a
m
eter
th
e
alg
o
r
ith
m
clas
s
i
f
ies
t
h
e
t
y
p
e
o
f
tis
s
u
e
p
r
esen
t
i
n
th
e
s
y
s
te
m
.
T
h
e
o
v
er
all
ac
cu
r
ac
y
is
9
1
% se
n
s
iti
v
it
y
.
Dan
iel
J
.
E
v
er
s
et
al
[
1
6
]
h
as
g
iv
e
n
t
h
e
s
t
u
d
y
to
ev
al
u
ate
wh
eth
er
t
h
e
o
p
ti
m
al
s
p
ec
tr
o
s
c
o
p
y
i
m
p
r
o
v
e
th
e
ac
c
u
r
ac
y
o
f
tr
a
n
s
t
h
o
r
ac
ic
l
u
n
g
b
io
p
s
ies
u
s
i
n
g
C
la
s
s
i
f
ica
ti
o
n
an
d
r
e
g
r
ess
io
n
tr
ee
(
C
A
R
T
)
alg
o
r
ith
m
.
B
ased
o
n
th
e
d
er
iv
ed
p
ar
a
m
eter
th
e
alg
o
r
ith
m
clas
s
i
f
ies
t
h
e
t
y
p
e
o
f
tis
s
u
e
p
r
esen
t
i
n
th
e
s
y
s
te
m
.
T
h
e
o
v
er
all
ac
cu
r
ac
y
is
9
1
% se
n
s
iti
v
it
y
.
M
in
-
C
h
u
n
Ya
n
g
et
al
[
1
7
]
en
h
an
ce
t
h
e
n
aï
v
e
b
a
y
es
clas
s
i
f
i
ca
tio
n
alg
o
r
ith
m
b
y
s
ep
ar
ati
n
g
th
e
u
l
tr
a
s
o
u
n
d
i
m
a
g
es
p
ix
el
-
by
-
p
i
x
el
th
en
t
h
e
i
m
ag
e
m
ea
s
u
r
ed
b
y
g
r
a
y
s
ca
le
i
s
co
n
v
er
ted
to
b
in
ar
y
i
m
ag
e
w
h
ich
i
s
th
en
e
v
al
u
ated
b
y
t
w
o
-
p
h
ase
c
r
iter
ia.
So
,
th
e
d
etec
tio
n
s
en
s
it
iv
it
y
ca
n
b
e
f
u
r
t
h
er
d
ev
elo
p
ed
.
Sh
e
n
g
j
u
n
Z
h
o
u
et
al
[
1
8
]
s
u
g
g
ested
th
a
t
in
t
h
e
m
ed
ical
ap
p
licatio
n
s
t
h
e
i
m
a
g
es
ar
e
s
e
g
m
en
ted
.
T
o
m
an
a
g
e
t
h
e
s
eg
m
e
n
tatio
n
,
f
u
z
z
y
c
-
m
ea
n
s
clu
s
ter
i
n
g
d
o
t
h
e
class
i
f
icatio
n
o
f
p
i
x
el
s
i
n
to
s
o
m
e
d
iv
is
io
n
s
.
T
h
en
th
e
alg
o
r
i
t
h
m
ass
ig
n
s
t
h
e
m
e
m
b
er
s
h
ip
v
al
u
es
f
o
r
th
o
s
e
p
ix
el
s
to
f
o
r
m
t
h
e
ce
n
tr
o
id
.
R
av
i
B
ab
u
et
al.
[
1
9
]
f
o
cu
s
ed
to
d
eter
m
i
n
e
th
e
i
m
a
g
e
class
if
icatio
n
r
ate
f
o
r
th
e
p
u
r
p
o
s
e
o
f
d
ig
ita
l
i
m
a
g
e
class
if
ica
tio
n
.
T
h
e
K
-
Nea
r
est
n
ei
g
h
b
o
r
alg
o
r
ith
m
u
s
es
t
h
e
lea
m
i
n
g
tec
h
n
iq
u
e
to
f
in
d
o
u
t
t
h
e
class
i
f
icatio
n
ti
m
e
o
f
t
h
o
s
e
i
m
ag
e
s
.
T
h
e
laz
y
b
ased
an
d
i
n
s
tan
ce
b
ased
ar
e
t
h
e
t
w
o
lea
m
i
n
g
tec
h
n
iq
u
e
s
.
T
o
co
m
p
ar
e
th
e
c
u
r
v
es
t
h
e
alg
o
r
ith
m
is
u
s
ed
w
h
ic
h
b
ased
o
n
s
o
m
e
co
m
p
ar
is
o
n
.
Fi
n
all
y
t
h
e
n
ea
r
est
n
ei
g
h
b
o
r
class
i
f
ier
s
u
s
ed
to
m
ea
s
u
r
e
t
h
e
d
is
tan
ce
o
f
t
h
e
t
w
o
cu
r
v
es [
2
0
]
.
3.
WO
RK
I
N
G
O
F
CL
AS
SI
F
I
CATI
O
N
SYS
T
E
M
T
o
au
to
m
atica
ll
y
ca
te
g
o
r
ize
m
ed
ical
i
m
ag
e
s
,
w
e
h
a
v
e
e
x
-
p
er
im
e
n
ted
o
n
r
ea
l
m
a
m
m
o
g
r
a
m
s
w
it
h
t
w
o
d
ata
m
in
i
n
g
tec
h
n
iq
u
e
s
,
a
s
s
o
ciatio
n
r
u
le
m
in
in
g
a
n
d
n
e
u
r
al
n
et
w
o
r
k
s
.
I
n
b
o
th
ca
s
es,
t
h
e
p
r
o
b
l
e
m
co
n
s
i
s
ts
o
f
b
u
ild
i
n
g
a
m
a
m
m
o
g
r
ap
h
y
c
lass
i
fi
ca
tio
n
m
o
d
el
u
s
i
n
g
a
ttrib
u
te
s
ex
tr
ac
ted
f
r
o
m
a
n
d
attac
h
ed
to
m
a
m
m
o
g
r
a
m
s
,
th
e
n
ev
al
u
ati
n
g
th
e
ef
f
ec
ti
v
e
n
ess
o
f
t
h
e
m
o
d
el
u
s
i
n
g
n
e
w
i
m
a
g
es.
T
h
e
p
r
o
ce
s
s
o
f
b
u
ild
in
g
th
e
class
i
fi
ca
tio
n
m
o
d
el
(
cla
s
s
i
fier
)
in
cl
u
d
es
p
r
ep
r
o
ce
s
s
i
n
g
an
d
e
x
tr
ac
tio
n
o
f
v
i
s
u
a
l
f
ea
t
u
r
es
f
r
o
m
alr
ea
d
y
lab
elle
d
i
m
a
g
es (
i.e
.
tr
ain
i
n
g
s
et)
.
Fig
u
r
e
1
s
h
o
w
s
a
n
o
v
er
v
ie
w
o
f
t
h
e
ca
te
g
o
r
izatio
n
p
r
o
ce
s
s
ad
o
p
ted
f
o
r
b
o
th
s
y
s
te
m
s
.
T
h
e
first
s
tep
is
r
ep
r
esen
ted
b
y
th
e
i
m
ag
e
ac
q
u
is
i
tio
n
an
d
i
m
a
g
e
en
h
a
n
ce
m
e
n
t,
f
o
llo
w
ed
b
y
f
ea
tu
r
e
ex
tr
ac
t
io
n
.
T
h
e
last
o
n
e
is
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8814
IJ
AA
S
Vo
l.
6
,
No
.
4
,
Dec
em
b
er
201
7
:
3
6
8
–
37
4
370
th
e
cla
s
s
i
ficatio
n
p
ar
t
w
h
er
e
th
e
tech
n
iq
u
e
f
o
r
s
u
p
er
v
is
e
d
lear
n
in
g
is
d
i
f
f
er
en
t.
All
th
ese
p
ar
ts
o
f
t
h
e
class
i
fi
ca
tio
n
s
y
s
te
m
s
ar
e
d
is
c
u
s
s
ed
in
m
o
r
e
d
etail
later
.
Fig
u
r
e
1
.
C
lass
if
ica
tio
n
S
y
s
t
em
3
.
1
.
I
m
a
g
e
Acquis
it
io
n
T
o
h
av
e
ac
ce
s
s
to
r
ea
l
m
ed
ic
al
i
m
a
g
es
f
o
r
e
x
p
er
i
m
e
n
tatio
n
is
a
v
er
y
d
i
f
fi
c
u
lt
u
n
d
er
tak
i
n
g
d
u
e
to
p
r
iv
ac
y
i
s
s
u
es
a
n
d
h
ea
v
y
b
u
r
ea
u
cr
atic
h
u
r
d
les.
T
h
e
d
ata
c
o
llectio
n
t
h
at
w
a
s
u
s
ed
i
n
o
u
r
ex
p
er
i
m
e
n
ts
w
a
s
tak
en
f
r
o
m
t
h
e
Ma
m
m
o
g
r
ap
h
i
c
I
m
ag
e
A
n
a
l
y
s
is
S
o
ciet
y
(
MI
AS)
[
1
8
]
.
T
h
is
s
a
m
e
co
llectio
n
h
as
b
ee
n
u
s
ed
i
n
o
th
er
s
tu
d
ie
s
o
f
au
to
m
atic
m
a
m
m
o
g
r
ap
h
y
clas
s
i
fi
ca
tio
n
.
3
.
2
.
I
m
a
g
e
E
nh
a
nce
m
ent
Ma
m
m
o
g
r
a
m
s
ar
e
i
m
a
g
es
d
i
f
fi
c
u
lt
to
in
ter
p
r
et,
an
d
a
p
r
ep
r
o
ce
s
s
in
g
p
h
a
s
e
o
f
th
e
i
m
a
g
es
i
s
n
ec
e
s
s
ar
y
to
i
m
p
r
o
v
e
th
e
q
u
ali
t
y
o
f
t
h
e
i
m
a
g
es
a
n
d
m
a
k
e
th
e
f
ea
t
u
r
e
ex
tr
ac
tio
n
p
h
a
s
e
m
o
r
e
r
eliab
l
e.
P
r
e
-
p
r
o
ce
s
s
in
g
is
al
w
a
y
s
a
n
ec
es
s
it
y
w
h
e
n
ev
er
th
e
d
ata
to
b
e
m
i
n
ed
in
n
o
is
y
,
in
co
n
s
is
ten
t
o
r
i
n
co
m
p
lete
a
n
d
p
r
e
-
p
r
o
ce
s
s
i
n
g
s
ig
n
i
fi
ca
n
tl
y
i
m
p
r
o
v
es t
h
e
ef
f
e
ctiv
e
n
ess
o
f
th
e
d
ata
m
i
n
in
g
te
ch
n
iq
u
es
I
m
ag
e
e
n
h
an
ce
m
e
n
t
h
elp
s
in
q
u
alitati
v
e
i
m
p
r
o
v
e
m
en
t
o
f
th
e
i
m
a
g
e
w
i
th
r
e
s
p
ec
t
to
a
s
p
ec
i
fi
c
ap
p
licatio
n
[
1
0
]
.
I
n
o
r
d
e
r
to
d
i
m
i
n
is
h
t
h
e
ef
f
ec
t
o
f
o
v
er
b
r
ig
h
t
n
es
s
o
r
o
v
er
d
ar
k
n
es
s
in
t
h
e
i
m
ag
e
s
an
d
ac
ce
n
tu
a
te
t
h
e
i
m
a
g
e
f
ea
t
u
r
es
,
w
e
ap
p
lied
a
w
id
el
y
u
s
ed
t
ec
h
n
i
q
u
e
i
n
i
m
a
g
e
p
r
o
ce
s
s
in
g
to
i
m
p
r
o
v
e
v
is
u
al
ap
p
ea
r
an
ce
o
f
i
m
a
g
es
k
n
o
w
n
as
His
to
g
r
a
m
E
q
u
aliza
tio
n
.
Hi
s
to
g
r
a
m
eq
u
aliza
tio
n
i
n
cr
ea
s
e
s
th
e
co
n
tr
ast
r
a
n
g
e
in
an
i
m
a
g
e
b
y
in
cr
ea
s
in
g
t
h
e
d
y
n
a
m
ic
r
an
g
e
o
f
g
r
e
y
lev
e
ls
(
o
r
co
lo
u
r
s
)
[
1
0
]
.
T
h
is
im
p
r
o
v
e
s
th
e
d
is
ti
n
ctio
n
o
f
f
ea
t
u
r
es
i
n
t
h
e
i
m
ag
e.
T
h
e
m
e
th
o
d
p
r
o
ce
ed
s
b
y
w
id
e
n
in
g
t
h
e
p
ea
k
s
i
n
t
h
e
i
m
ag
e
h
i
s
to
g
r
a
m
a
n
d
co
m
p
r
ess
i
n
g
th
e
v
alle
y
s
.
T
h
is
p
r
o
ce
s
s
eq
u
a
lizes
t
h
e
ill
u
m
i
n
atio
n
o
f
t
h
e
i
m
ag
e
an
d
ac
ce
n
t
u
ates
th
e
f
ea
t
u
r
es
to
b
e
e
x
tr
ac
ted
.
T
h
at
is
h
o
w
t
h
e
d
i
f
f
er
e
n
t ill
u
m
in
atio
n
co
n
d
itio
n
s
at
th
e
s
ca
n
n
in
g
p
h
ase
ar
e
r
ed
u
ce
d
.
3
.
3
.
F
ea
t
ure
E
x
t
ra
ct
io
n
Featu
r
e
ex
tr
ac
tio
n
is
a
b
est
f
o
r
m
o
f
d
i
m
en
s
io
n
al
it
y
r
ed
u
ce
s
.
W
h
en
th
e
in
p
u
t
to
th
e
v
ar
io
u
s
m
et
h
o
d
s
ar
e
to
o
b
ig
to
b
e
g
i
v
e
a
n
d
it
i
s
b
eliev
ed
to
b
e
d
is
r
ep
u
tab
l
y
u
n
n
ee
d
ed
(
m
o
r
e
d
ata,
b
u
t
n
o
t
m
o
r
e
in
f
o
r
m
a
tio
n
)
th
en
th
e
i
n
p
u
t
d
ata
w
ill
b
e
ch
an
g
ed
i
n
to
a
co
m
p
ac
t
v
er
s
io
n
w
it
h
d
i
f
f
er
e
n
t
n
u
m
b
er
o
f
f
ea
t
u
r
es
(
also
n
a
m
ed
a
s
f
ea
t
u
r
es
v
ec
to
r
)
.
Sto
r
in
g
t
h
e
i
n
p
u
t
d
ata
in
to
t
h
e
o
th
er
f
o
r
m
at
o
f
f
ea
t
u
r
es
is
ca
lled
f
ea
tu
r
es
ex
tr
ac
tio
n
.
T
h
e
n
u
m
b
er
s
o
f
tec
h
n
iq
u
es
f
o
r
f
ea
t
u
r
e
ex
tr
ac
tio
n
ar
e
g
iv
e
n
b
elo
w
3
.
3
.
1
.
T
ex
t
ure
T
ex
tu
r
e
d
em
o
n
s
tr
atio
n
ca
n
b
e
o
f
d
if
f
er
en
t
t
y
p
e
s
:
s
tr
u
ctu
r
al
an
d
s
tatis
tica
l.
Statis
t
ical
f
ea
t
u
r
es
ca
n
b
e
ca
lcu
lated
w
i
th
co
o
cc
u
r
r
en
ce
m
atr
ice
s
,
p
r
in
cip
al
co
m
p
o
n
e
n
t
a
n
al
y
s
i
s
.
[
1
3
]
T
h
e
f
ea
tu
r
e
s
lik
e
m
ea
n
v
ar
ia
n
ce
s
tan
d
ar
d
d
ev
iatio
n
,
e
n
er
g
y
,
en
tr
o
p
y
,
co
r
r
elat
io
n
,
in
er
tia
ar
e
c
alcu
lated
u
s
i
n
g
co
-
o
cc
u
r
r
en
ce
m
atr
i
x
.
C
o
n
tr
a
s
t
is
th
e
co
m
p
u
te
o
f
d
if
f
er
en
ce
i
n
t
h
e
g
r
a
y
lev
el
f
o
r
co
o
cc
u
r
r
en
ce
m
atr
ix
[
9
]
.
=
∑
∑
/
=
1
=
1
(1
)
=
1
∑
∑
−
=
1
=
1
(2
)
=
√
(3
)
=
∑
(
−
)
(
−
)
,
(
4
)
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
AA
S
I
SS
N:
2252
-
8814
C
la
s
s
i
fica
tio
n
o
f Co
n
ten
t B
a
s
e
d
Med
ica
l I
ma
g
e
R
etri
ev
a
l U
s
in
g
…
(
S
w
ee
ty
Ma
n
ia
r
)
371
=
−
∑
∑
(
,
)
l
og
(
,
)
(5
)
=
∑
∑
(
−
)
2
(
,
)
(6
)
3
.
3
.
2
.
Sh
a
pe
Sh
ap
e
f
ea
tu
r
es
h
a
v
e
a
s
ig
n
i
f
ic
an
t r
o
le
in
p
r
i
m
ar
y
g
r
o
u
p
o
f
m
ed
ical
i
m
a
g
es b
ased
o
n
th
eir
c
o
n
ten
t [
2
]
.
Featu
r
es
s
u
c
h
as
A
r
ea
,
E
d
g
e,
Fo
u
r
ier
Descr
ip
to
r
,
C
ir
cu
lar
it
y
,
ar
e
u
s
ed
to
r
etr
iev
e
m
ed
ical
i
m
ag
e
s
[
1
4
,
8
]
.
A
r
ea
:
A
r
ea
o
f
s
elec
tio
n
i
n
s
q
u
ar
e
p
ix
els o
r
in
ca
lib
r
ated
s
q
u
a
r
e
u
n
it
s
.
E
d
g
e:
Usi
n
g
ca
n
n
y
ed
g
e
d
etec
to
r
,
g
r
ad
ien
t,
an
d
o
th
er
o
p
er
at
o
r
s
.
Fo
u
r
ier
Descr
ip
to
r
:
Fo
u
r
ier
Descr
ip
to
r
s
(
FDs
)
is
a
p
o
w
er
f
u
l
f
ea
t
u
r
e
f
o
r
b
o
u
n
d
ar
ies an
d
o
b
jects r
ep
r
esen
tatio
n
.
(
)
=
∑
(
)
e
xp
[
−
2
]
−
1
=
0
,
0
≤
≤
−
1
(7
)
Dis
cr
ete
Fo
u
r
ier
T
r
an
s
f
o
r
m
o
f
z(
n
)
(
b
o
u
n
d
ar
y
p
o
in
t)
g
i
v
es
v
a
lu
e
o
f
Fo
u
r
ier
Descr
ip
to
r
.
=
4
(
2
)
(8
)
E
q
u
iv
ale
n
ce
d
ia
m
eter
(
cir
cle
w
it
h
s
a
m
e
ar
ea
as t
h
e
r
eg
io
n
)
=
√
4
∗
(9
)
4.
CL
AS
SI
F
I
CAT
I
O
N
AL
G
O
RIT
H
M
4
.
1
.
Neura
l N
et
w
o
rk
A
r
ti
ficial
n
eu
r
al
n
et
w
o
r
k
m
o
d
els
h
av
e
b
ee
n
s
tu
d
ied
f
o
r
m
a
n
y
y
ea
r
s
i
n
t
h
e
h
o
p
e
o
f
ac
h
ie
v
i
n
g
h
u
m
an
-
lik
e
p
er
f
o
r
m
an
ce
in
s
e
v
er
al
field
s
s
u
ch
as
s
p
ee
ch
a
n
d
i
m
a
g
e
u
n
d
er
s
tan
d
i
n
g
.
T
h
e
n
et
w
o
r
k
s
ar
e
co
m
p
o
s
ed
o
f
m
an
y
n
o
n
li
n
ea
r
co
m
p
u
ta
tio
n
al
ele
m
e
n
ts
o
p
er
atin
g
in
p
a
r
allel
an
d
ar
r
an
g
ed
in
p
atte
r
n
s
r
e
m
i
n
i
s
ce
n
t
o
f
b
io
lo
g
ical
n
eu
r
al
n
et
w
o
r
k
s
.
C
o
m
p
u
ta
tio
n
al
ele
m
en
ts
o
r
n
o
d
es
ar
e
co
n
n
ec
ted
in
s
ev
er
al
la
y
er
s
(
in
p
u
t,
h
id
d
en
an
d
o
u
tp
u
t)
v
ia
w
ei
g
h
ts
t
h
at
ar
e
ty
p
icall
y
a
d
ap
ted
d
u
r
in
g
t
h
e
tr
ain
i
n
g
p
h
ase
to
ac
h
iev
e
h
i
g
h
p
er
f
o
r
m
an
ce
.
I
n
s
tead
o
f
p
er
f
o
r
m
in
g
a
s
et
o
f
i
n
s
tr
u
ctio
n
s
s
eq
u
e
n
tiall
y
a
s
i
n
a
Vo
n
N
eu
m
a
n
n
co
m
p
u
ter
,
n
eu
r
al
n
et
w
o
r
k
m
o
d
els
ex
p
lo
r
e
s
i
m
u
lta
n
eo
u
s
l
y
m
a
n
y
h
y
p
o
th
eses
u
s
i
n
g
p
ar
allel
n
et
w
o
r
k
s
co
m
p
o
s
ed
o
f
m
an
y
co
m
p
u
tatio
n
a
l
ele
m
en
t
s
co
n
n
ec
ted
b
y
li
n
k
s
w
it
h
v
ar
ia
b
le
w
e
ig
h
t.
T
h
e
ar
c
h
itect
u
r
e
o
f
t
h
e
n
eu
r
al
n
et
w
o
r
k
co
n
s
i
s
t
s
o
f
th
r
ee
la
y
er
s
:
a
n
in
p
u
t
la
y
er
,
a
h
id
d
en
o
n
e
a
n
d
an
o
u
tp
u
t
la
y
er
.
T
h
e
n
u
m
b
er
o
f
n
o
d
es
in
th
e
in
p
u
t
la
y
er
i
s
eq
u
al
to
t
h
e
n
u
m
b
er
o
f
ele
m
e
n
ts
e
x
i
s
ti
n
g
i
n
o
n
e
tr
a
n
s
ac
tio
n
in
t
h
e
d
atab
ase.
W
h
il
e
th
e
o
u
tp
u
t la
y
er
w
as c
o
n
s
is
ti
n
g
o
f
o
n
e
n
o
d
e.
T
h
e
n
o
d
e
o
f
th
e
o
u
tp
u
t
la
y
er
is
th
e
o
n
e
t
h
a
t
g
iv
e
s
th
e
c
lass
i
fi
ca
tio
n
f
o
r
th
e
i
m
a
g
e.
I
t
clas
s
i
fi
es
it
a
s
n
o
r
m
al
o
r
ab
n
o
r
m
al.
I
n
th
e
tr
ain
i
n
g
p
h
ase,
th
e
in
ter
n
al
w
ei
g
h
t
s
o
f
t
h
e
n
eu
r
al
n
et
w
o
r
k
ar
e
ad
j
u
s
ted
ac
co
r
d
in
g
to
th
e
tr
a
n
s
ac
tio
n
s
u
s
ed
in
th
e
lear
n
i
n
g
p
r
o
ce
s
s
.
Fo
r
ea
ch
tr
ain
i
n
g
tr
an
s
ac
tio
n
th
e
n
e
u
r
al
n
et
w
o
r
k
r
ec
ei
v
es
i
n
ad
d
itio
n
th
e
ex
p
ec
ted
o
u
tp
u
t.
T
h
is
allo
w
s
th
e
m
o
d
i
fi
ca
tio
n
o
f
th
e
w
ei
g
h
ts
.
I
n
t
h
e
n
ex
t
s
te
p
,
th
e
tr
ain
ed
n
e
u
r
al
n
et
w
o
r
k
i
s
u
s
ed
to
clas
s
i
f
y
n
e
w
i
m
a
g
es.
Fig
u
r
e
2
.
Neu
r
al
Net
w
o
r
k
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8814
IJ
AA
S
Vo
l.
6
,
No
.
4
,
Dec
em
b
er
201
7
:
3
6
8
–
37
4
372
4
.
2
.
CART
T
h
e
class
if
icatio
n
a
n
d
r
eg
r
es
s
io
n
tr
ee
(
C
AR
T
)
alg
o
r
ith
m
is
m
a
in
l
y
u
s
ed
f
o
r
th
e
clas
s
i
f
icatio
n
o
f
d
if
f
er
e
n
t
tis
s
u
e
s
in
i
m
a
g
e
m
in
i
n
g
,
w
h
ich
i
s
o
n
th
e
b
asis
o
f
s
ev
er
al
d
er
iv
ed
p
ar
am
et
er
s
.
T
h
e
r
ec
u
r
s
iv
e
p
ar
titi
o
n
in
g
m
et
h
o
d
u
s
ed
i
n
th
e
C
AR
T
alg
o
r
ith
m
to
in
tr
o
d
u
ce
th
e
tr
ee
b
ased
m
o
d
ell
in
g
w
h
ich
i
s
later
co
n
v
er
ted
to
t
h
e
s
tati
s
tical
m
ain
s
tr
ea
m
.
T
o
s
elec
t
th
e
o
p
tim
al
tr
ee
v
a
lu
e
t
h
e
al
g
o
r
ith
m
in
v
o
lv
e
s
t
h
e
cr
o
s
s
v
alid
atio
n
s
c
h
e
m
e
f
r
o
m
s
o
m
e
r
ig
o
r
o
u
s
ap
p
r
o
ac
h
es.
B
ase
d
o
n
th
e
tech
n
iq
u
e
ca
lled
s
u
r
r
o
g
ate
s
p
lits
t
h
e
alg
o
r
i
th
m
a
u
to
m
at
icall
y
h
a
n
d
les
th
e
m
i
s
s
i
n
g
v
al
u
es.
Fo
r
e
x
a
m
p
le
t
h
e
v
ar
iab
le
(
x
=t1
)
is
s
elec
ted
th
e
n
th
e
g
r
ea
test
s
ep
ar
atio
n
i
s
p
r
o
d
u
ce
d
s
o
(
x
=t1
)
is
s
aid
to
b
e
s
p
lit.
I
f
th
i
s
v
ar
iab
le
X
it
s
en
d
s
to
w
h
ic
h
is
les
s
th
a
n
t1
th
en
t
h
e
d
ata
is
s
e
n
d
to
lef
t
o
r
else
it
s
en
d
s
to
r
ig
h
t.
T
h
e
p
r
o
c
ess
is
r
ep
ea
ted
f
o
r
all
th
e
n
o
d
es.
So
th
at
it
is
ea
s
y
to
co
n
clu
d
e
th
at
C
A
R
T
alg
o
r
it
h
m
u
s
e
s
o
n
l
y
t
h
e
b
in
ar
y
s
p
lits
.
4
.
3
.
K
-
M
ea
ns
K
-
Me
a
n
s
al
g
o
r
ith
m
is
s
aid
to
b
e
an
u
n
s
u
p
er
v
i
s
ed
clu
s
ter
in
g
alg
o
r
ith
m
.
I
t
w
o
r
k
s
w
ell
f
o
r
n
u
m
er
ica
l
d
ata
alo
n
e.
T
h
e
p
ix
el
-
by
-
p
ix
el
i
m
a
g
e
cla
s
s
i
f
icatio
n
is
p
o
s
s
ib
le
b
y
d
ef
i
n
i
n
g
s
i
n
g
le
a
n
d
m
u
lt
ip
le
th
r
es
h
o
ld
s
.
So
th
at
h
is
to
g
r
a
m
s
tatis
tics
i
s
u
s
ed
in
th
i
s
alg
o
r
it
h
m
f
o
r
t
h
e
p
ix
el
b
ased
clas
s
if
icatio
n
.
T
h
e
m
ai
n
w
o
r
k
o
f
t
h
is
p
r
o
ce
s
s
is
to
ch
ec
k
w
h
et
h
er
t
h
e
h
is
to
g
r
a
m
is
b
i
m
o
d
al
o
r
n
o
t.
I
f
it
is
th
e
n
th
e
g
r
a
y
v
al
u
e
w
ill
b
e
ap
p
ea
r
ed
o
th
er
w
is
e
t
h
e
i
m
a
g
es
g
et
p
ar
titi
o
n
ed
i
n
to
s
e
v
er
al
r
eg
io
n
s
.
T
h
e
th
r
esh
o
ld
o
f
g
r
a
y
v
al
u
e
ca
n
b
e
d
eter
m
i
n
ed
u
s
i
n
g
th
e
p
ea
k
v
al
u
es.
Ho
w
e
v
er
it
co
n
v
er
g
e
s
o
n
l
y
t
h
e
lo
ca
l
m
in
i
m
u
m
v
al
u
es.
So
th
e
alg
o
r
i
th
m
in
v
o
l
v
es
n
u
m
b
er
o
f
clu
s
ter
s
f
o
r
th
e
o
p
tim
izatio
n
.
4
.
4
.
Na
iv
e
B
a
y
es
T
h
e
Naiv
e
b
ay
es
al
g
o
r
ith
m
is
th
e
m
o
s
t
p
o
w
er
f
u
l
tech
n
iq
u
e.
I
t
d
o
es
th
e
test
in
g
p
r
o
ce
s
s
ea
s
i
l
y
an
d
th
e
class
i
f
icatio
n
p
r
o
b
le
m
s
ca
n
b
e
s
o
lv
ed
.
I
t
ca
n
b
e
ab
le
to
b
u
ild
a
m
o
d
el
f
a
s
tl
y
a
n
d
g
iv
in
g
b
et
ter
p
r
ed
ictio
n
s
.
T
o
f
i
n
d
th
e
m
is
s
i
n
g
d
ata
t
h
e
n
aï
v
e
b
a
y
es
alg
o
r
it
h
m
p
la
y
s
a
m
aj
o
r
r
o
le.
T
h
e
u
n
s
ee
n
d
ata
ca
n
b
e
ea
s
il
y
p
r
ed
icted
b
y
ch
ar
ac
ter
izi
n
g
t
h
e
p
r
o
b
le
m
in
n
a
ïv
e
b
a
y
es
m
eth
o
d
.
D
u
r
i
n
g
t
h
e
co
n
s
tr
u
c
tio
n
ti
m
e
a
n
d
p
r
ed
ictio
n
ti
m
e
t
h
is
alg
o
r
ith
m
s
ep
ar
a
tes
t
h
e
at
tr
ib
u
tes
v
alu
e.
T
h
e
p
r
o
b
ab
ilit
y
o
f
e
ac
h
attr
ib
u
tes
in
is
o
latio
n
p
r
o
ce
s
s
n
ee
d
s
o
n
l
y
t
h
e
en
o
u
g
h
d
ata.
So
,
th
er
e
is
n
o
n
ee
d
o
f
m
o
r
e
d
ata
co
llectio
n
in
th
i
s
alg
o
r
ith
m
.
Fi
n
all
y
,
if
th
e
d
ata
h
as
h
i
g
h
co
r
r
elate
d
f
ea
tu
r
es th
e
p
er
f
o
r
m
an
ce
w
i
ll b
e
d
eg
r
ad
ed
.
4
.
5
.
Dec
is
io
n T
re
e
Dec
is
io
n
tr
ee
alg
o
r
ith
m
is
o
n
e
o
f
th
e
class
i
f
ier
tech
n
iq
u
e
wh
ich
i
s
in
th
e
f
o
r
m
o
f
tr
ee
s
tr
u
ctu
r
e.
Fo
r
class
i
f
icatio
n
a
n
d
p
r
ed
ictio
n
,
t
h
e
p
o
w
er
f
u
l
to
o
ls
ar
e
av
a
ilab
l
e
in
th
is
al
g
o
r
it
h
m
.
I
t
h
as
f
o
u
r
d
iv
is
io
n
s
s
u
ch
as
Dec
is
io
n
n
o
d
e,
lea
f
n
o
d
e,
ed
g
e
a
n
d
p
ath
.
A
s
i
n
g
le
attr
ib
u
t
e
is
r
ep
r
esen
ted
in
th
e
d
ec
is
i
o
n
n
o
d
e.
L
ea
f
n
o
d
e
d
ef
in
e
s
t
h
e
tar
g
et
attr
ib
u
te.
S
p
litt
in
g
o
f
o
n
e
at
tr
ib
u
te
i
s
ed
g
e
an
d
t
h
e
p
ath
is
a
f
in
al
d
ec
is
io
n
.
Fo
r
co
n
ti
n
u
o
u
s
attr
ib
u
te
th
is
al
g
o
r
ith
m
is
n
o
t a
p
p
licab
le
5.
CO
M
P
ARIT
I
O
N
O
F
CL
AS
SI
F
I
CA
T
I
O
N
AL
G
O
RI
T
H
M
I
n
th
i
s
p
ar
t,
th
e
co
m
p
ar
ativ
e
r
e
s
u
lt
s
an
d
t
h
e
d
a
tasets
ar
e
lis
te
d
f
o
r
th
e
d
ata
m
in
in
g
al
g
o
r
ith
m
s
.
T
h
e
ac
cu
r
ac
y
o
f
v
ar
io
u
s
al
g
o
r
it
h
m
s
is
clea
r
l
y
n
o
ted
in
t
h
is
s
tu
d
y
.
5
.
1
.
Da
t
a
s
et
Descript
io
n
Var
io
u
s
i
m
ag
e
d
atase
ts
h
e
lp
s
t
o
f
in
d
th
e
cla
s
s
i
f
icat
io
n
p
er
f
o
r
m
an
ce
o
f
d
ata
m
i
n
i
n
g
al
g
o
r
ith
m
s
.
T
h
e
u
s
ed
d
ata
s
ets ar
e
s
h
o
w
n
i
n
ta
b
le
1
.
T
ab
le
1.
Data
s
et
C
o
m
p
ar
is
o
n
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
AA
S
I
SS
N:
2252
-
8814
C
la
s
s
i
fica
tio
n
o
f Co
n
ten
t B
a
s
e
d
Med
ica
l I
ma
g
e
R
etri
ev
a
l U
s
in
g
…
(
S
w
ee
ty
Ma
n
ia
r
)
373
5
.
2
.
Co
m
pa
riso
n o
f
Da
t
a
M
ini
ng
Alg
o
rit
h
m
s
T
h
is
p
ar
t lis
ts
o
u
t t
h
e
p
o
s
itiv
e
an
d
n
eg
a
tiv
e
a
s
p
ec
ts
u
s
ed
in
v
ar
io
u
s
alg
o
r
it
h
m
s
p
r
esen
t i
n
t
h
is
p
ap
er
f
o
r
th
e
d
ata
m
i
n
i
n
g
al
g
o
r
ith
m
.
T
ab
le
2
.
A
lg
o
r
ith
m
C
o
m
p
ar
is
o
n
s
6.
CL
AS
SI
F
I
CAT
I
O
N
P
ARA
M
E
T
E
R
T
h
e
co
n
f
u
s
io
n
m
atr
ix
ca
n
b
e
u
s
ed
to
d
eter
m
i
n
e
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
s
y
s
te
m
.
T
h
is
m
a
tr
ix
d
escr
ib
es
all
p
o
s
s
ib
le
o
u
tco
m
e
s
o
f
a
p
r
ed
ictio
n
r
esu
lts
i
n
tab
le
s
t
r
u
ctu
r
e.
T
h
e
p
o
s
s
ib
le
o
u
tco
m
es
o
f
a
t
w
o
cla
s
s
p
r
ed
ictio
n
b
e
r
ep
r
esen
ted
as
T
r
u
e
p
o
s
iti
v
e
(
T
P
)
,
T
r
u
e
n
eg
ati
v
e
(
T
N)
,
Fals
e
P
o
s
iti
v
e
(
FP
)
a
n
d
Fal
s
e
Ne
g
ati
v
e
(
FN)
.
T
h
e
n
o
r
m
al
a
n
d
ab
n
o
r
m
al
i
m
ag
e
s
ar
e
co
r
r
ec
tly
c
lass
i
f
ied
as
T
r
u
e
P
o
s
itiv
e
an
d
T
r
u
e
Neg
ativ
e
r
esp
ec
tiv
el
y
.
A
Fals
e
P
o
s
iti
v
e
is
w
h
e
n
th
e
o
u
tco
m
e
i
s
in
co
r
r
ec
tl
y
clas
s
if
ied
as
p
o
s
itiv
e
wh
en
it
i
s
a
n
e
g
ati
v
e.
Fals
e
P
o
s
iti
v
e
is
th
e
Fal
s
e
a
lar
m
i
n
th
e
clas
s
i
f
icatio
n
p
r
o
ce
s
s
.
A
f
alse
n
e
g
ati
v
e
is
w
h
en
t
h
e
o
u
tco
m
e
i
s
in
co
r
r
ec
tl
y
p
r
ed
icted
as n
eg
at
i
v
e
w
h
e
n
it s
h
o
u
ld
h
av
e
b
ee
n
i
n
f
ac
t p
o
s
iti
v
e.
I
n
o
u
r
s
y
s
te
m
co
n
s
id
er
,
T
P
=
Nu
m
b
er
o
f
A
b
n
o
r
m
al
i
m
ag
es c
o
r
r
ec
tl
y
clas
s
i
f
ied
T
N=
Nu
m
b
er
o
f
No
r
m
al
i
m
a
g
es c
o
r
r
ec
tly
cla
s
s
i
f
ied
FP
=
Nu
m
b
er
o
f
No
r
m
al
i
m
a
g
es c
lass
i
f
ied
as
A
b
n
o
r
m
al
FN=
Nu
m
b
er
o
f
A
b
n
o
r
m
al
i
m
ag
es c
las
s
i
f
ied
as No
r
m
al.
a.
P
r
ec
is
io
n
:
T
h
e
f
r
ac
tio
n
o
f
a
b
n
o
r
m
al
i
m
a
g
es
w
i
th
co
r
r
ec
t r
esu
lts
.
T
P
/(
FP
+
T
P
)
b.
Sen
s
iti
v
it
y
(
R
ec
all)
:
T
h
e
p
r
o
b
ab
ilit
y
o
f
t
h
e
test
f
i
n
d
in
g
t
h
e
ab
n
o
r
m
al
ca
s
e
a
m
o
n
g
all
ab
n
o
r
m
al
ca
s
e
s
.
T
P
/(
FN+T
P
)
c.
Sp
ec
if
icit
y
:
T
h
e
p
r
o
b
ab
ilit
y
o
f
th
e
test
f
i
n
d
in
g
th
e
n
o
r
m
al
ca
s
e
a
m
o
n
g
all
n
o
r
m
al
ca
s
e
s
.
T
N/(
FP
+
T
N)
d.
A
cc
u
r
ac
y
:
T
h
e
f
r
ac
tio
n
o
f
tes
t r
esu
lt
s
th
o
s
e
ar
e
co
r
r
ec
t.
(
T
N+
T
P
)
/FP
+
T
N+
FN+
T
P
7.
CO
NCLU
SI
O
N
I
n
th
is
s
t
u
d
y
,
t
h
e
o
v
er
all
p
er
f
o
r
m
an
ce
o
f
v
ar
io
u
s
al
g
o
r
ith
m
s
p
r
esen
t i
n
t
h
is
p
ap
er
w
a
s
an
a
l
y
ze
d
b
ased
o
n
t
h
e
clas
s
i
f
icatio
n
ac
c
u
r
ac
y
.
T
h
e
ac
cu
r
ac
y
i
n
i
m
a
g
e
cl
ass
i
f
icatio
n
is
t
h
e
m
ai
n
id
ea
o
f
e
v
al
u
ati
n
g
th
e
p
er
f
o
r
m
a
n
ce
in
d
ata
m
i
n
i
n
g
alg
o
r
ith
m
s
.
T
h
e
o
v
er
all
r
e
s
u
lt
s
h
o
w
n
i
n
th
is
p
ap
er
is
s
tep
in
to
f
u
r
t
h
er
d
ev
elo
p
m
en
t
i
n
f
u
t
u
r
e
tec
h
n
o
lo
g
y
.
T
o
ev
al
u
ate
th
e
b
est
i
n
d
icatio
n
s
c
lin
ical
s
tu
d
ie
s
ar
e
m
o
r
e
es
s
en
t
ial.
T
h
is
p
ap
er
p
r
o
j
ec
ted
s
ev
er
al
p
ar
am
eter
s
w
h
ic
h
is
u
s
ed
b
y
d
ata
m
in
i
n
g
m
et
h
o
d
s
.
B
y
co
m
p
ar
i
n
g
th
ese
al
g
o
r
ith
m
s
,
n
eu
r
al
n
et
w
o
r
k
r
es
u
lt
s
s
h
o
w
s
b
etter
p
er
f
o
r
m
a
n
ce
a
m
o
n
g
th
e
o
th
er
m
et
h
o
d
s
p
r
esen
ted
i
n
th
is
w
o
r
k
an
d
it
i
s
also
g
i
v
es
th
e
b
est
cla
s
s
i
f
icati
o
n
ac
cu
r
ac
y
as
w
ell
as
s
a
v
es
t
h
e
co
m
p
u
ti
n
g
ti
m
e.
I
n
f
u
t
u
r
e,
w
e
h
a
v
e
p
lan
n
ed
to
en
h
a
n
ce
n
e
u
r
al
n
et
w
o
r
k
m
et
h
o
d
to
im
p
r
o
v
e
th
e
clas
s
i
f
icati
o
n
ac
cu
r
ac
y
,
s
en
s
iti
v
it
y
,
s
p
ec
i
f
icit
y
a
n
d
as
w
e
ll
as
p
r
o
ce
s
s
in
g
t
i
m
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8814
IJ
AA
S
Vo
l.
6
,
No
.
4
,
Dec
em
b
er
201
7
:
3
6
8
–
37
4
374
RE
F
E
R
E
NC
E
S
[1
]
C.
L
a
k
s
h
m
i
D
e
v
a
s
e
n
a
,
T
.
S
u
ma
t
h
i
,
D
r
.
M
.
H
e
m
a
l
a
t
ha
“
A
n
E
xp
e
r
i
e
n
t
i
a
l
S
u
r
v
e
y
o
n
I
ma
ge
M
i
n
i
n
g
T
oo
l
s
,
T
e
c
hn
i
qu
e
s
a
n
d
A
pp
li
c
a
ti
o
n
s
”
I
n
t
e
r
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
n
C
o
m
pu
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
ee
r
i
n
g
(I
J
C
S
E
),
V
o
l
.
3,
N
o
.
3,
p
p
.
5
0
6
1
-
5
0
6
7
,
2
0
1
1
.
[2
]
E
.
A
r
n
o
l
d
i
,
M
.
G
e
b
r
e
g
z
i
a
b
h
e
r
,
U
.
J
.
S
c
h
o
e
p
f
,
R
.
G
o
l
d
e
n
b
e
r
g,
L
.
R
a
m
o
s
-
D
u
r
a
n
,
P
.
L
.
Z
w
e
rn
e
r
,
K
.
N
i
k
o
la
o
u,
M
.
F
.
R
e
i
s
e
r
,
P
.
C
o
s
t
e
ll
o
,
C.
T
h
il
o
“
A
u
t
o
ma
t
e
d
c
o
m
pu
t
e
r
ai
d
e
d
s
t
e
n
o
s
i
s
d
e
te
c
t
i
o
n
a
t
c
o
r
o
n
a
r
y
CT
a
n
g
i
o
g
r
a
p
h
y
:
i
n
iti
a
l
e
xp
e
r
i
e
n
ce
”
,
El
s
e
v
i
e
r
,
V
o
l
.
2
0
,
N
o
.
5,
p
p
.
1
1
6
0
-
1
1
6
7
,
2
0
1
0
.
[3
]
L
e
f
k
o
v
i
t
s
S
z
,
L
e
f
t
k
o
v
it
s
L
.
“
E
nh
a
n
ce
d
G
a
b
o
r
F
ilt
e
r
B
a
s
e
d
F
a
c
ia
l
F
e
at
u
r
e
D
e
t
ec
t
o
r
,
T
h
e
p
r
o
ce
e
d
i
n
g
of
t
h
e
E
u
r
o
p
e
a
n
I
n
t
e
gr
a
ti
o
n
-
B
e
t
w
e
e
n
T
r
a
d
i
o
n
a
n
d
m
o
d
e
r
n
it
y
”
,
El
s
e
v
i
e
r
,
c
o
m
pu
t
e
r
s
c
i
e
n
c
e
s
ec
t
i
o
n
,
vo
l
.
1,
p
p
.
7
8
7
3
8
-
8
8
6
3
,
2
0
1
3
.
[4
]
R
.
Ve
n
k
a
t
a
R
a
m
a
n
a
C
h
a
r
y
,
D
r
.
D
.
R
a
j
y
a
L
a
k
s
h
m
i
a
n
d
D
r
.
K
.
V.
N
S
u
n
i
t
h
a
“
F
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
m
e
t
h
o
d
s
f
o
r
c
o
l
o
r
im
a
g
e
sim
il
a
rit
y
”
,
A
d
v
a
n
c
e
d
Co
m
p
u
ti
n
g
:
An
I
n
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
(
ACIJ
),
Vo
l.
3
,
N
o
.
2
,
p
p
.
1
4
7
-
1
5
7
,
2
0
1
2
.
[5
]
A
n
k
i
t
A
g
r
a
w
al
,
S
a
n
c
h
i
t
M
i
s
r
a
,
R
am
a
n
a
t
h
a
n
N
a
r
a
y
a
n
a
n
,
L
a
li
t
h
P
o
l
e
p
e
dd
i
,
A
l
o
k
C
h
o
ud
h
a
r
y
,
“
A
L
u
n
g
C
a
n
ce
r
O
u
tc
o
m
e
C
al
c
u
la
t
o
r
Us
i
n
g
E
n
s
e
m
b
l
e
D
at
a
M
i
n
i
n
g
o
n
S
EE
R
D
a
t
a
”
,
B
I
O
KD
D
2
0
1
1
,
N
o
.
5,
2
0
1
1
.
[6
]
S
.
B
a
la
n
a
n
d
T
.
D
e
v
i
,
“
D
e
s
i
gn
a
n
d
D
e
v
e
l
op
m
e
n
t
of
a
n
A
l
go
r
it
hm
fo
r
I
ma
ge
C
l
u
s
t
e
r
i
n
g
I
n
T
e
x
t
il
e
I
ma
ge
R
e
t
r
i
e
v
a
l
Us
i
n
g
C
o
l
o
r
D
e
s
c
r
i
p
t
o
r
s
”
,
I
n
t
e
rn
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
C
o
m
pu
te
r
S
c
i
e
n
ce
,
V
o
l
.
2,
N
o
.
3,
p
p
.
19
9
-
2
1
1
,
2
0
1
2
.
[7
]
V
am
s
i
d
h
a
r
E
n
i
r
e
d
d
y
,
K
i
r
a
n
K
u
ma
r
R
e
dd
i
,
“
A
D
at
a
M
i
n
i
n
g
A
pp
r
o
a
c
h
fo
r
C
o
m
p
r
e
s
s
e
d
M
e
d
i
c
a
l
I
ma
ge
R
e
t
r
i
e
v
a
l”
,
I
n
t
e
rn
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
C
o
m
pu
t
e
r
A
pp
li
c
a
ti
o
n
s
(
0
9
7
5
–
88
7
)
V
o
l
.
52,
N
o
.
5,
2
0
1
2
.
[8
]
D
.
D
e
s
h
p
a
n
d
e
,
“
A
ss
o
c
ia
t
i
o
n
R
u
l
e
M
i
n
i
n
g
B
a
s
e
d
o
n
I
ma
ge
C
o
n
t
e
n
t”
,
I
n
t
e
rn
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
I
n
f
o
r
ma
ti
o
n
T
ec
hn
o
l
o
g
y
a
n
d
K
n
o
w
l
e
d
g
e
M
a
n
a
g
e
m
e
n
t
,
V
o
l
.
4
,
N
o
.
1,
pp.
14
3
-
1
4
6
,
2
0
1
1
.
[9
]
L
i
-
H
o
n
g
J
u
a
n
g
,
M
i
n
g
-
N
i
W
u
“
M
R
I
b
r
ai
n
l
e
s
i
o
n
ima
ge
d
e
te
c
t
i
o
n
b
a
s
e
d
o
n
c
o
l
o
r
-
c
o
n
v
e
r
t
e
d
K
-
m
e
a
n
s
c
l
u
s
t
e
r
i
n
g
s
e
g
m
e
n
ta
ti
o
n
”
,
e
l
s
e
v
i
e
r
,
c
o
m
pu
t
e
r
s
c
i
e
n
c
e
s
e
c
t
i
o
n
,
vo
l
.
43
,
N
o
.
7
,
pp.
94
1
-
9
4
9
,
2
0
1
0
.
[1
0
]
S
.
L
.
A
.
Le
e
a
,
A
.
Z
.
K
o
u
za
n
ia
,
E
.
J
.
H
u
b
“
R
a
n
d
o
m
f
o
r
e
s
t
b
a
s
e
d
l
u
n
g
n
o
du
l
e
c
la
s
s
i
f
i
c
a
ti
o
n
ai
d
e
d
by
c
l
u
s
t
e
r
i
n
g
”
,
e
l
s
e
v
i
e
r
,
c
o
m
pu
t
e
r
s
c
i
e
n
c
e
s
e
c
t
i
o
n
,
vo
l
.
3
4
,
N
o
.
7,
p
p
.
53
5
-
5
4
2
,
2
0
1
0
.
[1
1
]
M
a
h
n
a
z
Et
e
h
a
d
Ta
v
a
k
o
l
,
V
i
n
o
d
C
h
a
n
d
r
a
n
,
E
.
Y
.
K
.
N
g
,
R
a
h
e
l
e
h
K
a
f
i
e
h
,
“
B
r
e
a
s
t
c
a
n
ce
r
d
e
t
ec
t
i
o
n
f
r
o
m
t
h
e
r
m
a
l
i
m
a
g
e
s
u
s
i
n
g
b
i
s
p
e
c
t
r
a
l
i
n
v
a
r
i
a
nt
f
e
at
u
r
e
s
”
,
e
l
s
e
v
i
e
r
,
I
n
t
e
r
n
a
t
i
o
n
a
l
j
o
u
r
n
a
l
o
f
t
h
e
r
ma
l
s
c
i
e
n
ce
s
,
v
o
l
.
69
,
p
p
.
2
1
-
3
6
,
2
0
1
3
.
[1
2
]
M
i
n
g
-
Y
i
h
L
e
e
a
,
C
h
i
-
S
h
i
h
Y
a
n
g
,
“
E
n
t
r
o
p
y
-
b
a
s
e
d
f
e
at
u
r
e
e
x
t
r
a
c
ti
o
n
a
n
d
d
ec
i
s
i
o
n
t
r
e
e
i
n
du
c
ti
o
n
f
o
r
b
r
e
a
s
t
c
a
n
ce
r
d
ia
g
n
o
s
i
s
w
it
h
s
t
a
n
d
a
r
d
iz
e
d
t
h
e
r
m
o
g
r
a
p
h
i
ma
g
e
s
”
,
e
l
s
e
v
i
e
r
,
c
o
m
pu
t
e
r
s
c
i
e
n
c
e
m
e
t
h
o
ds
a
n
d
p
r
o
g
r
am
s
i
n
b
i
o
m
e
d
i
c
i
n
e
, v
o
l
.
100,
N
o
.
3,
p
p
.
26
9
-
2
8
2
,
2
0
1
0
.
[1
3
]
Y
e
C
h
e
n
a
,
J
udd
S
t
o
rr
s
,
L
i
r
o
n
g
T
a
n
a
,
L
a
w
r
e
n
c
e
J
.
M
a
z
l
a
c
k
c
,
J
i
n
g
-
H
u
e
i
L
e
e
b
,
L
o
n
g
J
.
L
u
a
,
“
D
e
t
ec
ti
n
g
b
r
ai
n
s
t
r
u
c
t
u
r
a
l
c
h
a
n
g
e
s
a
s
b
i
o
ma
r
k
e
r
f
r
o
m
ma
g
n
e
ti
c
r
e
s
o
n
a
n
c
e
i
m
a
g
e
s
u
s
i
n
g
a
l
o
c
a
l
f
e
a
t
u
r
e
b
a
s
e
d
SV
M
a
pp
r
o
a
c
h
”
,
e
l
s
e
v
i
e
r
,
J
o
u
r
n
a
l
of
N
e
u
r
o
s
c
i
e
n
c
e
m
e
t
h
o
d
s
,
v
o
l
.
2
2
1
,
N
o
.
15
,
p
p
.
2
2
-
3
1
,
20
1
4.
[1
4
]
We
n
-
J
i
e
W
u
a
,
S
h
i
h
-
We
i
L
i
n
a
,
W
o
o
K
y
u
n
g
M
o
o
n
b
,
“
C
o
m
b
i
n
i
n
g
s
u
p
p
o
r
t
v
e
c
t
o
r
ma
c
h
i
n
e
w
it
h
g
e
n
e
ti
c
al
g
o
r
it
hm
t
o
c
la
ss
i
fy
u
lt
r
a
s
o
u
n
d
b
r
e
a
s
t
t
u
m
o
r
ima
g
e
s
”
,
e
l
s
e
v
i
e
r
,
c
o
m
pu
te
r
s
c
i
e
n
c
e
s
e
c
t
i
o
n
,
vo
l
.
36
,
N
o
.
8
,
p
p
.
62
7
-
6
3
3
,
2
0
1
2
.
[1
5
]
J
a
r
i
c
h
W
.
S
p
li
et
h
o
ff
,
D
a
n
i
e
l
J
.
E
v
e
r
,
H
o
u
k
e
M
.
K
l
o
m
p
,
J
o
h
a
n
n
a
W
.
v
a
n
S
a
n
d
i
c
k,
M
i
c
h
e
l
W
.
W
o
u
t
e
r
s
,
R
a
m
i
N
a
c
h
a
b
e
,
G
e
r
al
d
W
.
L
u
c
a
s
s
e
n
,
B
e
nn
o
H
.
W
.
H
e
n
d
r
i
k
s
,
J
e
ll
e
We
s
s
e
li
n
g
,
T
h
e
o
J
.
M
.
R
u
e
r
s
,
”
I
m
p
r
o
v
e
d
i
d
e
n
ti
f
i
c
at
i
o
n
o
f
p
e
r
i
p
h
e
r
a
l
l
u
n
g
t
u
m
o
r
s
by
u
s
i
n
g
d
i
f
f
u
s
e
r
e
f
le
c
ta
n
c
e
a
n
d
f
l
u
o
r
e
s
ce
n
c
e
s
p
ec
t
r
o
s
c
o
p
y
”
,
e
l
s
e
v
i
e
r
,
c
o
m
pu
t
e
r
s
c
i
e
n
c
e
s
e
c
ti
o
n
,
v
o
l
.
80
,
N
o
.
2
,
p
p
.
16
5
-
1
7
1
,
2
0
1
3
.
[1
6
]
M
i
n
-
C
h
u
n
Y
a
n
g
,
C
h
i
u
n
-
S
h
e
n
g
H
u
a
n
g
,
“
W
h
o
l
e
b
r
e
a
s
t
l
e
s
i
o
n
d
e
t
ec
ti
o
n
u
s
i
n
g
n
ai
v
e
b
a
y
e
s
c
la
ss
i
f
i
e
r
fo
r
p
o
r
ta
b
l
e
u
l
t
r
a
s
o
u
n
d
”
,
e
l
s
e
v
i
e
r
,
c
o
m
pu
t
e
r
s
c
i
e
n
c
e
s
ec
ti
o
n
,
vo
l
.
3
8
,
N
o
.
1
1
,
p
p
.
1
8
7
0
-
18
8
0
,
2
0
1
2
.
[1
7
]
S
h
e
n
g
j
u
n
Z
h
o
u,
Y
u
a
n
z
hi
C
h
e
n
g
,
S
h
i
n
i
c
h
i
T
a
m
u
r
a
,
“
A
u
t
o
ma
t
e
d
l
u
n
g
s
e
g
m
e
n
t
a
ti
o
n
a
n
d
s
m
oo
t
h
i
n
g
t
ec
hn
i
qu
e
s
fo
r
i
n
c
l
u
s
i
o
n
o
f
j
ux
ta
p
l
e
u
r
a
l
n
o
du
l
e
s
a
n
d
p
u
lm
o
n
a
r
y
v
e
ss
e
l
s
o
n
c
h
e
s
t
CT
i
m
a
g
e
s
”
,
e
l
s
e
v
i
e
r
,
c
o
m
pu
t
e
r
s
c
i
e
n
c
e
s
ec
ti
o
n
, v
o
l
.
1
3
,
p
p
.
6
2
-
7
0
,
2
0
1
4.
[1
8
]
[U
R
a
v
i
B
a
b
u,
Y
.
V
e
n
k
a
s
w
a
r
l
u,
A
n
ee
l
K
u
ma
r
C
h
i
n
t
h
a
.
“
H
a
n
d
w
r
i
t
t
e
n
D
i
g
i
t
R
e
c
o
g
n
i
t
i
o
n
U
s
i
ng
K
-
N
ea
r
e
s
t
N
e
i
g
h
bo
u
r
C
la
ss
i
f
i
e
r”
I
E
E
E
W
o
r
l
d
C
o
n
g
r
e
s
s
o
n
C
o
m
pu
ti
n
g
a
n
d
C
o
mm
un
i
c
a
t
i
o
n
T
ec
h
n
o
l
o
g
i
e
s
I
S
B
N
:
97
8
-
1
-
4
7
9
9
-
287
6
-
7
p
p
6
0
-
6
5
,
2
0
1
4.
[1
9
]
U
R
a
v
i
B
a
b
u,
Y
.
V
e
n
k
a
s
w
a
r
l
u
,
A
n
ee
l
K
u
ma
r
C
h
i
n
t
h
a
.
“
H
a
n
d
w
r
i
t
t
e
n
D
i
g
i
t
R
e
c
o
g
n
i
t
i
o
n
U
s
i
ng
K
-
N
ea
r
e
s
t
N
e
i
g
h
bo
u
r
C
la
ss
i
f
i
e
r”
I
E
E
E
W
o
r
l
d
C
o
n
g
r
e
s
s
o
n
C
o
m
pu
ti
n
g
a
n
d
C
o
mm
un
i
c
a
t
i
o
n
T
ec
h
n
o
l
o
g
i
e
s
I
S
B
N
:
97
8
-
1
-
4
7
9
9
-
287
6
-
7
p
p
6
0
-
6
5
,
2
0
1
4.
[2
0
]
U
m
ut
K
o
n
u
r
,
F
i
k
r
e
t
S
.
G
u
r
g
e
n,
“
C
o
m
pu
t
e
r
ai
d
e
d
d
e
t
e
c
ti
o
n
o
f
s
p
i
na
b
i
f
i
da
u
s
i
ng
n
ea
r
e
s
t
n
e
i
g
h
bo
r
cl
a
ss
i
f
i
c
a
ti
o
n
w
i
t
h
c
u
r
v
at
u
r
e
s
ca
l
e
s
p
a
c
e
f
e
at
u
r
e
s
o
f
f
e
ta
l
s
ku
l
l
s
e
x
t
r
a
c
t
e
d
f
r
o
m
u
l
t
r
a
s
ou
n
d
i
ma
g
e
s
,
vo
l
.
8
5
,
p
p
.
8
0
-
9
5
,
2
0
1
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.