C
omp
u
te
r
S
c
i
e
n
c
e
an
d
I
n
for
mati
on
T
e
c
h
n
ol
ogi
e
s
V
ol
.
7
,
N
o
.
1
,
M
a
r
c
h
20
26
,
p
p.
1
11
~
1
20
IS
S
N
:
2722
-
3221
,
D
O
I
:
10.
1
1591
/
c
s
i
t
.
v
7
i
1
.
pp
1
11
-
1
20
111
Jou
r
n
al
h
o
m
e
pa
ge
:
ht
t
p:
/
/
i
ae
s
pr
i
m
e
.
c
om
/
i
nd
e
x
.
php
/
c
s
i
t
A
d
v
a
n
c
e
s
i
n
d
e
r
m
a
t
o
l
o
g
i
c
a
l
i
ma
g
i
n
g
:
e
n
h
a
n
c
i
n
g
sk
i
n
me
l
a
n
o
m
a
c
l
a
ssi
f
i
c
a
t
i
o
n
f
o
r
i
mpr
o
v
e
d
p
a
t
i
e
n
t
o
u
t
c
o
me
s
D
e
b
ad
u
tta
S
ah
oo
,
S
ou
mya
M
i
s
h
r
a
D
e
p
a
rt
m
e
n
t
o
f
E
l
e
c
t
ro
n
i
c
s
a
n
d
Co
m
m
u
n
i
c
a
t
i
o
n
E
n
g
i
n
e
e
ri
n
g
,
C.
V
.
Ra
m
a
n
G
l
o
b
a
l
U
n
i
v
e
rs
i
t
y
,
Bh
u
b
a
n
e
s
w
a
r,
In
d
i
a
A
r
ti
c
l
e
I
n
fo
A
BS
TR
A
C
T
Ar
t
i
c
l
e
h
i
s
t
or
y
:
Re
c
e
i
v
e
d
S
e
p
18
,
2
025
Re
vi
s
e
d
D
e
c
5
,
2
025
A
c
c
e
pt
e
d
J
a
n
1
,
2026
T
he
s
t
ud
y
pr
e
s
e
nt
s
a
n
e
nh
a
nc
e
d
A
l
e
xN
e
t
-
ba
s
e
d
de
e
p
l
e
a
r
ni
ng
s
ys
t
e
m
f
o
r
bi
na
r
y
c
l
a
s
s
i
f
i
c
a
t
i
o
n
o
f
m
e
l
a
no
m
a
s
k
i
n
c
a
nc
e
r
a
s
e
i
t
h
e
r
be
ni
g
n
o
r
m
a
l
i
gna
n
t
us
i
n
g
t
w
o
pa
i
r
e
d
de
r
m
a
t
o
s
c
opi
c
a
nd
c
l
i
n
i
c
a
l
i
m
a
ge
da
t
a
s
e
t
s
.
T
he
s
t
u
dy
e
va
l
ua
t
e
s
t
he
r
e
s
i
l
i
e
nc
e
o
f
t
he
m
od
e
l
s
a
c
r
o
s
s
di
f
f
e
r
e
nt
i
m
a
ge
s
e
t
s
w
i
t
h
c
om
m
on
p
r
e
p
r
o
c
e
s
s
i
ng
a
nd
s
pe
c
i
f
i
c
d
a
t
a
a
u
gm
e
n
t
a
t
i
o
n,
us
i
ng
a
m
e
l
a
n
om
a
da
t
a
s
e
t
c
on
t
a
i
ni
ng
1
0,
000
i
m
a
g
e
s
a
nd
a
be
n
i
gn
ve
r
s
u
s
m
a
l
i
gna
nt
da
t
a
s
e
t
w
i
t
h
3,
60
0
i
m
a
ge
s
.
T
he
A
l
e
xN
e
t
r
e
f
i
ne
m
e
n
t
e
xc
e
e
de
d
s
e
ve
r
a
l
s
t
a
nda
r
d
m
a
c
h
i
ne
l
e
a
r
n
i
ng
(
M
L
)
c
l
a
s
s
i
f
i
e
r
s
a
nd
o
t
he
r
de
e
p
a
r
c
h
i
t
e
c
t
ur
e
s
on
t
he
t
w
o
d
a
t
a
s
e
t
s
w
i
t
h
pr
a
c
t
i
c
a
l
t
r
a
i
n
i
ng
t
i
m
e
s
,
ga
i
n
i
ng
97.
12
%
a
nd
96.
21
%
i
n
ba
l
a
n
c
e
d
a
c
c
u
r
a
c
y.
T
h
e
t
r
a
i
n
i
ng
pr
oc
e
e
de
d
w
i
t
h
S
G
D
a
s
o
pt
i
m
i
s
e
r
a
nd
c
r
o
s
s
-
e
nt
r
opy
a
s
l
o
s
s
on
256×
2
56
i
m
a
g
e
s
.
B
e
nc
h
m
a
r
k
i
ng
a
ga
i
ns
t
s
upp
or
t
v
e
c
t
o
r
m
a
c
h
i
ne
(
S
V
M
)
,
k
-
ne
a
r
e
s
t
n
e
i
g
hbo
ur
(
K
N
N
)
,
a
nd
o
t
he
r
c
on
vol
ut
i
on
a
l
ne
u
r
a
l
ne
t
w
o
r
k
s
(
C
N
N
s)
de
s
i
g
ns
s
h
ow
s
t
ha
t
t
he
s
e
l
e
c
t
e
d
a
r
c
hi
t
e
c
t
ur
e
a
nd
hype
r
pa
r
a
m
e
t
e
r
s
a
c
hi
e
ve
d
t
h
e
h
i
g
he
s
t
pe
r
f
o
r
m
a
nc
e
o
n
c
os
t
-
e
f
f
e
c
t
i
ve
c
om
pu
t
a
t
i
o
n
f
o
r
t
he
r
ou
t
i
ne
m
e
l
a
n
o
m
a
t
r
i
a
g
e
.
T
he
r
e
po
r
t
hi
ghl
i
g
ht
s
t
he
ne
e
d
f
or
e
x
t
e
r
na
l
va
l
i
da
t
i
on,
i
n
c
o
r
po
r
a
t
i
o
n
i
nt
o
de
r
m
a
t
ol
ogi
c
a
l
w
o
r
k
f
l
ow
s
,
a
nd
e
xpl
a
i
na
b
i
l
i
t
y
t
o
i
m
pr
o
ve
t
r
us
t
,
d
i
m
i
ni
s
h
da
t
a
s
e
t
b
i
a
s
,
a
nd
s
up
por
t
t
h
e
s
a
f
e
c
l
i
ni
c
a
l
de
p
l
oy
m
e
nt
i
n
pr
a
c
t
i
c
e
.
Ke
y
w
or
d
s
:
A
l
e
x
ne
t
a
rc
hi
t
e
c
t
ur
e
D
e
e
p
l
e
a
rn
i
ng
D
i
a
gnos
t
i
c
a
c
c
ura
c
y
M
e
di
c
a
l
i
m
a
ge
a
na
l
ys
i
s
M
e
l
a
nom
a
de
t
e
c
t
i
on
S
ki
n
c
a
n
c
e
r
c
l
a
s
s
i
fi
c
a
t
i
on
T
hi
s
i
s
an
op
e
n
ac
c
e
s
s
ar
t
i
c
l
e
u
nde
r
t
he
C
C
B
Y
-
SA
l
i
c
e
n
s
e
.
Cor
r
e
s
pon
di
n
g
Au
t
h
or
:
D
e
ba
du
t
t
a
S
a
hoo
D
e
pa
r
t
m
e
nt
of
E
l
e
c
t
r
oni
c
s
a
nd
Com
m
un
i
c
a
t
i
on
E
ng
i
n
e
e
r
i
ng
,
C.
V
.
Ra
m
a
n
G
l
oba
l
U
n
i
ve
rs
i
t
y
Bhuba
n
e
s
w
a
r
752
054
,
Ind
i
a
E
m
a
i
l
:
de
b
a
du
t
t
a
s
a
hoo
2255
@
gm
a
i
l
.
c
o
m
1.
I
N
TR
O
D
U
C
TI
O
N
S
ki
n
c
a
n
c
e
r
i
s
one
of
t
h
e
m
os
t
pr
e
va
l
e
n
t
t
um
ours
w
or
l
dw
i
d
e
,
a
nd
m
e
l
a
no
m
a
t
y
pi
c
a
l
l
y
m
a
ke
s
ne
w
s
due
t
o
i
t
s
r
a
pi
d
s
pre
a
d
a
nd
hi
gh
de
a
t
h
ra
t
e
.
M
i
l
l
i
ons
of
n
e
w
c
a
s
e
s
of
s
k
i
n
c
a
n
c
e
r
a
r
e
re
p
ort
e
d
by
t
he
W
H
O
e
a
c
h
y
e
a
r
,
unde
rs
c
or
i
ng
t
he
i
m
por
t
a
n
c
e
of
e
a
rl
y
de
t
e
c
t
i
on
a
n
d
e
ff
e
c
t
i
v
e
t
re
a
t
m
e
nt
.
If
t
h
e
i
ni
t
i
a
l
s
y
m
pt
om
s
a
re
di
s
re
g
a
rd
e
d,
m
e
l
a
nom
a
,
w
hi
c
h
b
e
gi
ns
i
n
pi
g
m
e
n
t
-
pro
duc
i
ng
m
e
l
a
no
c
yt
e
s
,
c
a
n
t
urn
f
a
t
a
l
[
1],
[2]
.
D
oc
t
ors
now
ha
ve
s
ha
rp
e
r
op
t
i
c
s
a
n
d
a
d
di
t
i
on
a
l
e
ye
s
t
o
d
e
t
e
c
t
s
ki
n
i
s
s
u
e
s
t
h
a
nks
t
o
e
x
c
i
t
i
ng
a
dv
a
n
c
e
m
e
n
t
s
i
n
m
e
di
c
a
l
i
m
a
gi
ng
a
nd
m
a
c
hi
n
e
l
e
a
rn
i
ng
(M
L
)
t
oo
l
s
[3]
.
O
n
t
he
o
t
he
r
h
a
nd,
r
out
i
ne
i
ns
pe
c
t
i
ons
t
h
a
t
m
os
t
l
y
r
e
l
y
o
n
a
s
pe
c
i
a
l
i
s
t
'
s
gl
a
nc
e
m
a
y
oc
c
a
s
i
ona
l
l
y
ov
e
rl
oo
k
or
m
i
s
r
e
a
d
s
m
a
l
l
s
i
gns
.
In
t
hi
s
c
a
s
e
,
de
e
p
l
e
a
r
ni
ng
t
e
c
hni
q
ue
s
—
pa
rt
i
c
u
l
a
r
l
y
c
o
nvol
u
t
i
o
na
l
n
e
ur
a
l
n
e
t
w
orks
,
or
CN
N
s
—
fi
l
l
t
ha
t
n
e
e
d
a
s
t
rus
t
w
or
t
hy
h
e
l
p
e
rs
f
or
s
k
i
n
pho
t
o
t
a
gg
i
ng
[2]
,
[4]
,
[5
]
.
T
he
popul
a
r
A
l
e
xN
e
t
a
rc
h
i
t
e
c
t
ure
ha
s
a
l
re
a
dy
be
e
n
m
odi
f
i
e
d
for
t
hi
s
s
pe
c
i
f
i
c
t
a
s
k
;
i
t
w
a
s
onc
e
re
v
ol
u
t
i
on
a
ry
i
n
r
out
i
ne
pi
c
t
ur
e
s
e
a
rc
he
s
[1
]
.
T
h
e
c
urr
e
nt
e
ffor
t
fo
c
us
e
s
on
t
w
o
pri
m
a
ry
c
a
t
e
gor
i
e
s
of
s
ki
n
c
a
n
c
e
r:
t
h
e
os
t
e
ns
i
bl
y
b
e
n
i
gn
but
ob
vi
ous
l
y
c
a
nc
e
ro
us
m
ol
e
s
.
T
he
t
e
a
m
a
i
m
s
t
o
i
n
c
r
e
a
s
e
d
i
a
g
nos
i
s
ra
t
e
s
w
hi
l
e
m
a
i
nt
a
i
ni
ng
a
m
a
na
ge
a
bl
e
l
e
ve
l
of
fa
l
s
e
a
l
a
rm
s
by
o
pt
i
m
i
s
i
n
g
a
pr
e
t
r
a
i
n
e
d
v
e
rs
i
on
o
f
A
l
e
xN
e
t
.
In
ord
e
r
t
o
i
m
p
rove
d
e
rm
a
t
ol
og
i
s
t
s
'
c
l
i
n
i
c
a
l
pr
a
c
t
i
c
e
s
ki
l
l
s
,
t
he
c
urre
n
t
s
t
u
dy
i
nt
e
nds
t
o
a
dd
t
o
t
h
i
s
e
xp
a
ndi
ng
body
of
know
l
e
dge
by
s
how
c
a
s
i
ng
how
w
e
l
l
t
h
e
A
l
e
xN
e
t
m
ode
l
c
l
a
s
s
i
fi
e
s
s
ki
n
c
a
n
c
e
r
[1]
,
[2]
,
[6
]
–
[10]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2722
-
3221
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
,
V
o
l
.
7
,
N
o
.
1
,
M
a
rc
h
20
26
:
111
-
1
20
112
2.
R
ELA
TED
WO
R
K
T
he
a
dv
a
nc
e
m
e
n
t
s
i
n
m
e
l
a
no
m
a
d
e
t
e
c
t
i
o
n
ha
ve
r
e
s
ul
t
e
d
i
n
a
s
urge
i
n
c
a
nc
e
r
d
i
a
g
nos
t
i
c
c
om
put
i
ng
re
s
e
a
rc
h
.
T
h
e
f
i
rs
t
s
ys
t
e
m
s
u
t
i
l
i
z
e
d
c
l
a
s
s
i
c
a
l
i
m
a
g
e
a
n
a
l
ys
i
s
a
nd
M
L
pr
o
c
e
dur
e
s
i
n
w
h
i
c
h
f
e
a
t
ur
e
s
l
i
k
e
c
o
l
o
ur
a
nd
t
e
x
t
u
re
,
a
s
w
e
l
l
a
s
l
e
s
i
on
s
ha
p
e
,
w
e
re
c
re
a
t
e
d
b
y
d
om
a
i
n
e
xp
e
rt
s
a
n
d
c
l
a
s
s
i
f
i
e
d
b
y
s
u
ppo
rt
v
e
c
t
o
r
m
a
c
h
i
n
e
s
(S
V
M
s
)
,
k
-
n
e
a
r
e
s
t
n
e
i
ghb
ours
(
KNNs
)
,
d
e
c
i
s
i
o
n
t
r
e
e
s
,
ra
n
do
m
f
or
e
s
t
s
,
na
i
v
e
B
a
y
e
s
,
a
nd
fun
c
t
i
on
a
l
l
i
n
k
a
r
t
i
fi
c
i
a
l
n
e
u
ra
l
n
e
t
w
or
k
(
F
L
A
N
N
)
[
11]
,
[
12]
.
T
h
e
s
e
t
e
c
hn
i
q
ue
s
,
w
hi
l
e
ba
s
e
l
i
ne
s
w
e
re
us
e
ful
,
s
uf
fe
r
e
d
f
rom
poor
pe
r
form
a
n
c
e
d
ue
t
o
c
ons
i
de
r
a
b
l
e
i
nt
r
a
-
c
l
a
s
s
v
a
ri
a
bi
l
i
t
y
a
nd
i
m
a
ge
no
i
s
e
.
T
h
e
d
e
a
d
i
m
a
g
e
s
w
e
re
oft
e
n
pre
s
e
n
t
i
n
l
ow
-
q
ua
l
i
t
y
d
e
r
m
a
t
os
c
op
i
c
a
nd
c
l
i
n
i
c
a
l
pho
t
ogr
a
p
hs
.
T
h
e
p
r
e
c
i
s
e
a
n
d
t
i
m
e
l
y
d
i
a
g
no
s
i
s
o
f
m
e
l
a
n
o
m
a
a
nd
s
k
i
n
l
e
s
i
o
n
c
a
t
e
g
or
i
z
a
t
i
o
n
,
w
h
i
c
h
d
e
e
p
l
e
a
rn
i
ng
e
x
c
e
l
s
i
n
a
na
l
y
z
i
n
g
,
c
a
m
e
fr
o
m
d
e
p
l
o
y
a
b
l
e
CN
N
s
.
O
v
e
r
t
h
e
y
e
a
rs
,
s
t
u
di
e
s
h
a
v
e
do
c
u
m
e
n
t
e
d
a
l
m
os
t
e
v
e
r
y
a
r
c
h
i
t
e
c
t
u
r
e
t
o
b
e
c
o
m
p
e
t
i
n
g
w
i
t
h
a
d
e
r
m
a
t
ol
o
gi
s
t
i
n
p
e
r
f
or
m
a
n
c
e
,
i
f
no
t
ou
t
p
e
r
fo
r
m
i
n
g
t
h
e
m
[1
3
]
–
[1
8]
.
T
yp
i
c
a
l
CN
N
s
,
L
e
N
e
t
,
V
G
G
16,
R
e
s
N
e
t
,
G
oog
L
e
N
e
t
,
In
c
e
pt
i
on,
G
A
N
s
,
a
nd
t
h
e
re
c
e
nt
t
r
a
ns
for
m
e
r
-
i
ns
p
i
re
d
n
e
t
w
o
rks
ha
v
e
a
l
l
b
e
e
n
s
t
udi
e
d.
T
he
s
e
a
ppr
oa
c
he
s
,
how
e
ve
r
,
re
q
ui
r
e
not
onl
y
a
w
e
a
l
t
h
of
a
nno
t
a
t
e
d
d
a
t
a
s
e
t
s
a
nd
hi
gh
-
e
n
d
c
o
m
pu
t
e
rs
bu
t
a
l
s
o
o
ft
e
n
dra
w
up
on
c
ons
i
de
r
a
b
l
e
l
e
a
rni
ng
t
o
fi
n
a
l
i
z
e
a
s
ol
ut
i
on
a
n
d
c
a
pt
ur
e
i
m
a
ge
s
w
i
t
h
c
o
m
p
l
e
x
v
i
s
ua
l
p
a
t
t
e
rns
t
ha
t
w
oul
d
be
a
l
m
os
t
i
m
pos
s
i
b
l
e
t
o
de
s
c
r
i
be
m
a
nu
a
l
l
y
[
2],
[19
]
.
S
e
ve
r
a
l
w
orks
c
ont
i
nu
e
t
o
b
e
s
l
ow
e
d
dow
n
be
c
a
us
e
of
t
h
e
s
a
m
e
pr
a
c
t
i
c
a
l
i
s
s
u
e
s
t
ha
t
m
a
k
e
t
ra
ns
l
a
t
i
on
c
ha
l
l
e
ng
i
ng:
l
i
m
i
t
e
d
a
c
c
e
s
s
t
o
d
i
ve
rs
e
a
nd
w
e
l
l
-
a
nno
t
a
t
e
d
d
e
r
m
a
t
os
c
op
i
c
i
m
a
g
e
s
,
p
ri
v
a
c
y
a
nd
re
gul
a
t
ory
c
ons
t
ra
i
nt
s
of
m
e
d
i
c
a
l
da
t
a
s
h
a
ri
ng
,
a
n
d
di
ff
i
c
u
l
t
i
e
s
i
n
bui
l
di
n
g
c
os
t
l
y
a
nd
s
c
a
l
a
bl
e
da
t
a
i
nge
s
t
i
on
,
a
nnot
a
t
i
on
,
t
ra
i
ni
n
g,
a
n
d
v
a
l
i
da
t
i
on
p
i
pe
l
i
n
e
s
.
Cons
e
que
nt
l
y,
m
ore
r
e
c
e
nt
w
orks
h
a
v
e
d
e
v
e
l
op
e
d
ne
w
m
o
de
l
s
,
but
a
l
s
o
a
t
t
e
m
pt
e
d
t
o
bu
i
l
d
t
he
re
qui
s
i
t
e
i
nfra
s
t
ruc
t
ur
e
a
nd
l
e
ga
l
fr
a
m
e
w
orks
t
h
a
t
w
ou
l
d
e
n
a
bl
e
t
h
e
us
e
of
A
I
i
n
c
l
i
ni
c
a
l
pr
a
c
t
i
c
e
w
hi
l
e
s
a
fe
g
ua
rd
i
ng
a
no
nym
i
t
y
[20]
–
[22]
.
H
ow
e
ve
r
,
m
os
t
of
t
h
e
a
va
i
l
a
bl
e
a
nd
de
v
e
l
o
pe
d
re
s
e
a
rc
h
c
on
t
i
nue
s
t
o
e
x
a
m
i
ne
a
ve
r
y
n
a
rrow
r
a
ng
e
of
ML
b
a
s
e
l
i
ne
s
,
or
a
s
i
ngu
l
a
r
d
e
e
p
l
e
a
rn
i
ng
a
r
c
hi
t
e
c
t
ur
e
,
a
nd
t
ypi
c
a
l
l
y
,
us
i
ng
a
s
i
ngl
e
d
a
t
a
s
e
t
w
i
t
h
d
i
ff
e
re
n
t
d
a
t
a
p
re
pa
r
a
t
i
on
/
e
v
a
l
u
a
t
i
on
m
e
t
hod
ol
og
i
e
s
[23]
,
[24
]
.
Cons
e
qu
e
nt
l
y,
t
h
e
r
e
i
s
l
i
t
t
l
e
e
vi
d
e
n
c
e
t
o
s
up
port
t
h
e
op
t
i
m
a
l
m
od
e
l
s
a
nd
t
he
t
r
a
de
-
off
b
e
t
w
e
e
n
s
t
ru
c
t
u
re
,
fl
e
x
i
bi
l
i
t
y,
a
n
d
c
os
t
,
a
n
d
t
he
r
e
a
re
l
i
t
t
l
e
e
v
i
d
e
nc
e
a
nd
g
ui
d
a
n
c
e
a
v
a
i
l
a
b
l
e
t
o
phys
i
c
i
a
ns
t
o
h
e
l
p
t
h
e
m
s
e
l
e
c
t
t
he
m
os
t
a
pp
ropri
a
t
e
t
e
c
hno
l
og
i
e
s
for
c
l
i
ni
c
a
l
m
e
l
a
nom
a
s
c
re
e
n
i
ng
[25
]
.
T
he
c
urre
nt
w
ork
a
ddr
e
s
s
e
s
t
hi
s
g
a
p
by
pe
rfor
m
i
ng
a
m
e
t
hod
i
c
a
l
,
s
i
de
-
by
-
s
i
d
e
c
o
m
pa
ri
s
on
und
e
r
t
h
e
s
a
m
e
e
xpe
r
i
m
e
n
t
a
l
fr
a
m
e
w
ork
a
m
on
g
s
e
ve
r
a
l
c
o
nve
nt
i
ona
l
a
nd
d
e
e
p
l
e
a
rni
ng
a
ppro
a
c
h
e
s
f
or
s
k
i
n
c
a
nc
e
r
c
l
a
s
s
i
fi
c
a
t
i
on
.
T
h
i
s
a
rt
i
c
l
e
a
s
s
e
s
s
e
s
a
fi
n
e
-
t
une
d
A
l
e
xN
e
t
a
ga
i
ns
t
m
ul
t
i
pl
e
c
l
a
s
s
i
c
ML
c
l
a
s
s
i
fi
e
rs
,
s
u
c
h
a
s
S
V
M
,
K
N
N
,
r
a
ndo
m
for
e
s
t
,
a
n
d
fa
s
t
l
i
br
a
ry
for
a
pp
roxi
m
a
t
e
ne
a
r
e
s
t
n
e
i
gh
bours
,
a
nd
ot
h
e
r
d
e
e
p
ne
t
w
orks
,
i
nc
l
udi
n
g
a
s
t
a
nd
a
rd
CN
N
,
L
e
N
e
t
,
V
G
G
16,
Re
s
N
e
t
,
G
oo
gL
e
N
e
t
,
a
nd
a
g
e
n
e
ra
t
i
v
e
a
dv
e
rs
a
r
i
a
l
n
e
t
w
or
k
-
ba
s
e
d
m
ode
l
,
a
d
opt
i
ng
a
c
ohe
r
e
n
t
pr
e
pro
c
e
s
s
i
ng
,
t
ra
i
ni
n
g,
a
nd
t
e
s
t
i
ng
proc
e
dur
e
on
t
w
o
c
o
m
pl
e
m
e
n
t
a
ry
pub
l
i
c
l
y
a
va
i
l
a
bl
e
d
a
t
a
s
e
t
s
:
t
he
m
e
l
a
no
m
a
s
ki
n
c
a
n
c
e
r
d
a
t
a
s
e
t
,
c
on
s
i
s
t
i
ng
of
10
,
000
i
m
a
g
e
s
,
a
n
d
t
he
s
ki
n
c
a
n
c
e
r
m
a
l
i
gn
a
nt
vs
.
b
e
ni
gn
d
a
t
a
s
e
t
.
T
he
a
rt
i
c
l
e
a
l
l
ow
s
for
a
m
o
re
c
o
m
p
l
e
t
e
b
e
n
c
hm
a
rk
b
y
j
o
i
nt
l
y
c
ons
i
d
e
ri
ng
a
c
c
ura
c
y
,
A
U
C,
pr
e
c
i
s
i
on
,
re
c
a
l
l
,
F
1
-
s
c
ore
,
a
nd
t
ra
i
ni
ng
t
i
m
e
on
bot
h
da
t
a
s
e
t
s
a
nd
a
l
s
o
d
e
m
ons
t
r
a
t
e
s
t
h
a
t
a
n
opt
i
m
i
z
e
d
A
l
e
xN
e
t
c
a
n
a
c
hi
e
v
e
s
up
e
ri
or
di
a
gn
os
t
i
c
p
e
rfor
m
a
nc
e
w
h
i
l
e
s
t
i
l
l
b
e
i
ng
c
o
m
pu
t
a
t
i
o
na
l
l
y
vi
a
bl
e
fo
r
i
nt
e
gra
t
i
o
n
i
nt
o
pr
a
c
t
i
c
a
l
d
e
rm
a
t
ol
og
y
w
ork
fl
ow
s
[26]
.
3.
O
BJEC
TI
V
ES
E
va
l
ua
t
e
de
e
p
l
e
a
rni
ng
t
e
c
hni
que
s
:
t
o
a
s
s
e
s
s
t
he
e
f
fe
c
t
i
ve
ne
s
s
of
de
e
p
l
e
a
r
ni
ng
a
rc
hi
t
e
c
t
ure
s
,
pa
rt
i
c
ul
a
rl
y
t
he
A
l
e
x
N
e
t
m
ode
l
,
i
n
c
l
a
s
s
i
fyi
ng
s
ki
n
c
a
n
c
e
r
from
m
e
di
c
a
l
i
m
a
ge
s
.
Com
pa
re
w
i
t
h
ML
m
ode
l
s
:
t
o
c
om
pa
re
t
he
pe
rfo
rm
a
nc
e
of
t
he
A
l
e
xN
e
t
m
o
de
l
a
ga
i
ns
t
t
ra
di
t
i
ona
l
ML
c
l
a
s
s
i
fi
e
rs
,
s
uc
h
a
s
S
V
M
,
K
N
N
,
a
nd
ra
ndom
fore
s
t
s
,
i
n
t
e
rm
s
o
f
a
c
c
ura
c
y,
pre
c
i
s
i
on,
re
c
a
l
l
,
a
nd
F
1
s
c
ore
.
U
t
i
l
i
z
e
di
ve
r
s
e
da
t
a
s
e
t
s
:
t
o
ut
i
l
i
z
e
t
w
o
di
s
t
i
nc
t
da
t
a
s
e
t
s
,
na
m
e
l
y
t
he
m
e
l
a
nom
a
s
ki
n
c
a
nc
e
r
da
t
a
s
e
t
a
n
d
t
he
s
ki
n
c
a
nc
e
r:
m
a
l
i
gna
nt
vs
.
be
ni
gn
da
t
a
s
e
t
,
t
o
e
ns
ure
a
c
om
pre
he
n
s
i
ve
e
va
l
ua
t
i
on
of
t
he
m
ode
l
'
s
pe
rf
orm
a
nc
e
a
c
ros
s
va
ryi
ng
i
m
a
ge
c
ha
ra
c
t
e
ri
s
t
i
c
s
.
O
pt
i
m
i
z
e
m
ode
l
pe
rform
a
nc
e
:
t
o
fi
ne
-
t
une
t
he
A
l
e
x
N
e
t
m
ode
l
by
opt
i
m
i
z
i
n
g
hype
rpa
ra
m
e
t
e
r
s
,
i
nc
l
udi
ng
l
e
a
rni
n
g
ra
t
e
,
ba
t
c
h
s
i
z
e
,
a
nd
i
m
a
ge
pre
proc
e
s
s
i
n
g
t
e
c
hni
que
s
,
t
o
a
c
hi
e
ve
t
h
e
hi
ghe
s
t
po
s
s
i
bl
e
c
l
a
s
s
i
fi
c
a
t
i
on
a
c
c
ura
c
y
a
nd
e
f
fi
c
i
e
nc
y.
V
i
s
ua
l
i
z
e
re
s
ul
t
s
:
t
o
p
re
s
e
nt
t
he
pe
r
form
a
nc
e
m
e
t
ri
c
s
a
nd
m
ode
l
a
rc
hi
t
e
c
t
ure
t
hr
ough
fl
ow
c
ha
rt
s
a
n
d
gr
a
phs
,
fa
c
i
l
i
t
a
t
i
ng
a
c
l
e
a
r
u
nde
r
s
t
a
ndi
n
g
of
t
h
e
m
ode
l
'
s
s
t
r
uc
t
ure
a
nd
i
t
s
c
l
a
s
s
i
fi
c
a
t
i
on
c
a
pa
bi
l
i
t
i
e
s
.
Co
nt
ri
b
ut
e
t
o
c
l
i
ni
c
a
l
a
ppl
i
c
a
t
i
ons
:
t
o
c
ont
ri
b
ut
e
t
o
t
he
fi
e
l
d
of
de
rm
a
t
ol
ogy
by
pr
ovi
di
n
g
i
n
s
i
ght
s
i
nt
o
t
he
pot
e
nt
i
a
l
of
A
I
-
dri
ve
n
di
a
gn
os
t
i
c
t
o
ol
s
for
e
a
rl
y
de
t
e
c
t
i
on
a
nd
a
c
c
ura
t
e
c
l
a
s
s
i
fi
c
a
t
i
on
o
f
m
e
l
a
nom
a
s
ki
n
c
a
nc
e
r,
ul
t
i
m
a
t
e
l
y
a
i
m
i
ng
t
o
e
nha
n
c
e
pa
t
i
e
nt
out
c
om
e
s
[2
7]
.
4.
M
ET
H
O
D
O
L
O
G
Y
In
t
h
i
s
s
t
ud
y,
t
h
e
A
l
e
xN
e
t
a
rc
h
i
t
e
c
t
ure
—
a
pre
t
r
a
i
n
e
d
d
e
e
p
l
e
a
rni
n
g
m
ode
l
—
w
a
s
us
e
d
a
nd
re
fi
n
e
d
f
or
t
he
b
i
na
r
y
c
l
a
s
s
i
f
i
c
a
t
i
on
of
m
e
l
a
no
m
a
s
k
i
n
c
a
n
c
e
r
i
nt
o
be
n
i
g
n
a
nd
m
a
l
i
gn
a
nt
g
roups
[6]
,
[7]
.
T
h
e
s
i
z
e
t
ha
t
t
h
e
Evaluation Warning : The document was created with Spire.PDF for Python.
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
IS
S
N
:
2722
-
3221
A
d
v
an
c
e
s
i
n
de
r
m
a
t
ol
o
gi
c
al
i
m
ag
i
ng:
e
nhan
c
i
ng
s
k
i
n
m
e
l
a
n
om
a
c
l
as
s
i
f
i
c
a
t
i
o
n
f
or
…
(D
e
b
adut
t
a
Sahoo
)
113
ne
t
w
ork
re
qui
r
e
s
,
256
b
y
256
pi
x
e
l
s
,
w
a
s
a
pp
l
i
e
d
t
o
e
a
c
h
p
i
c
t
ure
t
h
a
t
w
a
s
pus
h
e
d
i
nt
o
t
he
pi
pe
l
i
n
e
.
A
fa
s
t
t
w
o
-
uni
t
l
i
n
e
a
r
l
a
y
e
r
t
ha
t
s
pi
t
s
e
i
t
he
r
be
ni
gn
or
m
a
l
i
gna
nt
m
a
t
e
ri
a
l
w
a
s
us
e
d
t
o
c
om
pl
e
t
e
t
he
t
a
s
k
,
re
p
l
a
c
i
ng
t
he
l
a
s
t
ful
l
y
c
o
nne
c
t
e
d
l
a
ye
r
[1]
,
[6]
,
[
7]
.
T
o
e
ns
u
r
e
s
m
o
o
t
h
a
nd
q
ui
c
k
l
e
a
r
n
i
n
g
,
pi
x
e
l
s
w
e
r
e
t
h
e
n
no
r
m
a
l
i
s
e
d
us
i
n
g
t
h
e
t
r
a
d
i
t
i
on
a
l
I
m
a
g
e
N
e
t
m
e
a
n
a
n
d
s
t
a
nd
a
rd
d
e
v
i
a
t
i
o
n
v
a
l
u
e
s
(
0
.
48
5
,
0
.
4
56
,
0
.
4
06
)
/
(0
.
22
9
,
0
.
22
4
,
0
.
2
25
)
.
S
t
oc
h
a
s
t
i
c
gra
d
i
e
n
t
de
s
c
e
nt
(S
G
D
)
a
t
0
.
001
w
i
t
h
0.
9
m
om
e
nt
u
m
w
a
s
us
e
d
for
t
ra
i
n
i
ng
[1],
[28
]
;
t
h
i
s
p
a
i
r
a
c
c
e
l
e
r
a
t
e
s
pro
gre
s
s
w
h
i
l
e
c
on
t
rol
l
i
n
g
no
i
s
y
j
u
m
ps
.
D
uri
n
g
t
r
a
i
n
i
ng
,
c
ros
s
-
e
nt
rop
y
l
os
s
w
a
s
us
e
d
a
s
t
h
e
obj
e
c
t
i
ve
fu
nc
t
i
on
t
o
m
e
a
s
ur
e
c
l
a
s
s
i
f
i
c
a
t
i
on
m
i
s
t
a
k
e
s
e
ff
i
c
i
e
nt
l
y
[28]
.
T
og
e
t
h
e
r
,
t
h
e
s
e
hyp
e
rp
a
ra
m
e
t
e
rs
a
nd
a
rc
h
i
t
e
c
t
ura
l
m
od
i
fi
c
a
t
i
ons
a
l
l
ow
e
d
t
he
A
l
e
xN
e
t
m
o
de
l
t
o
a
t
t
a
i
n
hi
gh
c
l
a
s
s
i
fi
c
a
t
i
on
a
c
c
u
ra
c
y
w
i
t
h
re
l
i
a
bl
e
t
ra
i
ni
ng
re
s
u
l
t
s
a
nd
r
e
a
l
i
s
t
i
c
e
x
e
c
u
t
i
on
t
i
m
e
s
,
e
s
t
a
bl
i
s
h
i
ng
i
t
a
s
a
r
e
l
i
a
bl
e
m
e
t
h
od
for
di
a
gnos
i
n
g
m
e
l
a
no
m
a
s
ki
n
c
a
n
c
e
r
[6]
,
[7]
,
[
29]
.
5.
D
A
TA
S
ET
D
ETA
I
LS
T
he
f
i
rs
t
c
o
l
l
e
c
t
i
on
of
d
e
r
m
a
t
os
c
op
i
c
i
m
a
ge
s
i
s
t
a
ke
n
f
rom
K
a
ggl
e
’s
m
e
l
a
nom
a
s
ki
n
c
a
n
c
e
r
d
a
t
a
s
e
t
of
1
0,
00
0
i
m
a
g
e
s
,
w
h
i
c
h
h
a
s
a
s
i
z
e
of
3
00
×
3
00
pi
x
e
l
s
[6]
.
T
h
e
m
o
de
l
h
a
s
a
l
o
t
of
v
i
s
u
a
l
d
e
t
a
i
l
t
o
w
or
k
w
i
t
h
b
e
c
a
us
e
t
h
e
p
ho
t
os
a
r
e
c
l
e
a
r
l
y
di
vi
de
d
i
nt
o
gro
ups
t
h
a
t
a
r
e
b
e
ni
g
n
a
nd
m
a
l
i
gn
a
n
t
,
a
s
i
n
F
i
gu
re
1.
T
h
e
r
e
m
a
i
n
i
n
g
1,
000
i
m
a
g
e
s
a
r
e
h
e
l
d
ba
c
k
fo
r
f
i
n
a
l
t
e
s
t
i
ng
,
w
hi
l
e
t
h
e
r
e
m
a
i
n
i
ng
9
,
6
00
a
r
e
d
e
s
i
g
na
t
e
d
for
t
r
a
i
n
i
n
g
[
6]
,
[
8]
.
T
h
e
s
e
c
ond
c
ol
l
e
c
t
i
on,
w
h
i
c
h
c
a
n
b
e
found
i
n
t
h
e
s
k
i
n
c
a
nc
e
r:
m
a
l
i
gn
a
nt
vs
.
b
e
n
i
gn
K
a
gg
l
e
pr
oj
e
c
t
,
c
o
nt
a
i
ns
3,
600
c
l
i
ni
c
a
l
i
m
a
g
e
s
of
s
i
z
e
2
24
×
244
pi
xe
l
s
t
ha
t
a
re
e
v
e
nl
y
d
i
vi
de
d
be
t
w
e
e
n
1,
800
b
e
ni
g
n
a
nd
1,
80
0
m
a
l
i
gn
a
nt
m
ol
e
s
[7]
.
T
he
c
o
m
b
i
na
t
i
on
of
t
h
e
t
w
o
d
a
t
a
s
e
t
s
pus
he
s
t
he
s
ys
t
e
m
t
o
a
ddr
e
s
s
i
s
s
ue
s
t
h
a
t
i
t
w
i
l
l
e
nc
o
unt
e
r
i
n
a
c
t
u
a
l
c
l
i
ni
c
s
b
y
di
s
p
l
a
y
i
ng
c
a
n
c
e
r
on
a
ra
ng
e
of
s
k
i
n
t
one
s
,
s
i
z
e
s
,
c
ol
ours
,
a
nd
t
e
x
t
ure
s
.
T
he
s
e
i
m
a
ge
c
o
l
l
e
c
t
i
ons
a
r
e
be
i
ng
us
e
d
be
c
a
us
e
do
c
t
o
rs
urg
e
nt
l
y
ne
e
d
a
w
a
y
t
o
s
pot
m
e
l
a
no
m
a
e
a
rl
y
a
nd
ge
t
pa
t
i
e
n
t
s
t
he
ri
g
ht
c
a
re
fa
s
t
[
1],
[6]
,
[7]
.
M
e
l
a
nom
a
i
s
da
ng
e
rous
-
i
t
s
pre
a
ds
qui
c
kl
y
a
nd
c
a
n
be
d
e
a
d
l
y
i
f
t
he
di
a
gnos
i
s
c
om
e
s
t
oo
l
a
t
e
.
W
he
n
c
l
i
ni
c
i
a
ns
s
e
a
r
c
h
for
e
a
rl
y
m
e
l
a
n
om
a
i
nd
i
c
a
t
i
ons
,
t
he
p
i
xe
l
s
i
n
t
he
f
i
rs
t
c
ol
l
e
c
t
i
o
n
—
w
h
i
c
h
w
a
s
c
row
ds
our
c
e
d
by
K
a
ggl
e
us
e
rs
—
s
how
m
i
nut
e
de
t
a
i
l
s
fr
om
de
rm
a
t
os
c
opi
c
phot
ogr
a
phs
t
a
ke
n
w
i
t
h
e
xp
e
ns
i
v
e
s
k
i
n
-
s
c
a
nn
i
ng
de
vi
c
e
s
.
E
v
e
ry
i
m
a
g
e
ha
s
a
t
ho
rough
l
a
b
e
l
t
h
a
t
i
s
e
i
t
he
r
hu
m
a
n
or
m
a
l
w
a
r
e
-
for
m
e
d
fol
l
ow
i
n
g
c
l
i
n
i
c
a
l
e
xa
m
i
n
a
t
i
ons
a
nd
m
i
c
ros
c
opy
s
l
i
d
e
s
[6]
,
[7]
,
[1
6]
.
T
h
i
s
a
l
l
ow
s
s
t
ude
n
t
s
t
o
re
l
y
o
n
s
uc
h
l
a
be
l
s
w
h
e
n
t
he
y
de
ve
l
op
n
e
w
s
up
e
rv
i
s
e
d
c
ode
.
A
s
t
he
pro
vi
d
e
r
obs
e
r
ve
s
,
t
he
s
e
c
ond
c
ol
l
e
c
t
i
on,
w
h
i
c
h
w
a
s
b
orrow
e
d
f
rom
t
he
IS
IC
a
r
c
hi
v
e
,
c
o
l
l
e
c
t
s
r
e
gu
l
a
r
c
l
i
ni
c
ph
ot
os
of
s
k
i
n
p
a
t
c
he
s
,
c
om
b
i
ni
ng
b
e
ni
g
n
a
nd
m
a
l
i
gna
nt
c
a
n
c
e
r
i
n
a
bou
t
e
qua
l
pr
opor
t
i
ons
[1]
,
[7]
.
T
h
e
t
e
a
m
s
r
e
m
ove
d
c
a
m
e
r
a
a
rt
e
f
a
c
t
s
a
nd
s
i
z
e
m
i
s
m
a
t
c
he
s
b
y
t
ri
m
m
i
ng
,
r
e
s
i
z
i
ng
,
a
nd
nor
m
a
l
i
s
i
n
g
t
he
p
i
x
e
l
s
t
o
m
a
k
e
e
a
c
h
b
a
t
c
h
e
a
s
i
e
r
t
o
i
nput
i
nt
o
n
e
ur
a
l
ne
t
s
or
vi
n
t
a
g
e
m
a
c
h
i
n
e
s
.
Com
b
i
ni
n
g
t
h
e
t
w
o
s
ou
rc
e
s
e
n
a
bl
e
s
re
s
e
a
r
c
he
rs
t
o
c
re
a
t
e
m
o
de
l
s
t
ha
t
l
e
a
rn
m
or
e
c
o
m
pr
e
h
e
ns
i
v
e
pa
t
t
e
rns
t
h
a
n
a
ny
one
c
l
i
n
i
c
c
a
n
i
d
e
nt
i
fy
a
n
d
t
he
n
.
D
a
t
a
s
pl
i
t
t
i
ng
a
nd
e
v
a
l
ua
t
i
on
s
t
ra
t
e
gy:
i
n
t
hi
s
i
m
pl
e
m
e
nt
a
t
i
on,
a
l
l
i
m
a
ge
s
i
n
t
he
ge
ne
r
a
t
e
d
d
a
t
a
s
e
t
w
e
re
pl
a
c
e
d
i
n
t
o
a
s
i
ngl
e
fol
d
e
r
s
t
ru
c
t
ur
e
a
nd
l
oa
d
e
d
t
hro
ugh
a
n
i
m
a
g
e
-
fol
de
r
i
n
t
e
rfa
c
e
.
D
u
ri
ng
t
ra
i
ni
ng
,
t
hos
e
i
m
a
ge
s
w
e
re
t
h
e
n
s
huff
l
e
d
ra
ndo
m
l
y
a
nd
p
re
s
e
nt
e
d
r
e
pe
a
t
e
dl
y
t
o
t
he
ne
t
w
ork
.
G
i
v
e
n
m
e
t
ri
c
s
re
fl
e
c
t
pe
rfor
m
a
nc
e
on
t
he
t
r
a
i
n
i
ng
di
s
t
r
i
bu
t
i
on
ra
t
he
r
t
h
a
n
a
s
t
r
i
c
t
l
y
i
nd
e
pe
n
de
n
t
v
a
l
i
da
t
i
on
s
e
t
,
b
e
c
a
us
e
t
h
e
s
a
m
e
s
e
t
of
i
m
a
ge
s
w
a
s
us
e
d
t
o
c
a
l
c
u
l
a
t
e
a
c
c
u
ra
c
y,
pr
e
c
i
s
i
on
,
re
c
a
l
l
,
a
nd
F
1
-
s
c
ore
a
t
t
h
e
e
nd
of
t
r
a
i
n
i
ng
.
A
l
t
h
ough
t
h
i
s
s
e
t
up
gi
v
e
s
a
n
i
n
i
t
i
a
l
i
de
a
of
how
w
e
l
l
t
h
e
f
i
ne
-
t
un
e
d
A
l
e
xN
e
t
c
a
n
fi
t
t
he
m
e
l
a
no
m
a
c
l
a
s
s
i
fi
c
a
t
i
o
n
t
a
s
k,
i
t
c
a
n
a
l
s
o
i
nf
l
a
t
e
g
e
ne
r
a
l
i
z
a
t
i
on
s
i
n
c
e
t
h
e
m
o
de
l
i
s
t
e
s
t
e
d
on
pho
t
o
s
t
h
a
t
i
t
ha
s
a
l
re
a
dy
s
e
e
n
duri
n
g
t
un
i
ng
.
T
he
n
e
xt
s
t
e
ps
i
n
t
hi
s
d
e
ve
l
op
m
e
n
t
w
i
l
l
fo
c
us
on
s
e
pa
r
a
t
i
ng
t
r
a
i
n
i
n
g,
va
l
i
d
a
t
i
on
a
nd
t
e
s
t
t
ra
i
ni
ng
s
e
t
s
t
ha
t
a
r
e
not
ove
rl
a
pp
i
ng
—
f
or
e
x
a
m
p
l
e
70%
,
t
r
a
i
n
i
ng
15%
,
va
l
i
d
a
t
i
on
1
5
%
t
e
s
t
i
ng
e
ffor
t
s
c
oul
d
b
e
e
m
pl
oy
e
d
t
o
a
voi
d
bi
a
s
a
nd
h
a
ve
a
n
unb
i
a
s
e
d
re
pre
s
e
nt
a
t
i
on
of
y
our
a
l
gor
i
t
h
m
'
s
t
rue
a
pp
l
i
c
a
b
i
l
i
t
y
w
i
t
hi
n
a
n
a
c
t
ua
l
prod
uc
t
i
on
e
nvi
r
onm
e
nt
.
A
dd
i
t
i
ona
l
l
y
,
i
m
pl
e
m
e
n
t
i
ng
c
ros
s
-
va
l
i
d
a
t
i
on
a
c
ros
s
m
ul
t
i
p
l
e
d
i
ff
e
re
n
t
d
a
t
a
s
e
t
s
e
t
s
c
a
n
p
rovi
de
you
w
i
t
h
a
n
un
bi
a
s
e
d
a
s
s
e
s
s
m
e
n
t
of
y
our
a
l
gor
i
t
h
m
'
s
p
e
rfor
m
a
n
c
e
a
nd
a
dd
t
o
t
h
e
re
l
i
a
bi
l
i
t
y
of
t
he
e
s
t
i
m
a
t
e
s
you
pr
ovi
d
e
.
6.
D
A
TA
P
R
E
-
P
R
O
C
ES
S
I
N
G
T
he
pr
e
-
pr
oc
e
s
s
i
ng
s
t
a
ge
o
f
t
h
e
da
t
a
w
a
s
e
s
s
e
nt
i
a
l
for
g
e
t
t
i
ng
t
he
m
e
l
a
no
m
a
s
ki
n
c
a
n
c
e
r
a
nd
m
a
l
i
gn
a
nt
ve
rs
us
b
e
n
i
gn
m
o
l
e
i
m
a
ge
da
t
a
s
e
t
s
r
e
a
d
y
for
e
ff
i
c
i
e
nt
t
ra
i
ni
ng
a
nd
t
he
be
s
t
pos
s
i
bl
e
p
e
rfor
m
a
nc
e
from
t
he
A
l
e
xN
e
t
m
ode
l
,
a
s
i
n
F
i
gure
2.
T
he
f
ol
l
ow
i
ng
a
re
i
m
port
a
n
t
pre
-
proc
e
s
s
i
ng
a
c
t
i
o
ns
a
nd
s
p
e
c
i
fi
c
s
o
f
hype
rp
a
ra
m
e
t
e
r
o
pt
i
m
i
s
a
t
i
on:
i)
Im
a
g
e
s
t
a
nd
a
rdi
s
a
t
i
on:
a
l
l
i
np
ut
i
m
a
g
e
s
w
e
r
e
r
e
s
i
z
e
d
u
ni
for
m
l
y
t
o
256
×
2
56
p
i
xe
l
s
t
o
m
a
i
n
t
a
i
n
c
ons
i
s
t
e
nt
di
m
e
ns
i
ons
s
ui
t
a
b
l
e
for
t
h
e
A
l
e
xN
e
t
a
r
c
hi
t
e
c
t
u
re
.
T
he
i
m
a
ge
s
w
e
r
e
no
rm
a
l
i
s
e
d
us
i
ng
c
h
a
nn
e
l
-
w
i
s
e
m
e
a
n
va
l
u
e
s
of
(0.
4
85,
0.
456
,
0
.
406)
a
nd
s
t
a
nda
r
d
de
vi
a
t
i
ons
o
f
(0.
2
29,
0.
224
,
0
.
225)
,
a
l
i
gni
n
g
w
i
t
h
t
he
pre
t
r
a
i
n
e
d
Im
a
g
e
N
e
t
m
od
e
l
’s
e
xp
e
c
t
e
d
i
nput
d
i
s
t
ri
b
ut
i
on
[30
]
.
i
i
)
D
a
t
a
a
ug
m
e
n
t
a
t
i
o
n:
t
o
i
m
p
rov
e
d
a
t
a
s
e
t
d
i
v
e
rs
i
t
y
a
nd
d
e
c
r
e
a
s
e
ov
e
rf
i
t
t
i
n
g
,
t
r
a
i
n
i
ng
m
e
t
h
ods
l
i
k
e
r
a
n
do
m
rot
a
t
i
on
,
ho
ri
z
o
n
t
a
l
a
nd
v
e
r
t
i
c
a
l
fl
i
p
pi
ng
,
s
c
a
l
i
ng
,
s
h
e
a
r
i
ng
,
a
n
d
t
r
a
ns
l
a
t
i
o
n
w
e
re
us
e
d
[2]
,
[
17]
,
[
23]
,
[
31]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2722
-
3221
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
,
V
o
l
.
7
,
N
o
.
1
,
M
a
rc
h
20
26
:
111
-
1
20
114
iii)
T
ra
i
ni
n
g
c
on
fi
gu
ra
t
i
on
:
t
h
e
fi
na
l
ful
l
y
-
c
onn
e
c
t
e
d
l
a
y
e
r
o
f
t
h
e
pre
t
r
a
i
ne
d
A
l
e
xN
e
t
m
od
e
l
w
a
s
a
d
j
us
t
e
d
t
o
provi
d
e
t
w
o
c
l
a
s
s
e
s
(
m
a
l
i
gna
n
t
a
n
d
be
n
i
gn)
.
U
s
i
ng
S
G
D
a
s
t
he
opt
i
m
i
s
e
r
,
t
h
e
m
o
m
e
nt
u
m
w
a
s
s
e
t
a
t
0.
9
,
a
nd
t
he
l
e
a
r
ni
ng
ra
t
e
w
a
s
s
e
t
a
t
0.
001
.
Cros
s
-
e
nt
r
opy
l
os
s
w
a
s
t
he
c
l
a
s
s
i
f
i
c
a
t
i
on
l
os
s
fun
c
t
i
on
,
w
h
i
c
h
w
a
s
s
ui
t
a
b
l
e
for
j
obs
i
nvo
l
vi
ng
b
i
na
r
y
c
l
a
s
s
e
s
.
i
v)
M
od
e
l
op
t
i
m
i
s
a
t
i
on
:
o
ut
pe
rfo
r
m
i
ng
fr
a
m
e
w
or
ks
f
or
d
e
e
p
l
e
a
r
ni
n
g
for
t
he
t
w
o
da
t
a
s
e
t
s
,
t
h
e
10
,
0
00
-
i
m
a
g
e
s
m
e
l
a
no
m
a
s
k
i
n
c
a
n
c
e
r
d
a
t
a
s
e
t
h
a
d
a
m
ode
l
a
c
c
ur
a
c
y
of
97
.
1
2%
a
ft
e
r
6453
s
e
c
onds
of
t
ra
i
ni
ng.
T
h
e
s
k
i
n
c
a
n
c
e
r:
m
a
l
i
gn
a
n
t
vs
.
b
e
n
i
gn
d
a
t
a
s
e
t
,
i
n
c
on
t
r
a
s
t
,
d
e
m
o
ns
t
r
a
t
e
d
t
he
e
f
f
i
c
a
c
y
a
nd
e
f
fi
c
i
e
n
c
y
of
t
h
e
A
l
e
xN
e
t
m
o
de
l
i
n
s
k
i
n
c
a
nc
e
r
c
l
a
s
s
i
f
i
c
a
t
i
on
w
i
t
h
a
n
a
c
c
ur
a
c
y
o
f
96
.
2
1
%
a
n
d
a
t
r
a
i
ni
ng
t
i
m
e
of
1
70
0
s
e
c
o
nds
.
F
i
gure
1
.
S
a
m
p
l
e
i
m
a
g
e
fro
m
t
he
c
om
b
i
ne
d
d
a
t
a
s
e
t
i
l
l
us
t
ra
t
i
n
g
be
ni
gn
a
nd
m
a
l
i
gna
n
t
s
k
i
n
c
a
nc
e
r
F
i
gure
2
.
O
v
e
rvi
e
w
of
t
he
m
e
t
hodo
l
ogy
:
d
a
t
a
proc
e
s
s
i
ng
,
a
ug
m
e
n
t
a
t
i
on
,
t
r
a
i
n
i
ng
7.
EX
P
ER
I
M
EN
TS
T
hi
s
r
e
s
e
a
rc
h
pre
s
e
nt
s
a
c
om
pre
h
e
ns
i
v
e
a
na
l
ys
i
s
o
f
how
v
a
ri
ous
s
k
i
n
c
a
n
c
e
r
s
pot
s
c
a
n
b
e
s
ort
e
d
us
i
ng
d
e
e
p
l
e
a
rni
ng
a
nd
c
on
ve
n
t
i
o
na
l
ML
t
e
c
hni
q
ue
s
de
s
c
ri
b
e
d
i
n
F
i
gu
r
e
3
[1]
–
[4],
[6]
,
[7]
,
[2
3]
.
T
w
o
i
m
a
ge
pool
s
w
e
re
e
m
pl
oye
d
i
n
t
he
t
e
s
t
r
uns
:
t
h
e
m
a
l
i
gn
a
n
t
-
ve
rs
us
-
be
ni
g
n
s
e
t
a
nd
t
h
e
1
0,
0
00
-
i
m
a
ge
m
e
l
a
no
m
a
s
ki
n
c
a
n
c
e
r
d
a
t
a
s
e
t
[6
],
[7]
.
T
o
i
de
n
t
i
fy
t
he
m
e
t
hod
t
h
a
t
b
e
s
t
c
l
a
s
s
i
fi
e
s
c
a
n
c
e
r
,
m
o
de
l
s
fro
m
bo
t
h
fa
m
i
l
i
e
s
w
e
re
s
c
ore
d
s
i
d
e
by
s
i
de
[25]
,
[32]
.
7.
1
.
D
e
e
p
l
e
a
r
n
i
n
g
n
e
tw
o
r
k
s
S
t
a
nd
a
rd
CN
N
s
,
Re
s
N
e
t
,
L
e
N
e
t
,
G
oogL
e
N
e
t
,
V
G
G
16
,
In
c
e
p
t
i
onV
3
,
G
A
N
s
,
a
nd
A
l
e
xN
e
t
w
e
r
e
a
m
on
g
t
he
pow
e
rfu
l
de
e
p
l
e
a
rn
i
ng
m
o
de
l
s
t
ha
t
t
he
r
e
s
e
a
r
c
he
rs
i
n
ve
s
t
i
ga
t
e
d
[
1]
–
[4]
,
[11]
,
[23]
.
T
h
e
s
k
i
n
c
a
n
c
e
r
da
t
a
s
e
t
s
w
e
r
e
us
e
d
t
o
re
f
i
ne
e
a
c
h
m
od
e
l
,
w
hi
c
h
be
ga
n
w
i
t
h
pr
e
t
r
a
i
n
e
d
w
e
i
gh
t
s
fro
m
A
l
e
xN
e
t
.
A
c
c
ur
a
c
y
a
nd
t
r
a
i
n
i
ng
p
a
c
e
va
r
i
e
d
s
i
g
ni
f
i
c
a
nt
l
y
a
c
ros
s
s
y
s
t
e
m
s
,
de
s
p
i
t
e
t
he
f
a
c
t
t
h
a
t
ne
a
rl
y
a
l
l
of
t
h
e
m
ge
ne
r
a
t
e
d
pos
i
t
i
ve
r
e
s
ul
t
s
.
F
or
e
xa
m
p
l
e
,
L
e
N
e
t
a
nd
V
G
G
16
w
e
re
f
a
s
t
e
r
but
t
ra
i
l
e
d
b
e
hi
n
d,
w
h
e
re
a
s
Re
s
N
e
t
Evaluation Warning : The document was created with Spire.PDF for Python.
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
IS
S
N
:
2722
-
3221
A
d
v
an
c
e
s
i
n
de
r
m
a
t
ol
o
gi
c
al
i
m
ag
i
ng:
e
nhan
c
i
ng
s
k
i
n
m
e
l
a
n
om
a
c
l
as
s
i
f
i
c
a
t
i
o
n
f
or
…
(D
e
b
adut
t
a
Sahoo
)
115
a
nd
G
oogL
e
N
e
t
a
c
hi
e
ve
d
r
e
s
pe
c
t
a
bl
e
a
c
c
ur
a
c
i
e
s
.
W
h
e
n
c
o
m
bi
n
e
d
,
t
he
f
i
ndi
ngs
u
ne
qu
i
vo
c
a
l
l
y
d
e
m
ons
t
ra
t
e
t
ha
t
ove
r
a
l
l
de
s
i
gn,
d
e
pt
h,
a
nd
c
o
m
p
l
e
x
i
t
y
i
nfl
u
e
n
c
e
—
a
nd
oc
c
a
s
i
on
a
l
l
y
i
m
pe
d
e
—
s
k
i
n
c
a
nc
e
r
c
a
t
e
gor
i
s
a
t
i
on
pe
rfor
m
a
nc
e
[3
3],
[34]
.
7.
2
.
M
ac
h
i
n
e
l
e
a
r
n
i
n
g
n
e
tw
o
r
k
s
S
V
M
s
,
K
N
N
s
,
de
c
i
s
i
on
t
re
e
s
,
ra
ndom
fore
s
t
s
,
na
i
ve
B
a
ye
s
,
a
nd
fl
e
xi
bl
e
l
oc
a
l
a
pproxi
m
a
t
e
ne
a
re
s
t
ne
i
ghbou
r
a
re
s
om
e
of
t
he
w
e
l
l
-
kno
w
n
c
l
a
s
s
i
fi
e
r
s
t
ha
t
t
he
t
e
a
m
a
l
s
o
e
xa
m
i
ne
d
a
nd
vi
s
ua
l
i
s
e
d
[1],
[
2],
[4
],
[5],
[11],
[13],
[14
]
,
[20],
[21],
[28
],
[31]
.
Be
fore
m
a
ki
ng
a
fi
na
l
a
pproxi
m
a
t
i
on,
t
he
s
e
m
ode
l
s
re
l
i
e
d
on
m
a
nua
l
l
y
c
re
a
t
e
d
fe
a
t
ure
s
t
a
ke
n
from
a
t
y
pi
c
a
l
i
m
a
ge
-
proc
e
s
s
i
n
g
pi
pe
l
i
ne
.
D
e
e
pe
r
ne
ura
l
ne
t
w
ork
s
w
on
t
h
e
gol
d
m
e
da
l
de
s
pi
t
e
S
V
M
a
n
d
K
N
N
s
l
i
p
pi
ng
i
nt
o
t
he
m
i
d
-
80%
a
c
c
ura
c
y
ra
n
ge
be
c
a
us
e
of
t
he
i
r
s
upe
ri
o
r
a
bi
l
i
t
y
t
o
ha
ndl
e
t
he
noi
s
e
a
nd
fi
ne
de
t
a
i
l
s
pre
s
e
nt
i
n
s
k
i
n
c
a
nc
e
r
i
m
a
ge
s
[1]
.
N
e
ve
rt
he
l
e
s
s
,
t
he
ol
de
r
a
l
gori
t
hm
s
e
s
t
a
bl
i
s
he
d
re
l
i
a
bl
e
ba
s
e
l
i
ne
s
a
n
d
s
e
r
ve
d
a
s
a
re
m
i
nde
r
of
t
he
a
dva
nt
a
ge
s
a
nd
di
s
a
dva
nt
a
ge
s
o
f
t
he
t
ra
di
t
i
ona
l
t
ool
ki
t
.
F
i
gure
3
.
V
i
s
u
a
l
r
e
pre
s
e
nt
a
t
i
on
of
i
m
a
g
e
p
re
pr
oc
e
s
s
i
ng
s
t
e
ps
:
fro
m
r
a
w
i
npu
t
t
o
nor
m
a
l
i
z
a
t
i
on
7.
3
.
P
r
op
os
e
d
mod
e
l
A
l
e
xN
e
t
c
om
pri
s
e
s
e
i
ght
l
a
ye
r
s
i
n
t
ot
a
l
,
w
i
t
h
t
hre
e
ful
l
y
c
onne
c
t
e
d
l
a
ye
rs
c
om
i
ng
a
ft
e
r
fi
ve
c
onvol
ut
i
o
na
l
l
a
ye
rs
,
a
s
s
ho
w
n
i
n
F
i
gu
re
4
[1]
.
T
oge
t
he
r,
t
he
y
e
na
bl
e
t
he
m
ode
l
t
o
pi
c
k
up
di
ffe
re
nt
de
t
a
i
l
s
i
n
t
he
de
pt
h
of
unp
roc
e
s
s
e
d
ph
ot
os
.
T
he
l
a
ye
r
i
s
fol
l
o
w
e
d
by
a
3
×
3
m
a
x
pool
i
ng
l
a
ye
r
w
i
t
h
a
s
t
ri
de
of
2
.
A
s
i
t
e
nha
nc
e
s
ge
ne
ra
l
i
s
a
t
i
on
a
nd
de
c
re
a
s
e
s
t
he
s
pa
t
i
a
l
di
m
e
ns
i
on
s
of
t
he
fe
a
t
ure
m
a
ps
,
i
t
c
a
pt
ure
s
t
he
m
os
t
prom
i
ne
nt
fe
a
t
ure
s
.
T
he
m
ode
l
l
e
a
r
ns
t
he
m
o
s
t
re
l
e
va
nt
i
nform
a
t
i
on
or
fe
a
t
ure
s
of
t
he
pi
c
t
u
re
,
w
hi
c
h
re
duc
e
s
t
he
c
om
p
ut
i
ng
l
oa
d
i
n
a
ddi
t
i
o
n
t
o
l
e
a
rni
n
g
t
o
ge
ne
r
a
l
i
s
e
fa
c
t
ua
l
l
y
[1
],
[2],
[
6],
[7
]
.
F
or
i
ns
t
a
nc
e
,
t
h
e
s
e
c
ond
l
a
ye
r
c
ont
a
i
n
s
256
5
×
5
fi
l
t
e
rs
,
w
hi
l
e
t
he
t
hi
rd,
fou
rt
h,
a
nd
fi
ft
h
l
a
ye
rs
ha
ve
384
a
nd
256
3
×
3
fi
l
t
e
rs
,
re
s
pe
c
t
i
ve
l
y.
T
o
ge
t
he
r
w
i
t
h
t
he
fi
r
s
t
l
a
ye
r,
a
l
l
o
f
t
he
s
e
l
a
ye
rs
c
a
re
f
ul
l
y
de
s
c
ri
be
t
he
rou
gh
t
o
e
xa
c
t
pa
t
t
e
rn
s
or
s
ha
pe
s
,
w
hi
c
h
i
s
c
r
uc
i
a
l
for
de
t
e
rm
i
ni
ng
i
f
a
s
ki
n
di
s
e
a
s
e
i
s
be
ni
gn
o
r
c
a
nc
e
rou
s
[6],
[
7]
.
Re
c
t
i
fi
e
d
l
i
ne
a
r
uni
t
(Re
L
U
)
a
c
t
i
va
t
i
on
fu
nc
t
i
on
s
a
re
a
l
s
o
i
nc
orpo
ra
t
e
d
i
nt
o
a
l
l
of
t
he
s
e
l
e
ve
l
s
;
t
he
s
e
a
c
t
i
va
t
i
on
fu
nc
t
i
on
s
a
re
c
ons
i
s
t
e
nt
t
hro
ugho
ut
t
he
c
onfl
i
c
t
i
ng
l
a
ye
rs
.
S
i
nc
e
t
he
A
l
e
xN
e
t
m
ode
l
a
c
hi
e
ve
s
a
good
ba
l
a
nc
e
of
a
c
c
ura
c
y,
pre
c
i
s
i
on,
re
c
a
l
l
,
F
1
s
c
ore
,
a
nd
m
a
na
ge
a
bl
e
t
ra
i
ni
n
g
t
i
m
e
,
w
e
de
c
i
de
d
t
o
e
m
pl
o
y
i
t
.
T
he
ne
t
w
or
k
produc
e
d
re
m
a
rka
bl
e
re
s
ul
t
s
a
ft
e
r
a
dj
us
t
i
ng
a
fe
w
hype
r
pa
ra
m
e
t
e
rs
for
e
a
c
h
da
t
a
s
e
t
:
a
0.
001
l
e
a
rni
ng
ra
t
e
w
i
t
h
0.
9
m
om
e
nt
um
u
s
i
ng
S
G
D
,
a
25
6
×
256
-
pi
xe
l
i
nput
s
i
z
e
,
a
nd
c
ros
s
e
nt
rop
y
l
os
s
.
O
n
t
he
m
e
l
a
nom
a
s
e
t
,
t
he
a
c
c
u
ra
c
y
w
a
s
9
7.
12%
i
n
64
53
s
e
c
ond
s
,
a
nd
on
t
he
m
a
l
i
gna
nt
-
ve
rs
us
-
be
ni
gn
s
e
t
,
i
t
w
a
s
96.
21
%
i
n
1700.
8
5
s
e
c
on
ds
.
T
he
s
e
i
m
pre
s
s
i
ve
re
s
ul
t
s
,
c
om
bi
ne
d
w
i
t
h
re
a
s
ona
bl
e
t
ra
i
ni
ng
t
i
m
e
s
,
i
m
pl
y
t
ha
t
A
l
e
x
N
e
t
m
a
y
be
a
de
pe
nda
bl
e
t
ool
for
ro
ut
i
ne
c
l
i
ni
c
a
l
di
a
gno
s
e
s
.
T
he
m
ode
l
'
s
e
ffe
c
t
i
ve
ne
s
s
i
s
f
urt
he
r
de
m
on
s
t
ra
t
e
d
by
t
he
t
ra
i
ni
ng
t
i
m
e
s
of
64
53
s
e
c
ond
s
f
or
t
he
m
e
l
a
nom
a
da
t
a
s
e
t
a
n
d
1700.
85
s
e
c
ond
s
fo
r
t
he
m
a
l
i
gna
nt
vs
.
be
ni
gn
da
t
a
s
e
t
,
w
hi
c
h
m
a
ke
i
t
a
vi
a
bl
e
opt
i
on
fo
r
pra
c
t
i
c
a
l
de
rm
a
t
ol
ogy
a
ppl
i
c
a
t
i
o
ns
.
T
he
fi
ndi
ng
s
of
t
hi
s
s
t
udy
n
ot
o
nl
y
hi
ghl
i
g
ht
t
he
A
l
e
x
N
e
t
m
ode
l
'
s
p
ot
e
nt
i
a
l
t
o
i
m
prove
di
a
gn
os
t
i
c
p
re
c
i
s
i
o
n,
but
t
he
y
a
l
s
o
a
dd
t
o
t
he
e
x
pa
ndi
ng
c
o
rpu
s
of
re
s
e
a
rc
h
o
n
t
he
u
s
e
of
A
I
i
n
m
e
di
c
a
l
i
m
a
ge
proc
e
s
s
i
n
g.
8.
R
ES
U
LTS
A
N
D
D
I
S
C
U
S
S
I
O
N
T
he
ne
w
A
l
e
xN
e
t
m
od
e
l
c
ons
i
s
t
e
n
t
l
y
a
nd
a
c
c
ur
a
t
e
l
y
de
t
e
c
t
s
m
e
l
a
nom
a
s
po
t
s
a
s
e
i
t
he
r
be
n
i
gn
or
m
a
l
i
gn
a
nt
,
a
c
c
ordi
n
g
t
o
t
e
s
t
s
c
on
duc
t
e
d
o
n
t
w
o
di
s
t
i
n
c
t
c
o
l
l
e
c
t
i
ons
of
s
ki
n
c
a
n
c
e
r
i
m
a
g
e
s
.
O
n
t
h
e
b
i
g
g
e
r
m
e
l
a
n
om
a
s
k
i
n
c
a
n
c
e
r
d
a
t
a
s
e
t
,
i
t
a
c
hi
e
ve
d
97
.
1
2%
a
c
c
ur
a
c
y
a
nd
96
%
A
U
C
a
f
t
e
r
a
ppr
ox
i
m
a
t
e
l
y
64
53
s
e
c
onds
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2722
-
3221
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
,
V
o
l
.
7
,
N
o
.
1
,
M
a
rc
h
20
26
:
111
-
1
20
116
O
n
t
h
e
b
a
l
a
nc
e
d
M
a
l
i
gn
a
nt
vs
B
e
ni
g
n
s
e
t
,
i
t
r
e
c
or
de
d
96
.
21
%
a
c
c
u
ra
c
y
a
nd
96%
A
U
C
i
n
a
pprox
i
m
a
t
e
l
y
1700
s
e
c
onds
,
v
i
s
ua
l
i
s
e
d
i
n
F
i
gur
e
5
.
T
he
m
od
e
l
d
e
m
ons
t
ra
t
e
d
i
t
s
s
t
r
e
ngt
h
w
he
n
pr
e
s
e
n
t
e
d
w
i
t
h
a
va
ri
e
t
y
o
f
c
l
i
ni
c
a
l
i
m
a
g
e
s
by
prod
uc
i
ng
ba
l
a
nc
e
d
pr
e
c
i
s
i
on
,
r
e
c
a
l
l
,
a
nd
F
1
s
c
or
e
s
a
c
ros
s
bo
t
h
ru
ns
,
a
s
di
s
pl
a
ye
d
i
n
T
a
bl
e
1
.
In
t
e
r
m
s
of
ra
w
a
c
c
ur
a
c
y
,
a
re
a
-
und
e
r
-
c
urv
e
v
a
l
u
e
s
,
a
n
d
t
ra
i
ni
n
g
s
p
e
e
d
,
A
l
e
xN
e
t
out
pe
rform
e
d
s
e
ve
r
a
l
m
ore
re
c
e
nt
de
e
p
l
e
a
rn
i
ng
a
nd
c
onv
e
nt
i
ona
l
M
L
a
ppro
a
c
he
s
i
n
he
a
d
-
to
-
he
a
d
c
om
p
a
ri
s
o
ns
,
de
m
ons
t
ra
t
i
ng
t
he
i
m
p
ort
a
nc
e
of
t
he
s
e
l
e
c
t
e
d
a
rc
h
i
t
e
c
t
ure
a
nd
a
d
j
us
t
e
d
h
ype
r
pa
r
a
m
e
t
e
rs
.
W
he
n
c
om
b
i
n
e
d,
t
he
f
i
ndi
ngs
i
m
p
l
y
t
h
a
t
t
h
e
s
ys
t
e
m
m
a
y
b
e
a
re
l
i
a
b
l
e
t
oo
l
for
e
a
rl
y
,
pre
c
i
s
e
m
e
l
a
n
om
a
i
de
nt
i
fi
c
a
t
i
on
,
w
hi
c
h
w
oul
d
e
nh
a
nc
e
phys
i
c
i
a
n
j
udg
m
e
nt
a
n
d,
e
ve
n
t
ua
l
l
y
,
pa
t
i
e
nt
out
c
om
e
s
[6
],
[35]
.
F
i
g
ure
6
i
l
l
us
t
ra
t
e
s
t
he
a
ve
r
a
ge
pe
r
for
m
a
n
c
e
a
n
a
l
ys
i
s
of
m
a
c
h
i
ne
l
e
a
rni
n
g
m
od
e
l
s
e
v
a
l
u
a
t
e
d
us
i
ng
t
w
o
di
s
t
i
n
c
t
s
ki
n
c
a
nc
e
r
da
t
a
s
e
t
s
.
F
i
gure
4
.
O
v
e
rvi
e
w
of
t
he
pr
opos
e
d
m
ode
l
a
r
c
hi
t
e
c
t
ur
e
a
nd
e
va
l
u
a
t
i
on
pro
c
e
s
s
Evaluation Warning : The document was created with Spire.PDF for Python.
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
IS
S
N
:
2722
-
3221
A
d
v
an
c
e
s
i
n
de
r
m
a
t
ol
o
gi
c
al
i
m
ag
i
ng:
e
nhan
c
i
ng
s
k
i
n
m
e
l
a
n
om
a
c
l
as
s
i
f
i
c
a
t
i
o
n
f
or
…
(D
e
b
adut
t
a
Sahoo
)
117
F
i
gure
5
.
A
v
e
ra
ge
t
i
m
e
c
om
pl
e
xi
t
y
c
o
m
pa
r
i
s
on
of
t
h
e
d
e
e
p
l
e
a
rn
i
ng
m
o
de
l
s
i
n
s
e
c
onds
T
a
b
l
e
1
.
A
v
e
ra
g
e
pe
r
form
a
n
c
e
c
o
m
p
a
ri
s
on
of
t
h
e
t
w
o
da
t
a
s
e
t
s
a
c
ros
s
e
va
l
ua
t
i
on
m
e
t
ri
c
s
D
L
m
o
d
e
l
A
c
c
u
ra
c
y
(%
)
P
re
c
i
s
i
o
n
(%
)
Re
c
a
l
l
(%
)
F
1
s
c
o
re
(%
)
A
l
e
x
N
e
t
9
7
.
1
2
96
95
97
CN
N
91
95
85
90
L
e
N
e
t
8
8
.
4
5
4
6
.
7
9
47
4
6
.
9
5
G
o
o
g
L
e
N
e
t
87
47
51
49
S
N
N
87
86
87
87
F
L
A
N
N
86
89
80
84
Re
s
N
e
t
6
6
.
8
6
47
16
24
GAN
53
52
67
58
V
G
G
1
6
51
50
44
47
F
i
gure
6
.
A
v
e
ra
ge
pe
r
form
a
n
c
e
a
n
a
l
ys
i
s
of
m
a
c
hi
n
e
l
e
a
rn
i
ng
m
ode
l
s
us
i
ng
t
w
o
di
s
t
i
n
c
t
s
ki
n
c
a
nc
e
r
da
t
a
s
e
t
s
9.
LI
M
I
TA
TI
O
N
S
A
N
D
F
U
TU
R
E
D
I
R
EC
TI
O
N
S
N
ot
e
w
o
rt
hy
l
i
m
i
t
a
t
i
ons
m
us
t
be
t
a
ke
n
i
nt
o
c
ons
i
de
ra
t
i
o
n
re
ga
rdi
ng
t
he
s
a
i
d
A
l
e
xN
e
t
-
ba
s
e
d
m
e
t
hodol
ogy,
de
s
pi
t
e
t
he
a
c
c
ura
c
y,
A
U
C,
a
nd
ba
l
a
nc
e
d
pre
c
i
s
i
on
–
re
c
a
l
l
a
c
hi
e
ve
d
on
t
w
o
publ
i
c
l
y
a
c
c
e
s
s
i
bl
e
s
ki
n
c
a
nc
e
r
da
t
a
s
e
t
s
be
i
ng
re
l
a
t
i
ve
l
y
g
ood.
T
o
be
gi
n
w
i
t
h,
t
h
e
m
e
l
a
nom
a
s
ki
n
c
a
nc
e
r
da
t
a
s
e
t
a
nd
s
ki
n
c
a
nc
e
r
m
a
l
i
gna
nt
vs
.
be
ni
gn
da
t
a
s
e
t
a
re
c
ul
l
e
d
from
pa
rt
i
c
ul
a
r
onl
i
ne
re
po
s
i
t
ori
e
s
a
nd,
t
he
re
fore
,
m
a
y
not
a
c
c
ura
t
e
l
y
re
pre
s
e
nt
t
he
va
ri
a
t
i
on
o
f
t
he
re
a
l
-
w
o
rl
d
c
l
i
n
i
c
a
l
popul
a
t
i
on,
i
m
a
gi
ng
de
vi
c
e
s
,
a
nd
proc
e
s
s
e
s
of
a
c
qui
s
i
t
i
on.
U
nde
r
-
re
pre
s
e
nt
a
t
i
o
n
of
c
e
rt
a
i
n
s
ki
n
t
one
s
,
t
he
e
a
rl
y
-
s
t
a
ge
m
e
l
a
nom
a
s
,
or
l
e
s
i
on
t
ype
,
m
a
y
l
e
a
d
t
o
da
t
a
s
e
t
bi
a
s
a
nd
l
i
m
i
t
t
he
a
ppl
i
c
a
bi
l
i
t
y
of
t
he
re
p
ort
e
d
pe
rfo
rm
a
nc
e
t
o
ot
he
r
hos
pi
t
a
l
s
or
re
gi
on
s
[
1]
.
In
a
ddi
t
i
on,
s
uc
h
s
t
udi
e
s
oft
e
n
u
s
e
t
he
l
a
be
l
s
c
re
a
t
e
d
by
da
t
a
s
e
t
de
ve
l
ope
r
s
.
W
hi
l
e
t
he
s
e
l
a
be
l
s
m
a
y
be
of
poor
q
ua
l
i
t
y,
t
he
re
i
s
no
w
a
y
t
o
c
o
nt
rol
t
he
di
ffe
re
nc
e
s
t
ha
t
m
a
y
e
xi
s
t
w
i
t
hi
n
t
he
da
t
a
s
e
t
i
t
s
e
l
f
duri
ng
t
h
e
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2722
-
3221
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
,
V
o
l
.
7
,
N
o
.
1
,
M
a
rc
h
20
26
:
111
-
1
20
118
c
ours
e
of
t
he
s
t
ud
y.
E
ve
n
w
i
t
h
t
he
popul
a
ri
t
y
of
t
he
s
e
da
t
a
s
e
t
s
,
w
e
s
u
s
pe
c
t
t
ha
t
t
he
re
i
s
no
t
rue
ground
fo
r
t
he
m
e
t
ri
c
s
u
s
e
d,
a
nd
t
h
us
no
t
r
ue
ba
s
i
s
f
or
t
he
m
e
t
ri
c
s
,
due
t
o
t
he
e
xi
s
t
e
nc
e
of
w
ha
t
w
e
ha
ve
c
a
l
l
e
d
‘l
a
be
l
l
i
ng
n
oi
s
e
’,
di
ffe
re
nc
e
s
re
l
a
t
e
d
t
o
t
he
hi
s
t
opa
t
hol
o
gy
a
ppr
ova
l
a
t
ha
nd,
a
nd
t
he
c
ont
rol
o
f
t
he
di
a
gno
s
i
s
.
A
l
s
o,
t
he
s
t
udy
l
a
c
k
s
t
e
s
t
s
on
a
m
o
re
s
i
g
ni
fi
c
a
nt
dom
a
i
n
s
hi
ft
,
s
uc
h
a
s
i
m
a
ge
s
t
a
ke
n
from
a
va
ri
e
t
y
of
de
rm
a
t
os
c
ope
s
or
s
m
a
rt
ph
one
c
a
m
e
ra
s
,
a
s
t
he
foc
us
of
t
he
a
s
s
e
s
s
m
e
nt
i
s
on
t
w
o
s
i
m
i
l
a
r
da
t
a
s
e
t
s
.
T
hi
s
m
a
y
i
ndi
c
a
t
e
a
ga
p
i
n
t
he
s
t
u
dy
c
onc
e
rni
ng
s
uc
h
a
dom
a
i
n
s
hi
ft
.
L
a
s
t
l
y,
none
of
t
he
t
ri
a
l
s
ha
ve
be
e
n
d
one
i
n
t
he
s
e
t
t
i
n
g
of
a
fut
ure
s
t
udy
o
n
re
a
de
r
s
(de
rm
a
t
ol
o
gi
s
t
s
or
ot
he
r
he
a
l
t
hc
a
re
provi
de
r
s
).
T
he
re
fo
re
,
m
a
ny
of
t
he
i
m
p
ort
a
nt
i
s
s
ue
s
,
i
nc
l
udi
n
g
t
rus
t
i
n
t
he
m
ode
l
by
t
he
us
e
r
s
,
ho
w
i
nt
e
rp
re
t
a
bl
e
a
nd
how
w
e
l
l
w
ork
fl
ow
s
c
a
n
be
i
nc
orpora
t
e
d
i
nt
o
da
i
l
y
pra
c
t
i
c
e
,
a
nd
how
m
uc
h
A
I
a
s
s
i
s
t
a
nc
e
a
ffe
c
t
s
t
he
s
pe
e
d
of
d
i
a
gno
s
e
s
a
nd
ho
w
qui
c
kl
y
de
c
i
s
i
o
ns
c
a
n
be
m
a
de
ba
s
e
d
on
di
a
gno
s
i
s
,
w
e
re
not
t
e
s
t
e
d.
T
he
c
urre
nt
m
ode
l
onl
y
doe
s
a
n
i
m
a
ge
-
l
e
ve
l
c
l
a
s
s
i
fi
c
a
t
i
on
a
nd
doe
s
not
ye
t
do
l
e
s
i
o
n
s
e
gm
e
nt
a
t
i
on
o
r
pro
vi
de
s
t
ruc
t
ure
d
e
xpl
a
na
t
i
on
s
t
ha
t
w
o
ul
d
ha
ve
be
e
n
u
s
e
ful
i
n
provi
ng
t
he
c
l
i
ni
c
a
l
va
l
i
di
t
y
of
t
he
m
ode
l
s
.
T
he
l
i
m
i
t
a
t
i
ons
of
our
fi
ndi
n
gs
pro
vi
d
e
m
a
n
y
a
ve
nue
s
for
fut
ur
e
r
e
s
e
a
rc
h
.
F
u
t
ur
e
r
e
s
e
a
rc
h
s
hou
l
d
c
orrobo
ra
t
e
our
m
od
e
l
'
s
re
s
u
l
t
s
by
v
a
l
i
da
t
i
n
g
i
t
a
g
a
i
ns
t
i
nde
p
e
nd
e
nt
,
m
ul
t
i
-
c
e
n
t
re
da
t
a
s
e
t
s
w
i
t
h
m
or
e
e
xt
e
ns
i
v
e
va
r
i
a
t
i
o
ns
i
n
s
ki
n
t
ype
s
,
l
e
s
i
on
t
ype
s
,
a
n
d
i
m
a
g
e
a
c
qui
s
i
t
i
on
m
e
t
ho
ds
,
a
s
w
e
l
l
a
s
by
ut
i
l
i
z
i
ng
c
r
os
s
-
da
t
a
s
e
t
e
v
a
l
u
a
t
i
on
m
e
t
hods
t
o
qua
n
t
i
f
y
t
h
e
i
m
p
a
c
t
of
do
m
a
i
n
s
hi
ft
.
A
dd
i
t
i
on
a
l
l
y,
unc
e
rt
a
i
nt
y
e
s
t
i
m
a
t
i
on
a
nd
c
a
l
i
br
a
t
i
on
s
houl
d
b
e
i
n
c
orpor
a
t
e
d
i
n
t
o
fut
ure
s
t
u
di
e
s
t
o
a
l
l
o
w
be
t
t
e
r
qua
nt
i
fi
c
a
t
i
on
of
t
he
c
e
rt
a
i
n
t
y
a
s
s
o
c
i
a
t
e
d
w
i
t
h
i
nd
i
vi
d
ua
l
pr
e
di
c
t
i
ons
,
p
a
rt
i
c
u
l
a
r
l
y
for
bord
e
rl
i
ne
or
i
nfre
que
n
t
l
y
obs
e
rve
d
i
ns
t
a
nc
e
s
.
L
a
s
t
l
y
,
fut
ur
e
s
t
udi
e
s
s
h
oul
d
provi
de
f
or
t
h
e
i
nc
or
pora
t
i
o
n
of
e
xpl
a
n
a
t
i
ons
of
t
h
e
A
I'
s
de
c
i
s
i
ons
t
hrou
gh
t
he
us
e
of
t
e
c
hni
q
ue
s
l
i
k
e
s
a
l
i
e
n
c
y
m
a
ps
a
nd
c
l
a
s
s
-
a
c
t
i
va
t
i
on
v
i
s
ua
l
i
z
a
t
i
ons
,
a
nd
t
h
e
c
ondu
c
t
of
pros
p
e
c
t
i
v
e
s
t
udi
e
s
e
m
p
l
oyi
ng
h
um
a
n
-
A
I
c
o
l
l
a
bor
a
t
i
on
i
n
a
c
t
u
a
l
c
l
i
ni
c
a
l
s
e
t
t
i
ngs
[4]
,
[3
6]
b
e
for
e
A
I
-
ba
s
e
d
d
e
c
i
s
i
on
-
s
u
ppor
t
s
ys
t
e
m
s
c
a
n
be
i
m
pl
e
m
e
n
t
e
d
a
n
d
i
nt
e
gr
a
t
e
d
i
nt
o
c
l
i
ni
c
a
l
pr
a
c
t
i
c
e
.
10.
C
O
N
C
LU
S
I
O
N
T
hi
s
s
t
udy
c
on
fi
r
m
s
t
ha
t
f
i
ne
-
t
un
e
d
A
l
e
xN
e
t
e
x
c
e
l
s
a
t
c
l
a
s
s
i
fyi
ng
s
ki
n
c
a
nc
e
r
i
m
a
g
e
s
,
a
c
hi
e
vi
n
g
97.
12
%
a
c
c
ur
a
c
y
on
t
he
m
e
l
a
n
om
a
s
ki
n
c
a
n
c
e
r
da
t
a
s
e
t
a
nd
96.
21
%
on
t
he
s
ki
n
c
a
nc
e
r
m
a
l
i
gn
a
nt
vs
.
b
e
ni
g
n
da
t
a
s
e
t
,
w
hi
l
e
ou
t
pe
rf
or
m
i
ng
t
r
a
di
t
i
o
na
l
ML
m
e
t
hods
a
nd
s
e
v
e
ra
l
c
o
m
pe
t
i
n
g
d
e
e
p
n
e
t
w
o
rks
.
T
he
m
ode
l
s
houl
d
be
i
n
c
orpor
a
t
e
d
i
n
t
o
t
he
d
e
rm
a
t
o
l
ogy
t
r
i
a
g
e
proc
e
s
s
w
i
t
h
s
a
l
i
e
nc
y
m
a
ps
for
c
l
i
ni
c
a
l
us
e
s
o
t
h
a
t
t
h
e
m
ode
l
i
s
t
r
a
ns
p
a
re
n
t
,
v
a
l
i
da
t
e
d
t
hrou
gh
m
u
l
t
i
c
e
nt
r
e
t
r
i
a
l
s
a
c
ros
s
va
r
i
ous
popu
l
a
t
i
ons
,
a
n
d
us
e
d
a
s
a
c
ol
l
a
bo
ra
t
i
v
e
t
oo
l
t
o
fa
c
i
l
i
t
a
t
e
e
a
rl
y
de
t
e
c
t
i
on
of
m
e
l
a
no
m
a
a
nd
i
m
pro
ve
p
a
t
i
e
n
t
c
a
re
.
In
de
r
m
a
t
o
l
ogi
c
a
l
pra
c
t
i
c
e
,
w
h
e
re
our
m
o
de
l
w
i
l
l
s
e
rv
e
a
s
a
t
r
i
a
g
i
ng
t
oo
l
,
w
e
propos
e
t
h
a
t
t
h
e
c
l
i
n
i
c
a
l
i
m
pl
e
m
e
n
t
a
t
i
on
of
our
m
ode
l
w
i
l
l
o
c
c
ur
i
n
t
h
e
f
ut
ur
e
,
w
i
t
h
a
n
e
m
pha
s
i
s
on
us
i
n
g
s
a
l
i
e
n
c
y
m
a
ppi
ng
t
e
c
hni
que
s
t
o
hi
gh
l
i
g
ht
t
he
re
l
e
va
n
t
vi
s
u
a
l
fe
a
t
ur
e
s
t
ha
t
a
i
d
i
n
t
h
e
d
i
a
g
nos
t
i
c
proc
e
s
s
.
T
o
va
l
i
d
a
t
e
t
he
us
e
of
our
m
ode
l
i
n
t
h
e
c
l
i
ni
c
a
l
s
e
t
t
i
ng,
w
e
pl
a
n
t
o
c
o
nduc
t
m
u
l
t
i
-
s
i
t
e
s
t
ud
i
e
s
t
h
a
t
i
n
c
l
ude
a
l
a
rge
n
um
b
e
r
of
pa
t
i
e
nt
s
a
c
c
e
s
s
e
d
by
m
ul
t
i
p
l
e
phot
ogr
a
ph
i
c
c
a
pt
ur
e
d
e
vi
c
e
s
.
O
ur
m
o
de
l
a
c
t
s
a
s
a
s
upp
l
e
m
e
n
t
a
r
y
t
o
ol
t
o
a
i
d
de
r
m
a
t
o
l
ogi
s
t
s
;
t
hus
,
bot
h
c
om
m
un
i
t
y
a
nd
pra
c
t
i
c
e
e
xp
e
ri
e
n
c
e
s
w
i
l
l
s
u
pport
o
ur
m
o
de
l
'
s
us
e
t
o
e
xp
e
d
i
t
e
m
e
l
a
n
om
a
de
t
e
c
t
i
on
a
nd
i
m
pro
ve
t
h
e
s
t
a
n
da
rd
of
c
a
re
pr
a
c
t
i
c
e
s
for
a
l
l
pa
t
i
e
nt
s
.
F
U
N
D
I
N
G
I
N
F
O
R
M
A
TI
O
N
T
he
a
u
t
hors
de
c
l
a
re
t
h
a
t
no
funds
,
gra
n
t
s
,
or
ot
h
e
r
s
up
port
w
e
re
re
c
e
i
ve
d
dur
i
ng
t
h
e
pre
p
a
r
a
t
i
on
o
f
t
hi
s
m
a
nus
c
r
i
pt
.
A
U
TH
O
R
C
O
N
TR
I
BU
TI
O
N
S
S
TA
T
EM
EN
T
T
hi
s
j
ou
rna
l
us
e
s
t
h
e
Cont
r
i
bu
t
o
r
Rol
e
s
T
a
xono
m
y
(C
Re
di
T
)
t
o
r
e
c
o
gni
z
e
i
ndi
v
i
du
a
l
a
u
t
hor
c
ont
r
i
bu
t
i
ons
,
r
e
du
c
e
a
ut
hors
h
i
p
di
s
pu
t
e
s
,
a
nd
fa
c
i
l
i
t
a
t
e
c
o
l
l
a
bora
t
i
on
.
N
ame
of
A
u
th
o
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
D
e
ba
du
t
t
a
S
a
hoo
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
S
oum
ya
M
i
s
hr
a
✓
✓
✓
✓
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
ft
w
a
re
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
rm
a
l
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
s
o
u
rc
e
s
D
:
D
a
t
a
Cu
ra
t
i
o
n
O
:
W
ri
t
i
n
g
-
O
ri
g
i
n
a
l
D
ra
ft
E
:
W
ri
t
i
n
g
-
Re
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
s
u
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
rv
i
s
i
o
n
P
:
P
ro
j
e
c
t
a
d
m
i
n
i
s
t
ra
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
s
i
t
i
o
n
Evaluation Warning : The document was created with Spire.PDF for Python.
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
IS
S
N
:
2722
-
3221
A
d
v
an
c
e
s
i
n
de
r
m
a
t
ol
o
gi
c
al
i
m
ag
i
ng:
e
nhan
c
i
ng
s
k
i
n
m
e
l
a
n
om
a
c
l
as
s
i
f
i
c
a
t
i
o
n
f
or
…
(D
e
b
adut
t
a
Sahoo
)
119
C
O
N
F
LI
C
T
O
F
I
N
T
ER
ES
T
S
TA
T
EM
EN
T
T
he
a
u
t
hors
s
t
a
t
e
no
c
onfl
i
c
t
o
f
i
nt
e
re
s
t
.
D
A
TA
A
V
A
I
LA
BI
LI
TY
D
a
t
a
a
va
i
l
a
bi
l
i
t
y
do
e
s
not
a
p
pl
y
t
o
t
h
i
s
p
a
p
e
r
a
s
n
o
n
e
w
da
t
a
w
e
re
c
r
e
a
t
e
d
or
a
na
l
yz
e
d
i
n
t
h
i
s
s
t
u
dy.
R
EF
ER
EN
C
ES
[1
]
K
.
S
u
d
h
a
rs
o
n
,
R.
S
a
n
t
h
i
y
a
,
a
n
d
S
.
V
a
rs
h
a
,
“
A
d
e
e
p
l
y
c
o
n
n
e
c
t
e
d
m
o
d
e
l
fo
r
h
i
g
h
-
p
re
c
i
s
i
o
n
c
l
a
s
s
i
fi
c
a
t
i
o
n
o
f
b
e
n
i
g
n
a
n
d
m
a
l
i
g
n
a
n
t
s
k
i
n
l
e
s
i
o
n
s
,
”
i
n
2
0
2
5
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
e
n
c
e
o
n
A
d
v
a
n
c
e
d
Co
m
p
u
t
i
n
g
T
e
c
h
n
o
l
o
g
i
e
s
(ICo
A
CT
)
,
M
a
r.
2
0
2
5
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
ICo
A
CT
6
3
3
3
9
.
2
0
2
5
.
1
1
0
0
4
8
1
4
.
[2
]
N
.
D
a
ra
p
a
n
e
n
i
e
t
a
l
.
,
“
E
a
r
l
y
d
e
t
e
c
t
i
o
n
o
f
s
k
i
n
c
a
n
c
e
r
-
s
o
l
u
t
i
o
n
fo
r
i
d
e
n
t
i
fy
i
n
g
a
n
d
d
e
fi
n
i
n
g
s
k
i
n
c
a
n
c
e
rs
u
s
i
n
g
A
I,
”
i
n
2
0
2
2
In
t
e
r
d
i
s
c
i
p
l
i
n
a
r
y
R
e
s
e
a
r
c
h
i
n
T
e
c
h
n
o
l
o
g
y
a
n
d
M
a
n
a
g
e
m
e
n
t
(IR
T
M
)
,
F
e
b
.
2
0
2
2
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
IRT
M
5
4
5
8
3
.
2
0
2
2
.
9
7
9
1
6
7
6
.
[3
]
In
t
e
rn
a
t
i
o
n
a
l
A
g
e
n
c
y
fo
r
Re
s
e
a
rc
h
o
n
Ca
n
c
e
r
W
o
rl
d
H
e
a
l
t
h
O
rg
a
n
i
z
a
t
i
o
n
,
“
G
l
o
b
a
l
c
a
n
c
e
r
o
b
s
e
rv
a
t
o
ry
,
”
In
t
e
rn
a
t
i
o
n
a
l
A
g
e
n
c
y
fo
r
Re
s
e
a
rc
h
o
n
Ca
n
c
e
r
W
o
rl
d
H
e
a
l
t
h
O
r
g
a
n
i
z
a
t
i
o
n
,
2
0
2
2
.
[4
]
P
.
K
a
u
s
h
i
k
,
Y
.
Ch
o
p
ra
,
A
.
K
a
j
l
a
,
M
.
P
o
o
n
i
a
,
A
.
K
h
a
n
,
a
n
d
D
.
Y
a
d
a
v
,
“
A
I
-
p
o
w
e
re
d
d
e
rm
a
t
o
l
o
g
y
:
a
c
h
i
e
v
i
n
g
d
e
rm
a
t
o
l
o
g
i
s
t
-
g
ra
d
e
s
k
i
n
c
a
n
c
e
r
c
l
a
s
s
i
fi
c
a
t
i
o
n
,
”
i
n
2
0
2
4
IE
E
E
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
e
n
c
e
o
n
In
t
e
r
d
i
s
c
i
p
l
i
n
a
r
y
A
p
p
r
o
a
c
h
e
s
i
n
T
e
c
h
n
o
l
o
g
y
a
n
d
M
a
n
a
g
e
m
e
n
t
f
o
r
S
o
c
i
a
l
In
n
o
v
a
t
i
o
n
(IA
T
M
S
I)
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
IA
T
M
S
I6
0
4
2
6
.
2
0
2
4
.
1
0
5
0
2
6
6
4
.
[5
]
A
.
E
s
t
e
v
a
e
t
a
l
.
,
“
D
e
rm
a
t
o
l
o
g
i
s
t
-
l
e
v
e
l
c
l
a
s
s
i
fi
c
a
t
i
o
n
o
f
s
k
i
n
c
a
n
c
e
r
w
i
t
h
d
e
e
p
n
e
u
ra
l
n
e
t
w
o
rk
s
,
”
Na
t
u
r
e
,
v
o
l
.
5
4
2
,
n
o
.
7
6
3
9
,
p
p
.
1
1
5
–
1
1
8
,
F
e
b
.
2
0
1
7
,
d
o
i
:
1
0
.
1
0
3
8
/
n
a
t
u
re
2
1
0
5
6
.
[6
]
M
.
H
.
J
a
v
i
d
,
“
M
e
l
a
n
o
m
a
s
k
i
n
c
a
n
c
e
r
d
a
t
a
s
e
t
o
f
1
0
0
0
0
i
m
a
g
e
s
,
”
k
a
g
g
l
e
.
c
o
m
.
[O
n
l
i
n
e
].
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
w
w
w
.
k
a
g
g
l
e
.
c
o
m
/
d
a
t
a
s
e
t
s
/
h
a
s
n
a
i
n
j
a
v
e
d
/
m
e
l
a
n
o
m
a
-
s
k
i
n
-
c
a
n
c
e
r
-
d
a
t
a
s
e
t
-
of
-
1
0
0
0
0
-
i
m
a
g
e
s
[7
]
C.
F
a
n
c
o
n
i
,
“
S
k
i
n
c
a
n
c
e
r:
m
a
l
i
g
n
a
n
t
v
s
.
b
e
n
i
g
n
,
”
k
a
g
g
l
e
.
c
o
m
.
[O
n
l
i
n
e
].
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
w
w
w
.
k
a
g
g
l
e
.
c
o
m
/
d
a
t
a
s
e
t
s
/
fa
n
c
o
n
i
c
/
s
k
i
n
-
c
a
n
c
e
r
-
m
a
l
i
g
n
a
n
t
-
vs
-
b
e
n
i
g
n
[8
]
O
.
A
k
i
n
ri
n
a
d
e
a
n
d
C.
D
u
,
“
S
k
i
n
c
a
n
c
e
r
d
e
t
e
c
t
i
o
n
u
s
i
n
g
d
e
e
p
m
a
c
h
i
n
e
l
e
a
rn
i
n
g
t
e
c
h
n
i
q
u
e
s
,
”
In
t
e
l
l
i
g
e
n
c
e
-
B
a
s
e
d
M
e
d
i
c
i
n
e
,
v
o
l
.
1
1
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
b
m
e
d
.
2
0
2
4
.
1
0
0
1
9
1
.
[9
]
G
.
S
.
V
i
r
g
e
n
s
,
J
.
A
.
T
e
o
d
o
r
o
,
E
.
Ia
r
u
s
s
i
,
T
.
R
o
d
ri
g
u
e
s
,
a
n
d
D
.
T
.
A
m
a
r
a
l
,
“
E
n
h
a
n
c
i
n
g
a
n
d
a
d
v
a
n
c
e
m
e
n
t
s
i
n
d
e
e
p
l
e
a
r
n
i
n
g
f
o
r
m
e
l
a
n
o
m
a
d
e
t
e
c
t
i
o
n
:
a
c
o
m
p
r
e
h
e
n
s
i
v
e
r
e
v
i
e
w
,
”
C
o
m
p
u
t
e
r
s
i
n
B
i
o
l
o
g
y
a
n
d
M
e
d
i
c
i
n
e
,
v
o
l
.
1
9
4
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
m
p
b
i
o
m
e
d
.
2
0
2
5
.
1
1
0
5
3
3
.
[1
0
]
A
.
A
.
S
h
i
rs
a
t
h
,
T
.
S
h
a
h
,
a
n
d
V
.
A
.
S
h
i
rs
a
t
h
,
“
D
e
e
p
l
e
a
rn
i
n
g
b
a
s
e
d
s
y
s
t
e
m
f
o
r
e
a
rl
y
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
s
s
i
fi
c
a
t
i
o
n
o
f
m
e
l
a
n
o
m
a
s
k
i
n
c
a
n
c
e
r
-
a
s
y
s
t
e
m
a
t
i
c
re
v
i
e
w
,
”
J
o
u
r
n
a
l
o
f
Ca
r
c
i
n
o
g
e
n
e
s
i
s
,
v
o
l
.
2
4
,
n
o
.
2
,
p
p
.
2
6
–
3
4
,
2
0
2
5
,
d
o
i
:
1
0
.
6
4
1
4
9
/
J
.
Ca
rc
i
n
o
g
.
2
4
.
2
.
2
6
-
3
4
.
[1
1
]
S
.
J
a
i
n
,
V
.
J
a
g
t
a
p
,
a
n
d
N
.
P
i
s
e
,
“
Co
m
p
u
t
e
r
a
i
d
e
d
m
e
l
a
n
o
m
a
s
k
i
n
c
a
n
c
e
r
d
e
t
e
c
t
i
o
n
u
s
i
n
g
i
m
a
g
e
p
ro
c
e
s
s
i
n
g
,
”
P
r
o
c
e
d
i
a
Co
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
4
8
,
p
p
.
7
3
5
–
7
4
0
,
2
0
1
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
ro
c
s
.
2
0
1
5
.
0
4
.
2
0
9
.
[1
2
]
E
.
V
o
c
a
t
u
r
o
,
D
.
P
e
r
n
a
,
a
n
d
E
.
Z
u
m
p
a
n
o
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s
f
o
r
a
u
t
o
m
a
t
e
d
m
e
l
a
n
o
m
a
d
e
t
e
c
t
i
o
n
,
”
P
r
o
c
e
e
d
i
n
g
s
-
2
0
1
9
IE
E
E
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
B
i
o
i
n
f
o
r
m
a
t
i
c
s
a
n
d
B
i
o
m
e
d
i
c
i
n
e
,
(B
IB
M
)
,
p
p
.
2
3
1
0
–
2
3
1
7
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
BIBM
4
7
2
5
6
.
2
0
1
9
.
8
9
8
3
1
6
5
.
[1
3
]
A
.
M
a
h
b
o
d
,
G
.
S
c
h
a
e
f
e
r,
I.
E
l
l
i
n
g
e
r,
R.
E
c
k
e
r,
A
.
P
i
t
i
o
t
,
a
n
d
C.
W
a
n
g
,
“
F
u
s
i
n
g
fi
n
e
-
t
u
n
e
d
d
e
e
p
f
e
a
t
u
re
s
fo
r
s
k
i
n
l
e
s
i
o
n
c
l
a
s
s
i
fi
c
a
t
i
o
n
,
”
C
o
m
p
u
t
e
r
i
z
e
d
M
e
d
i
c
a
l
Im
a
g
i
n
g
a
n
d
G
r
a
p
h
i
c
s
,
v
o
l
.
7
1
,
p
p
.
1
9
–
2
9
,
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
m
p
m
e
d
i
m
a
g
.
2
0
1
8
.
1
0
.
0
0
7
.
[1
4
]
H
.
A
.
H
a
e
n
s
s
l
e
e
t
a
l
.
,
“
M
a
n
a
g
a
i
n
s
t
m
a
c
h
i
n
e
:
d
i
a
g
n
o
s
t
i
c
p
e
rfo
rm
a
n
c
e
o
f
a
d
e
e
p
l
e
a
rn
i
n
g
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
ra
l
n
e
t
w
o
rk
fo
r
d
e
rm
o
s
c
o
p
i
c
m
e
l
a
n
o
m
a
re
c
o
g
n
i
t
i
o
n
i
n
c
o
m
p
a
ri
s
o
n
t
o
5
8
d
e
rm
a
t
o
l
o
g
i
s
t
s
,
”
A
n
n
a
l
s
o
f
O
n
c
o
l
o
g
y
,
v
o
l
.
2
9
,
n
o
.
8
,
p
p
.
1
8
3
6
–
1
8
4
2
,
A
u
g
.
2
0
1
8
,
d
o
i
:
1
0
.
1
0
9
3
/
a
n
n
o
n
c
/
m
d
y
1
6
6
.
[1
5
]
H
.
B
h
a
t
t
,
V
.
S
h
a
h
,
K
.
S
h
a
h
,
R
.
S
h
a
h
,
a
n
d
M
.
S
h
a
h
,
“
S
t
a
t
e
-
of
-
t
h
e
-
a
rt
m
a
c
h
i
n
e
l
e
a
rn
i
n
g
t
e
c
h
n
i
q
u
e
s
f
o
r
m
e
l
a
n
o
m
a
s
k
i
n
c
a
n
c
e
r
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
s
s
i
fi
c
a
t
i
o
n
:
a
c
o
m
p
r
e
h
e
n
s
i
v
e
r
e
v
i
e
w
,
”
I
n
t
e
l
l
i
g
e
n
t
M
e
d
i
c
i
n
e
,
v
o
l
.
3
,
n
o
.
3
,
p
p
.
1
8
0
–
1
9
0
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
m
e
d
.
2
0
2
2
.
0
8
.
0
0
4
.
[1
6
]
F
.
A
l
a
m
,
A
.
U
l
l
a
h
,
D
.
S
h
a
h
,
S
.
A
l
i
,
a
n
d
M
.
T
a
h
i
r
,
“
A
rt
i
fi
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
i
n
m
e
l
a
n
o
m
a
d
e
t
e
c
t
i
o
n
:
a
re
v
i
e
w
o
f
c
u
rre
n
t
t
e
c
h
n
o
l
o
g
i
e
s
a
n
d
fu
t
u
re
d
i
re
c
t
i
o
n
s
,
”
In
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
In
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
,
n
o
.
1
,
J
a
n
.
2
0
2
5
,
d
o
i
:
1
0
.
1
1
5
5
/
i
n
t
/
3
1
6
4
9
5
2
.
[1
7
]
M
.
K
re
o
u
z
i
e
t
a
l
.
,
“
D
e
e
p
l
e
a
rn
i
n
g
fo
r
m
e
l
a
n
o
m
a
d
e
t
e
c
t
i
o
n
:
a
d
e
e
p
l
e
a
rn
i
n
g
a
p
p
ro
a
c
h
t
o
d
i
ffe
re
n
t
i
a
t
i
n
g
m
a
l
i
g
n
a
n
t
m
e
l
a
n
o
m
a
fro
m
b
e
n
i
g
n
m
e
l
a
n
o
c
y
t
i
c
n
e
v
i
,
”
Ca
n
c
e
r
s
,
v
o
l
.
1
7
,
n
o
.
1
,
D
e
c
.
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
c
a
n
c
e
rs
1
7
0
1
0
0
2
8
.
[1
8
]
H
.
L
.
G
u
ru
ra
j
,
N
.
M
a
n
j
u
,
A
.
N
a
g
a
rj
u
n
,
V
.
N
.
M
.
A
ra
d
h
y
a
,
a
n
d
F
.
F
l
a
m
m
i
n
i
,
“
D
e
e
p
S
k
i
n
:
a
d
e
e
p
l
e
a
r
n
i
n
g
a
p
p
ro
a
c
h
fo
r
s
k
i
n
c
a
n
c
e
r
c
l
a
s
s
i
fi
c
a
t
i
o
n
,
”
IE
E
E
A
c
c
e
s
s
,
v
o
l
.
1
1
,
p
p
.
5
0
2
0
5
–
5
0
2
1
4
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
A
CCE
S
S
.
2
0
2
3
.
3
2
7
4
8
4
8
.
[1
9
]
U
.
B
.
A
n
s
a
ri
a
n
d
T
.
S
a
ro
d
e
,
“
S
k
i
n
c
a
n
c
e
r
d
e
t
e
c
t
i
o
n
u
s
i
n
g
i
m
a
g
e
p
ro
c
e
s
s
i
n
g
,
”
In
t
e
r
n
a
t
i
o
n
a
l
R
e
s
e
a
r
c
h
J
o
u
r
n
a
l
o
f
E
n
g
i
n
e
e
r
i
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
(IR
J
E
T
)
,
v
o
l
.
4
,
n
o
.
4
,
p
p
.
2
8
7
5
–
2
8
8
1
,
2
0
1
7
.
[2
0
]
V
.
G
u
l
s
h
a
n
e
t
a
l
.
,
“
D
e
v
e
l
o
p
m
e
n
t
a
n
d
v
a
l
i
d
a
t
i
o
n
o
f
a
d
e
e
p
l
e
a
rn
i
n
g
a
l
g
o
ri
t
h
m
fo
r
d
e
t
e
c
t
i
o
n
o
f
d
i
a
b
e
t
i
c
re
t
i
n
o
p
a
t
h
y
i
n
re
t
i
n
a
l
fu
n
d
u
s
p
h
o
t
o
g
r
a
p
h
s
,
”
J
A
M
A
-
J
o
u
r
n
a
l
o
f
t
h
e
A
m
e
r
i
c
a
n
M
e
d
i
c
a
l
A
s
s
o
c
i
a
t
i
o
n
,
v
o
l
.
3
1
6
,
n
o
.
2
2
,
p
p
.
2
4
0
2
–
2
4
1
0
,
2
0
1
6
,
d
o
i
:
1
0
.
1
0
0
1
/
j
a
m
a
.
2
0
1
6
.
1
7
2
1
6
.
[2
1
]
J
.
D
e
F
a
u
w
e
t
a
l
.
,
“
Cl
i
n
i
c
a
l
l
y
a
p
p
l
i
c
a
b
l
e
d
e
e
p
l
e
a
rn
i
n
g
f
o
r
d
i
a
g
n
o
s
i
s
a
n
d
r
e
fe
rra
l
i
n
re
t
i
n
a
l
d
i
s
e
a
s
e
,
”
Na
t
u
r
e
M
e
d
i
c
i
n
e
,
v
o
l
.
2
4
,
n
o
.
9
,
p
p
.
1
3
4
2
–
1
3
5
0
,
S
e
p
.
2
0
1
8
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
9
1
-
0
1
8
-
0
1
0
7
-
6.
[2
2
]
T.
-
M
.
Ch
i
u
,
Y
.
-
C
.
L
i
,
I
.
-
C.
Ch
i
,
a
n
d
M
.
-
H
.
T
s
e
n
g
,
“
A
I
-
d
ri
v
e
n
e
n
h
a
n
c
e
m
e
n
t
o
f
s
k
i
n
c
a
n
c
e
r
d
i
a
g
n
o
s
i
s
:
a
t
w
o
-
s
t
a
g
e
v
o
t
i
n
g
e
n
s
e
m
b
l
e
a
p
p
ro
a
c
h
u
s
i
n
g
d
e
rm
o
s
c
o
p
i
c
d
a
t
a
,
”
Ca
n
c
e
r
s
,
v
o
l
.
1
7
,
n
o
.
1
,
2
0
2
5
,
d
o
i
:
1
0
.
3
3
9
0
/
c
a
n
c
e
rs
1
7
0
1
0
1
3
7
.
[2
3
]
S
.
A
b
u
ra
s
s
,
O
.
D
o
rg
h
a
m
,
J
.
A
l
S
h
a
q
s
i
,
M
.
A
.
Ru
m
m
a
n
,
a
n
d
O
.
A
l
-
K
a
d
i
,
“
V
i
s
i
o
n
t
ra
n
s
fo
rm
e
rs
i
n
m
e
d
i
c
a
l
i
m
a
g
i
n
g
:
a
c
o
m
p
re
h
e
n
s
i
v
e
re
v
i
e
w
o
f
a
d
v
a
n
c
e
m
e
n
t
s
a
n
d
a
p
p
l
i
c
a
t
i
o
n
s
a
c
r
o
s
s
m
u
l
t
i
p
l
e
d
i
s
e
a
s
e
s
,
”
J
o
u
r
n
a
l
o
f
Im
a
g
i
n
g
In
f
o
r
m
a
t
i
c
s
i
n
M
e
d
i
c
i
n
e
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
2
7
8
-
0
2
5
-
0
1
4
8
1
-
y.
[2
4
]
M
.
S
t
e
fa
ń
c
z
y
k
a
n
d
T
.
Bo
c
h
e
ń
s
k
i
,
“
M
i
x
i
n
g
d
e
e
p
l
e
a
rn
i
n
g
w
i
t
h
c
l
a
s
s
i
c
a
l
v
i
s
i
o
n
fo
r
o
b
j
e
c
t
re
c
o
g
n
i
t
i
o
n
,
”
J
o
u
r
n
a
l
o
f
W
S
CG
,
v
o
l
.
2
8
,
n
o
.
1
–
2
,
p
p
.
1
4
7
–
1
5
4
,
2
0
2
0
,
d
o
i
:
1
0
.
2
4
1
3
2
/
J
W
S
CG
.
2
0
2
0
.
2
8
.
1
8
.
[2
5
]
R.
S
a
b
i
r
a
n
d
T
.
M
e
h
m
o
o
d
,
“
Cl
a
s
s
i
fi
c
a
t
i
o
n
o
f
m
e
l
a
n
o
m
a
s
k
i
n
c
a
n
c
e
r
b
a
s
e
d
o
n
i
m
a
g
e
d
a
t
a
s
e
t
u
s
i
n
g
d
i
ffe
re
n
t
n
e
u
ra
l
n
e
t
w
o
rk
s
,
”
S
c
i
e
n
t
i
f
i
c
R
e
p
o
r
t
s
,
v
o
l
.
1
4
,
n
o
.
1
,
N
o
v
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
9
8
-
024
-
7
5
1
4
3
-
4.
[2
6
]
P
.
N
.
S
.
C
.
T
e
j
a
,
T
.
B.
K
ri
s
h
n
a
,
A
.
K
.
R
.
P
o
re
d
d
y
,
a
n
d
P
.
K
o
k
i
l
,
“
L
i
g
h
t
w
e
i
g
h
t
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
fo
r
m
e
l
a
n
o
m
a
c
l
a
s
s
i
fi
c
a
t
i
o
n
i
n
d
e
rm
o
s
c
o
p
y
i
m
a
g
e
s
fo
r
s
m
a
rt
h
e
a
l
t
h
c
a
re
,
”
i
n
2
0
2
4
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
e
n
c
e
o
n
W
i
r
e
l
e
s
s
Co
m
m
u
n
i
c
a
t
i
o
n
s
S
i
g
n
a
l
P
r
o
c
e
s
s
i
n
g
a
n
d
Ne
t
wo
r
k
i
n
g
(W
i
S
P
NE
T
)
,
M
a
r.
2
0
2
4
,
p
p
.
1
–
6,
d
o
i
:
1
0
.
1
1
0
9
/
W
i
S
P
N
E
T
6
1
4
6
4
.
2
0
2
4
.
1
0
5
3
2
9
2
3
.
[2
7
]
M
.
F
.
A
l
m
u
fa
re
h
,
N
.
T
a
ri
q
,
M
.
H
u
m
a
y
u
n
,
a
n
d
F
.
A
.
K
h
a
n
,
“
M
e
l
a
n
o
m
a
i
d
e
n
t
i
fi
c
a
t
i
o
n
a
n
d
c
l
a
s
s
i
fi
c
a
t
i
o
n
m
o
d
e
l
b
a
s
e
d
o
n
fi
n
e
-
t
u
n
e
d
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
ra
l
n
e
t
w
o
r
k
,
”
D
i
g
i
t
a
l
H
e
a
l
t
h
,
v
o
l
.
1
0
,
S
e
p
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
7
7
/
2
0
5
5
2
0
7
6
2
4
1
2
5
3
7
5
7
.
[2
8
]
Y
.
Q
i
n
e
t
a
l
.
,
“
A
u
t
o
fo
c
u
s
l
a
y
e
r
fo
r
s
e
m
a
n
t
i
c
s
e
g
m
e
n
t
a
t
i
o
n
,
”
i
n
M
e
d
i
c
a
l
I
m
a
g
e
Co
m
p
u
t
i
n
g
a
n
d
Co
m
p
u
t
e
r
A
s
s
i
s
t
e
d
In
t
e
r
v
e
n
t
i
o
n
(M
ICCA
I),
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
e
n
c
e
o
n
M
e
d
i
c
a
l
Im
a
g
e
Co
m
p
u
t
i
n
g
a
n
d
Co
m
p
u
t
e
r
-
A
s
s
i
s
t
e
d
In
t
e
r
v
e
n
t
i
o
n
,
2
0
1
8
,
p
p
.
6
0
3
–
61
1
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
030
-
0
0
9
3
1
-
1
_
6
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2722
-
3221
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
,
V
o
l
.
7
,
N
o
.
1
,
M
a
rc
h
20
26
:
111
-
1
20
120
[2
9
]
Y
.
M
o
u
s
a
,
R
.
T
a
h
a
,
R
.
K
a
u
r,
a
n
d
S
.
A
fi
fi
,
“
M
e
l
a
n
o
m
a
c
l
a
s
s
i
fi
c
a
t
i
o
n
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
i
n
P
r
o
c
e
e
d
i
n
g
s
,
Im
a
g
e
a
n
d
V
i
d
e
o
T
e
c
h
n
o
l
o
g
y
:
1
1
t
h
P
a
c
i
f
i
c
-
R
i
m
S
y
m
p
o
s
i
u
m
,
P
S
I
V
T
2
0
2
3
,
A
u
c
k
l
a
n
d
,
Ne
w
Z
e
a
l
a
n
d
,
No
v
e
m
b
e
r
2
2
–
2
4
,
2023
,
2
0
2
4
,
p
p
.
2
5
9
–
2
7
2
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
9
8
1
-
97
-
0
3
7
6
-
0
_
2
0
.
[3
0
]
G
.
V
e
ro
n
e
s
i
,
N
.
Cu
rt
i
,
A
.
G
a
rd
i
n
i
,
G
.
Q
u
e
rz
o
l
i
,
G
.
Ca
s
t
e
l
l
a
n
i
,
a
n
d
E
.
D
i
k
a
,
“
M
a
c
h
i
n
e
l
e
a
rn
i
n
g
t
o
d
e
t
e
c
t
m
e
l
a
n
o
m
a
e
x
p
l
o
i
t
i
n
g
n
u
c
l
e
i
m
o
rp
h
o
l
o
g
y
a
n
d
s
p
a
t
i
a
l
o
rg
a
n
i
z
a
t
i
o
n
,
”
S
c
i
e
n
t
i
f
i
c
R
e
p
o
r
t
s
, v
o
l
.
1
5
,
n
o
.
1
,
J
u
l
.
2
0
2
5
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
9
8
-
0
2
5
-
0
2
9
1
3
-
z.
[3
1
]
N
.
D
a
ra
p
a
n
e
n
i
e
t
a
l
.
,
“
CO
V
ID
1
9
s
e
v
e
ri
t
y
o
f
p
n
e
u
m
o
n
i
a
a
n
a
l
y
s
i
s
u
s
i
n
g
c
h
e
s
t
x
ra
y
s
,
”
2
0
2
0
IE
E
E
1
5
t
h
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
e
n
c
e
o
n
In
d
u
s
t
r
i
a
l
a
n
d
In
f
o
r
m
a
t
i
o
n
S
y
s
t
e
m
s
(ICIIS
)
P
r
o
c
e
e
d
i
n
g
s
,
p
p
.
3
8
1
–
3
8
6
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
ICIIS
5
1
1
4
0
.
2
0
2
0
.
9
3
4
2
7
0
2
.
[3
2
]
M
.
Q
.
K
h
a
n
e
t
a
l
.
,
“
C
l
a
s
s
i
fi
c
a
t
i
o
n
o
f
m
e
l
a
n
o
m
a
a
n
d
n
e
v
u
s
i
n
d
i
g
i
t
a
l
i
m
a
g
e
s
fo
r
d
i
a
g
n
o
s
i
s
o
f
s
k
i
n
c
a
n
c
e
r,
”
IE
E
E
A
c
c
e
s
s
,
v
o
l
.
7
,
p
p
.
9
0
1
3
2
–
9
0
1
4
4
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
A
CCE
S
S
.
2
0
1
9
.
2
9
2
6
8
3
7
.
[3
3
]
H
.
C.
Re
i
s
a
n
d
V
.
T
u
rk
,
“
F
u
s
i
o
n
o
f
t
ra
n
s
fo
rm
e
r
a
t
t
e
n
t
i
o
n
a
n
d
CN
N
fe
a
t
u
re
s
fo
r
s
k
i
n
c
a
n
c
e
r
d
e
t
e
c
t
i
o
n
,
”
A
p
p
l
i
e
d
S
o
f
t
Co
m
p
u
t
i
n
g
,
v
o
l
.
1
6
4
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
s
o
c
.
2
0
2
4
.
1
1
2
0
1
3
.
[3
4
]
V
.
V
e
n
u
g
o
p
a
l
,
N
.
I.
Ra
j
,
M
.
K
.
N
a
t
h
,
a
n
d
N
.
S
t
e
p
h
e
n
,
“
A
d
e
e
p
n
e
u
ra
l
n
e
t
w
o
rk
u
s
i
n
g
m
o
d
i
fi
e
d
E
ffi
c
i
e
n
t
N
e
t
fo
r
s
k
i
n
c
a
n
c
e
r
d
e
t
e
c
t
i
o
n
i
n
d
e
rm
o
s
c
o
p
i
c
i
m
a
g
e
s
,
”
D
e
c
i
s
i
o
n
A
n
a
l
y
t
i
c
s
J
o
u
r
n
a
l
,
v
o
l
.
8
,
S
e
p
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
d
a
j
o
u
r.
2
0
2
3
.
1
0
0
2
7
8
.
[3
5
]
N
.
M
e
l
a
rk
o
d
e
,
K
.
S
ri
n
i
v
a
s
a
n
,
S
.
M
.
Q
a
i
s
a
r
,
a
n
d
P
.
P
l
a
w
i
a
k
,
“
A
I
-
p
o
w
e
re
d
d
i
a
g
n
o
s
i
s
o
f
s
k
i
n
c
a
n
c
e
r:
a
c
o
n
t
e
m
p
o
ra
ry
re
v
i
e
w
,
o
p
e
n
c
h
a
l
l
e
n
g
e
s
a
n
d
fu
t
u
re
re
s
e
a
rc
h
d
i
re
c
t
i
o
n
s
,
”
Ca
n
c
e
r
s
,
v
o
l
.
1
5
,
n
o
.
4
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
c
a
n
c
e
rs
1
5
0
4
1
1
8
3
.
[3
6
]
A
.
A
.
L
o
u
i
s
o
n
a
n
d
B.
B.
S
u
j
i
t
h
a
,
“
M
O
O
N
-
N
E
T
:
d
e
e
p
L
e
a
rn
i
n
g
b
a
s
e
d
m
e
l
a
n
o
m
a
s
k
i
n
l
e
s
i
o
n
c
l
a
s
s
i
fi
c
a
t
i
o
n
u
s
i
n
g
S
E
-
D
N
e
t
a
n
d
D
BN
,
”
i
n
2
0
2
5
IE
E
E
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
e
n
c
e
o
n
Co
m
p
u
t
e
,
Co
n
t
r
o
l
,
Ne
t
w
o
r
k
&
P
h
o
t
o
n
i
c
s
(ICCC
NP
),
B
a
n
g
a
l
o
r
e
,
In
d
i
a
,
2
0
2
5
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
i
c
c
c
n
p
6
3
9
1
4
.
2
0
2
5
.
1
1
2
3
3
4
2
7
.
BI
O
G
R
A
P
H
I
ES
O
F
A
U
T
H
O
R
S
D
e
b
ad
u
t
t
a
S
ah
oo
i
s
a
n
e
n
t
h
us
i
a
s
t
i
c
e
l
e
c
t
r
o
ni
c
s
&
c
o
m
m
u
ni
c
a
t
i
on
e
n
gi
ne
e
r
s
ki
l
l
e
d
i
n
P
C
B
d
e
s
i
gn
,
V
L
S
I
,
a
n
d
s
i
gna
l
pr
oc
e
s
s
i
ng
.
A
s
a
l
i
c
e
ns
e
d
H
A
M
op
e
r
a
t
or
,
I
t
hr
i
ve
i
n
R
&
D
a
nd
t
r
ou
bl
e
s
ho
ot
i
ng
w
i
t
h
i
n
e
m
be
dde
d
s
ys
t
e
m
s
.
P
r
o
f
i
c
i
e
nt
i
n
C
a
nd
P
y
t
hon
,
H
i
s
pa
s
s
i
on
a
t
e
a
b
ou
t
l
e
ve
r
a
g
i
n
g
A
I
t
o
i
nn
o
va
t
e
a
n
d
e
n
h
a
n
c
e
t
e
c
hn
o
l
og
y
f
o
r
e
f
f
e
c
t
i
v
e
p
r
o
b
l
e
m
-
s
o
l
v
i
n
g
i
n
t
h
e
e
ng
i
n
e
e
r
i
n
g
do
m
a
i
n
.
H
e
c
a
n
b
e
c
o
n
t
a
c
t
e
d
a
t
e
m
a
i
l
:
d
e
b
a
d
u
t
t
a
s
a
ho
o
22
5
5
@g
m
a
i
l
.
c
o
m
.
S
ou
m
ya
M
i
s
h
r
a
i
s
a
s
e
a
s
o
ne
d
pr
o
f
e
s
s
i
ona
l
w
i
t
h
ov
e
r
1
6
ye
a
r
s
o
f
d
i
ve
r
s
e
e
xpe
r
i
e
nc
e
e
nc
o
m
p
a
s
s
i
ng
t
e
a
c
hi
ng,
r
e
s
e
a
r
c
h
a
nd
d
e
ve
l
op
m
e
n
t
(
R
&
D
)
,
a
nd
a
d
m
i
ni
s
t
r
a
t
i
o
n.
C
ur
r
e
nt
l
y
s
e
r
vi
n
g
a
s
t
he
I
Q
A
C
i
ns
t
i
t
u
t
i
ona
l
c
o
or
d
i
n
a
t
o
r
,
c
us
t
o
di
a
n
o
f
s
p
a
c
e
c
l
u
b
-
C
G
U
,
a
nd
I
nt
e
r
n
a
t
i
o
na
l
P
r
og
r
a
m
s
&
O
ut
r
e
a
c
h
C
oo
r
d
i
na
t
o
r
,
s
he
b
r
i
n
gs
e
x
t
e
n
s
i
v
e
e
xp
e
r
t
i
s
e
i
n
e
nha
nc
i
ng
a
c
a
d
e
m
i
c
qu
a
l
i
t
y
a
s
s
ur
a
nc
e
a
nd
f
o
s
t
e
r
i
ng
g
l
oba
l
pa
r
t
n
e
r
s
h
i
ps
.
H
e
r
r
e
s
e
a
r
c
h
s
p
e
c
i
a
l
i
z
a
t
i
o
n
l
i
e
s
i
n
w
i
r
e
l
e
s
s
c
om
m
uni
c
a
t
i
on,
s
i
gna
l
p
r
oc
e
s
s
i
n
g,
a
nd
t
he
a
ppl
i
c
a
t
i
on
of
a
r
t
i
f
i
c
i
a
l
i
nt
e
l
l
i
ge
n
c
e
/
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
(
A
I
/
M
L
)
i
n
t
he
s
e
do
m
a
i
n
s
.
S
he
c
a
n
be
c
on
t
a
c
t
e
d
a
t
e
m
a
i
l
:
s
ou
m
ya
@c
g
u
-
o
di
s
ha
.
a
c
.
i
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.