T
E
L
K
O
M
NIKA
T
elec
o
mm
un
ica
t
io
n Co
m
pu
t
i
ng
E
lect
ro
nics
a
nd
Co
ntr
o
l
Vo
l.
24
,
No
.
4
,
A
u
g
u
s
t
20
26
,
p
p
.
1
3
3
1
~
1
3
4
1
I
SS
N:
1
6
9
3
-
6
9
3
0
,
DOI
:
1
0
.
1
2
9
2
8
/
T
E
L
KOM
NI
K
A
.
v
24
i
4
.
27700
1331
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//telko
mn
ika
.
u
a
d
.
a
c.
i
d
Co
m
pa
ra
tive
ev
a
lua
tion o
f
cla
ss
ica
l
and
m
a
chine
lea
r
ning
m
e
thods
f
o
r
m
edi
ca
l i
m
a
g
e e
nha
nc
e
m
e
nt
M
d.
M
ehedi H
a
s
a
n
1
,
Su
j
o
n Cha
nd
ra
Su
t
ra
dh
a
r
1
,
Z
a
n
na
t
ul F
er
du
s
h
ie
2
,
Ra
bey
a
B
a
s
ri
1
1
D
e
p
a
r
t
me
n
t
o
f
G
e
n
e
r
a
l
Ed
u
c
a
t
i
o
n
,
U
n
i
v
e
r
si
t
y
o
f
F
r
o
n
t
i
e
r
T
e
c
h
n
o
l
o
g
y
,
G
a
z
i
p
u
r
,
B
a
n
g
l
a
d
e
sh
2
D
e
p
a
r
t
me
n
t
o
f
E
d
u
c
a
t
i
o
n
a
l
T
e
c
h
n
o
l
o
g
y
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
U
n
i
v
e
r
si
t
y
o
f
F
r
o
n
t
i
e
r
T
e
c
h
n
o
l
o
g
y
,
G
a
z
i
p
u
r
,
B
a
n
g
l
a
d
e
s
h
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Dec
3
,
2025
R
ev
i
s
ed
A
p
r
18
,
2
0
2
6
A
cc
ep
ted
Ma
y
25
,
2026
M
e
d
ica
l
im
a
g
in
g
is
c
rit
ica
l
f
o
r
d
iag
n
o
stic
a
c
c
u
ra
c
y
,
y
e
t
ra
w
i
m
a
g
e
s
o
f
ten
su
ff
e
r
f
ro
m
n
o
ise
a
n
d
lo
w
c
o
n
tras
t.
T
h
is
stu
d
y
p
ro
v
id
e
s
a
c
o
m
p
a
ra
ti
v
e
e
v
a
lu
a
ti
o
n
o
f
c
las
sic
a
l
m
e
th
o
d
s,
n
a
m
e
l
y
th
e
L
a
p
lac
e
tran
s
f
o
r
m
(
LT
),
S
o
b
e
l
o
p
e
ra
to
r
(S
O),
a
n
d
h
isto
g
ra
m
e
q
u
a
li
z
a
t
io
n
(HE),
a
g
a
in
st
a
d
a
ta
-
d
riv
e
n
c
o
n
v
o
lu
ti
o
n
a
l
n
e
u
ra
l
n
e
tw
o
rk
(CNN
)
u
sin
g
th
e
m
u
sc
u
lo
sk
e
leta
l
r
a
d
io
g
ra
p
h
s
(M
URA
)
a
n
d
Hu
m
a
n
M
e
ta
p
n
e
u
m
o
v
iru
s
(HM
P
V
)
l
u
n
g
c
o
m
p
u
ted
to
m
o
g
ra
p
h
y
(C
T
)
d
a
tas
e
ts.
W
h
il
e
q
u
a
n
ti
tativ
e
a
n
a
ly
sis
sh
o
w
s
th
a
t
HE
a
n
d
S
O
sig
n
if
ica
n
tl
y
o
u
tp
e
rf
o
rm
o
th
e
r
m
e
th
o
d
s
in
iso
lat
e
d
c
o
n
tras
t
e
n
h
a
n
c
e
m
e
n
t
a
n
d
e
d
g
e
d
e
f
in
it
io
n
,
t
h
e
y
o
f
ten
in
tro
d
u
c
e
a
rti
f
a
c
ts.
In
c
o
n
tras
t,
t
h
e
CNN
-
b
a
se
d
a
p
p
r
o
a
c
h
d
e
m
o
n
stra
tes
su
p
e
rio
r
d
e
tail
p
re
se
rv
a
ti
o
n
a
n
d
e
n
tro
p
y
,
o
f
fe
rin
g
a
m
o
re
b
a
lan
c
e
d
a
n
d
a
d
a
p
ti
v
e
so
lu
ti
o
n
f
o
r
d
iv
e
rse
d
ia
g
n
o
stic
re
q
u
irem
e
n
ts.
Ou
r
f
in
d
in
g
s
sta
ti
stica
ll
y
v
a
li
d
a
te
th
a
t
a
lt
h
o
u
g
h
c
las
sic
a
l
o
p
e
ra
to
rs
re
m
a
in
h
ig
h
ly
e
ff
e
c
ti
v
e
f
o
r
sp
e
c
if
ic
b
o
u
n
d
a
ry
d
e
tec
ti
o
n
tas
k
s,
m
a
c
h
in
e
lea
rn
in
g
(
M
L
)
f
ra
m
e
w
o
r
k
s
p
ro
v
id
e
t
h
e
m
o
st
ro
b
u
st
p
e
rf
o
r
m
a
n
c
e
f
o
r
c
ro
ss
-
m
o
d
a
li
t
y
ima
g
e
e
n
h
a
n
c
e
m
e
n
t,
b
rid
g
i
n
g
th
e
g
a
p
b
e
t
w
e
e
n
ra
w
d
a
ta
a
c
q
u
isit
io
n
a
n
d
c
li
n
ica
l
in
ter
p
re
tatio
n
.
K
ey
w
o
r
d
s
:
C
las
s
ical
i
m
ag
e
e
n
h
an
ce
m
e
n
t
C
o
n
v
o
lu
tio
n
al
n
e
u
r
al
n
e
t
w
o
r
k
E
d
g
e
d
etec
tio
n
I
m
ag
e
e
n
h
a
n
ce
m
e
n
t p
ip
elin
e
Me
d
ical
i
m
a
g
e
en
h
an
ce
m
e
n
t
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Z
a
n
n
at
u
l Fer
d
u
s
h
ie
Dep
ar
t
m
en
t o
f
E
d
u
ca
tio
n
al
T
ec
h
n
o
lo
g
y
a
n
d
E
n
g
in
ee
r
i
n
g
,
U
n
iv
er
s
it
y
o
f
Fro
n
tier
T
ec
h
n
o
lo
g
y
Kaliak
air
,
Gaz
ip
u
r
,
B
an
g
lad
es
h
E
m
ail:
za
n
n
at
u
l0
0
0
1
@
u
f
tb
.
ac
.
b
d
1.
I
NT
RO
D
UCT
I
O
N
Me
d
ical
i
m
a
g
i
n
g
is
in
d
is
p
e
n
s
ab
le
f
o
r
m
o
d
er
n
h
ea
lt
h
ca
r
e,
p
r
o
v
id
in
g
cr
itical
in
s
i
g
h
ts
t
h
r
o
u
g
h
m
o
d
ali
ties
li
k
e
X
-
r
a
y
s
,
co
m
p
u
ted
to
m
o
g
r
ap
h
y
(
C
T
)
,
an
d
m
ag
n
et
ic
r
eso
n
a
n
ce
i
m
ag
i
n
g
(
MRI)
.
Ho
w
e
v
er
,
r
a
w
i
m
a
g
es
f
r
eq
u
e
n
tl
y
s
u
f
f
er
f
r
o
m
lo
w
co
n
tr
ast,
n
o
is
e,
a
n
d
p
o
o
r
ed
g
e
d
ef
in
itio
n
,
w
h
ich
ca
n
o
b
s
cu
r
e
v
ita
l
d
iag
n
o
s
t
ic
d
etails
[
1
]
,
[
2
]
.
E
n
h
a
n
ce
m
en
t
tech
n
iq
u
es
g
e
n
er
all
y
f
all
i
n
to
t
w
o
ca
teg
o
r
ies:
f
r
eq
u
en
c
y
d
o
m
ai
n
m
et
h
o
d
s
,
w
h
ich
m
a
n
ip
u
late
t
h
e
i
m
a
g
e
s
p
ec
tr
u
m
v
ia
Fo
u
r
ie
r
tr
an
s
f
o
r
m
s
,
an
d
s
p
atial
d
o
m
ain
m
et
h
o
d
s
,
w
h
ic
h
o
p
er
ate
d
ir
ec
tly
o
n
p
ix
e
l
v
al
u
e
s
th
r
o
u
g
h
tr
an
s
f
o
r
m
atio
n
s
li
k
e
h
is
to
g
r
a
m
eq
u
aliza
tio
n
(
HE
)
an
d
s
h
ar
p
en
i
n
g
[
2
]
.
T
r
a
d
itio
n
al
tech
n
iq
u
es
s
u
c
h
as
th
e
L
ap
lace
tr
an
s
f
o
r
m
(
L
T
)
,
So
b
el
o
p
er
ato
r
(
SO)
,
an
d
H
E
ar
e
ef
f
ec
tiv
e
f
o
r
s
p
ec
if
ic
tas
k
s
.
L
T
en
h
an
ce
s
h
ig
h
-
f
r
eq
u
e
n
c
y
d
etails,
SO
h
i
g
h
li
g
h
t
s
a
n
ato
m
ical
b
o
u
n
d
ar
ies
,
an
d
HE
i
m
p
r
o
v
es
o
v
er
all
co
n
tr
ast.
Desp
ite
th
ei
r
u
tili
t
y
,
t
h
ese
clas
s
ical
m
et
h
o
d
s
o
f
ten
lack
ad
ap
tab
ilit
y
;
t
h
e
y
t
y
p
icall
y
ap
p
l
y
u
n
i
f
o
r
m
tr
a
n
s
f
o
r
m
atio
n
s
t
h
at
m
a
y
n
o
t
ac
co
u
n
t
f
o
r
th
e
h
i
g
h
v
ar
iab
ilit
y
i
n
p
atien
t
a
n
ato
m
y
o
r
th
e
s
p
ec
if
ic
n
o
is
e
p
r
o
f
iles
o
f
d
if
f
er
en
t i
m
a
g
i
n
g
h
ar
d
w
ar
e.
A
s
i
g
n
i
f
ica
n
t
r
e
s
ea
r
ch
g
ap
ex
i
s
ts
i
n
s
y
s
te
m
atica
ll
y
co
m
p
ar
i
n
g
t
h
ese
cla
s
s
ica
l
o
p
er
ato
r
s
ag
a
in
s
t
d
ata
-
d
r
iv
en
co
n
v
o
l
u
tio
n
a
l
n
eu
r
al
n
et
w
o
r
k
s
(
C
NNs)
ac
r
o
s
s
m
u
lti
-
m
o
d
al
d
atasets
li
k
e
m
u
s
c
u
lo
s
k
eleta
l
r
ad
io
g
r
ap
h
s
(
MU
R
A
)
an
d
H
u
m
an
Me
tap
n
eu
m
o
v
ir
u
s
(
HM
P
V)
[
3
]
,
[
4
]
.
W
h
ile
m
o
s
t
s
t
u
d
ies
f
o
cu
s
o
n
s
i
n
g
le
-
m
o
d
alit
y
o
p
tim
izatio
n
s
,
th
i
s
p
ap
er
p
r
o
v
i
d
es a
cr
o
s
s
-
m
o
d
alit
y
co
m
p
ar
at
iv
e
an
al
y
s
is
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
4
,
A
u
g
u
s
t
20
26
:
1
331
-
1
3
4
1
1332
T
h
e
s
p
ec
if
ic
co
n
tr
ib
u
t
io
n
s
o
f
th
is
s
tu
d
y
ar
e
as:
(
i)
c
o
m
p
a
r
ativ
e
f
r
a
m
e
w
o
r
k
:
a
r
ig
o
r
o
u
s
s
y
s
te
m
a
tic
co
m
p
ar
is
o
n
b
et
w
ee
n
th
r
ee
c
lass
ical
o
p
er
ato
r
s
(
L
T
,
SO,
HE
)
an
d
m
ac
h
i
n
e
lear
n
in
g
(
ML
)
-
b
ased
C
N
N
ar
ch
itect
u
r
es
ac
r
o
s
s
m
u
s
c
u
lo
s
k
eleta
l
an
d
l
u
n
g
C
T
d
atasets
[
5
]
;
(
ii)
p
er
f
o
r
m
a
n
ce
b
en
ch
m
ar
k
in
g
:
q
u
a
n
titat
iv
e
v
alid
atio
n
d
e
m
o
n
s
tr
ati
n
g
th
at
w
h
ile
SO
an
d
HE
ex
ce
l
in
s
p
ec
if
ic
b
o
u
n
d
ar
y
d
etec
tio
n
,
ML
f
r
a
m
e
w
o
r
k
s
p
r
o
v
id
e
s
u
p
er
io
r
d
etail
p
r
eser
v
atio
n
(
en
tr
o
p
y
)
an
d
ad
ap
tiv
e
clar
it
y
[
6
]
;
an
d
(
iii)
m
o
d
alit
y
-
s
p
ec
if
ic
i
n
s
ig
h
t
s
:
i
de
n
ti
f
icatio
n
o
f
o
p
ti
m
al
en
h
a
n
ce
m
e
n
t
s
tr
ateg
ie
s
b
ased
o
n
d
iag
n
o
s
tic
r
eq
u
ir
e
m
e
n
t
s
,
b
r
id
g
in
g
t
h
e
g
ap
b
et
w
ee
n
co
m
p
u
tatio
n
al
s
p
ee
d
an
d
clin
i
ca
l r
o
b
u
s
tn
es
s
.
T
h
e
r
em
ain
d
er
o
f
th
i
s
p
ap
er
i
s
s
tr
u
ct
u
r
ed
as:
s
ec
tio
n
2
r
ev
ie
w
s
r
elate
d
w
o
r
k
;
s
ec
tio
n
3
d
etails
th
e
m
et
h
o
d
o
lo
g
ies;
s
ec
tio
n
4
ex
p
er
i
m
en
tal
s
etu
p
,
r
esu
lt
s
,
an
d
d
is
cu
s
s
io
n
;
an
d
s
ec
tio
n
5
d
is
cu
s
s
es
f
i
n
d
in
g
s
an
d
co
n
clu
d
es t
h
e
s
t
u
d
y
.
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
As
ev
id
e
n
ce
d
in
T
ab
le
1
[
7
]
-
[
1
4
]
,
w
h
ile
r
ec
en
t
ad
v
a
n
ce
m
en
ts
li
k
e
d
if
f
u
s
io
n
m
o
d
els
[
1
3
]
an
d
t
r
an
s
f
o
r
m
er
s
[
1
4
]
o
f
f
er
h
ig
h
p
ea
k
s
i
g
n
al
-
to
-
n
o
i
s
e
r
atio
(
P
SNR
)
v
al
u
es,
t
h
e
y
o
f
te
n
co
m
e
at
th
e
co
s
t
o
f
h
i
g
h
co
m
p
u
tatio
n
al
o
v
er
h
ea
d
.
C
o
n
v
er
s
el
y
,
tr
ad
itio
n
al
m
et
h
o
d
s
[
7
]
,
[
1
0
]
a
r
e
co
m
p
u
tatio
n
all
y
e
f
f
i
cien
t
b
u
t
lack
t
h
e
ad
ap
tab
ilit
y
r
eq
u
ir
ed
f
o
r
co
m
p
lex
m
o
d
ali
ties
li
k
e
MU
R
A
.
Ou
r
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
b
r
id
g
es
th
is
g
ap
b
y
in
te
g
r
atin
g
f
o
u
n
d
atio
n
a
l
s
p
ati
al
o
p
er
ato
r
s
w
ith
a
C
NN
-
b
ase
d
f
r
am
e
w
o
r
k
,
ac
h
ie
v
in
g
a
b
alan
ce
b
et
w
ee
n
h
i
g
h
-
in
te
n
s
it
y
ed
g
e
d
etec
tio
n
a
n
d
in
f
o
r
m
atio
n
p
r
eser
v
atio
n
(
e
n
tr
o
p
y
)
,
a
s
d
etailed
in
th
e
f
o
llo
w
i
n
g
s
ec
tio
n
s
.
T
ab
le
1
.
C
o
m
p
ar
ativ
e
s
u
m
m
ar
y
o
f
ex
is
ti
n
g
liter
at
u
r
e
an
d
p
r
o
p
o
s
ed
w
o
r
k
S
t
u
d
y
M
e
t
h
o
d
(
s)
D
a
t
a
se
t
/
m
o
d
a
l
i
t
y
A
d
v
a
n
t
a
g
e
s
L
i
mi
t
a
t
i
o
n
s
Li
-
p
i
n
g
a
n
d
L
i
a
n
-
q
i
n
g
[
7
]
L
o
c
a
l
l
y
r
e
d
i
st
r
i
b
u
t
e
d
h
i
s
t
o
g
r
a
ms
(
L
H
R
)
G
e
n
e
r
a
l
me
d
i
c
a
l
B
e
t
t
e
r
g
r
a
y
-
l
e
v
e
l
b
a
l
a
n
c
e
;
n
o
i
se
su
p
p
r
e
ssi
o
n
C
l
a
r
i
t
y
i
mp
r
o
v
e
me
n
t
is
n
o
t
si
g
n
i
f
i
c
a
n
t
N
a
g
e
sh
a
a
n
d
K
u
mar
[
8
]
H
o
mo
g
e
n
e
ity
+
HE
M
a
mm
o
g
r
a
ms
U
t
i
l
i
z
e
s
b
o
t
h
g
l
o
b
a
l
a
n
d
l
o
c
a
l
i
mag
e
i
n
f
o
r
mat
i
o
n
;
i
mp
r
o
v
e
s
l
o
c
a
l
c
o
n
t
r
a
s
t
a
n
d
p
r
e
se
r
v
e
s
i
mp
o
r
t
a
n
t
d
e
t
a
i
l
s.
C
o
mp
u
t
a
t
i
o
n
a
l
l
y
mo
r
e
c
o
mp
l
e
x
t
h
a
n
g
l
o
b
a
l
HE
P.
Y
a
n
[
9
]
L
o
c
a
l
a
d
a
p
t
i
v
e
h
i
s
t
o
g
r
a
m
e
q
u
a
l
i
z
a
t
i
o
n
(
L
A
H
E
)
+
w
a
v
e
l
e
t
t
r
a
n
sf
o
r
ms
L
u
n
g
CT
A
d
d
r
e
sse
s
a
r
t
i
f
a
c
t
n
o
i
se
a
n
d
w
i
d
e
d
y
n
a
m
i
c
r
a
n
g
e
R
e
q
u
i
r
e
s
c
o
mp
l
e
x
i
mag
e
se
g
me
n
t
a
t
i
o
n
st
e
p
s
Wu
et
al
.
[
1
0
]
L
a
p
l
a
c
e
+
g
r
a
d
i
e
n
t
+
HE
D
i
g
i
t
a
l
X
-
r
a
y
C
l
e
a
r
e
r
d
e
t
a
i
l
t
h
a
n
t
r
a
d
i
t
i
o
n
a
l
si
n
g
l
e
-
me
t
h
o
d
s
S
u
sce
p
t
i
b
l
e
to
o
v
e
r
-
sh
a
r
p
e
n
i
n
g
a
r
t
i
f
a
c
t
s
Z
h
a
n
g
et
a
l
.
[
1
1
]
M
u
l
t
i
-
sca
l
e
f
u
s
i
o
n
n
e
t
w
o
r
k
M
R
I
1
5
%
i
mp
r
o
v
e
me
n
t
in
l
e
si
o
n
d
e
t
e
c
t
i
o
n
M
o
d
a
l
i
t
y
sp
e
c
i
f
i
c
;
l
e
ss
t
e
st
e
d
on
X
-
r
a
y
/
M
U
R
A
L
a
o
u
sy
et
al
.
[
1
2
]
C
N
N
+
e
d
g
e
d
e
t
e
c
t
i
o
n
L
u
n
g
CT
H
i
g
h
se
g
me
n
t
a
t
i
o
n
a
c
c
u
r
a
c
y
(
d
i
c
e
:
0
.
9
2
)
.
H
i
g
h
c
o
m
p
u
t
a
t
i
o
n
a
l
c
o
s
t
f
o
r
t
r
a
i
n
i
n
g
Ho
et
al
.
[
1
3
]
D
i
f
f
u
si
o
n
mo
d
e
l
(
U
n
i
M
I
E)
CT
a
n
d
M
R
I
S
t
a
t
e
-
of
-
t
h
e
-
a
r
t
P
S
N
R
(
3
4
.
5
d
B
)
S
l
o
w
i
n
f
e
r
e
n
c
e
sp
e
e
d
c
o
mp
a
r
e
d
to
C
N
N
s
W
a
n
g
a
n
d
S
a
mo
n
t
e
[
1
4
]
T
r
a
n
sf
o
r
me
r
s
(
mu
l
t
i
-
sca
l
e
)
L
o
w
-
d
o
s
e
CT
2
0
%
n
o
i
se
r
e
d
u
c
t
i
o
n
;
f
i
n
e
d
e
t
a
i
l
p
r
e
se
r
v
a
t
i
o
n
R
e
q
u
i
r
e
s
l
a
r
g
e
-
sca
l
e
d
a
t
a
se
t
s
f
o
r
a
t
t
e
n
t
i
o
n
t
u
n
i
n
g
P
r
o
p
o
se
d
w
o
r
k
C
N
N
+
LT
+
SO
+
HE
M
U
R
A
a
n
d
H
M
P
V
B
a
l
a
n
c
e
d
e
n
t
r
o
p
y
(
6
.
0
3
)
a
n
d
e
d
g
e
i
n
t
e
n
s
i
t
y
(
3
5
8
2
)
R
e
q
u
i
r
e
s
g
r
a
p
h
i
c
s
p
r
o
c
e
ssi
n
g
u
n
i
t
(
G
P
U
)
f
o
r
o
p
t
i
m
a
l
ML
p
e
r
f
o
r
man
c
e
3.
M
E
T
H
O
DO
L
O
G
Y
I
n
th
i
s
s
ec
tio
n
,
i
m
a
g
e
en
h
a
n
ce
m
en
t
f
r
a
m
e
w
o
r
k
is
d
esi
g
n
ed
as
a
m
o
d
u
lar
,
s
eq
u
en
t
ial
p
ip
elin
e
to
ad
d
r
ess
th
e
li
m
itatio
n
s
o
f
r
aw
m
ed
ical
i
m
ag
in
g
d
ata,
s
u
ch
as
lo
w
co
n
tr
ast
an
d
n
o
is
e.
T
h
e
ar
ch
itectu
r
e,
illu
s
tr
ated
in
F
ig
u
r
e
1
,
in
te
g
r
ates
clas
s
ical
s
p
atial
d
o
m
a
i
n
p
r
o
ce
s
s
in
g
w
it
h
ad
v
an
ce
d
ML
r
ef
in
e
m
e
n
t.
T
h
e
F
ig
u
r
e
1
f
o
llo
w
s
a
s
eq
u
e
n
t
ial
w
o
r
k
f
lo
w
,
b
eg
i
n
n
i
n
g
w
it
h
n
o
r
m
aliza
tio
n
an
d
r
esizi
n
g
to
s
ta
n
d
ar
d
ize
th
e
in
p
u
t
g
r
a
y
s
ca
le
i
m
a
g
es.
T
h
e
co
r
e
m
et
h
o
d
in
te
g
r
ates
HE
f
o
r
g
lo
b
al
co
n
tr
ast
s
tr
etch
in
g
f
o
llo
wed
b
y
a
C
NN
-
b
ased
r
ef
in
e
m
e
n
t
b
lo
ck
th
at
u
tili
ze
s
p
er
ce
p
tu
al
an
d
SS
I
M
lo
s
s
f
u
n
ctio
n
s
to
e
n
s
u
r
e
h
i
g
h
-
f
id
e
lit
y
r
ec
o
n
s
tr
u
ctio
n
.
P
ar
allel
ev
alu
atio
n
p
ath
s
f
o
r
L
ap
lace
,
So
b
el
,
an
d
Un
s
h
ar
p
m
as
k
i
n
g
allo
w
f
o
r
co
m
p
r
eh
e
n
s
iv
e
ab
latio
n
s
tu
d
ie
s
an
d
co
m
p
ar
ativ
e
an
al
y
s
is
.
T
h
e
f
i
n
al
o
u
tp
u
t
d
eli
v
er
s
a
r
o
b
u
s
t
en
h
a
n
ce
m
en
t
ch
ar
ac
ter
ized
b
y
i
m
p
r
o
v
ed
ed
g
e
clar
it
y
,
o
p
ti
m
ized
co
n
tr
ast,
a
n
d
s
ig
n
if
ica
n
t
n
o
is
e
r
ed
u
ctio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
C
o
mp
a
r
a
tive
ev
a
lu
a
tio
n
o
f c
la
s
s
ica
l a
n
d
ma
ch
in
e
le
a
r
n
in
g
m
eth
o
d
s
fo
r
med
ica
l
…
(
Md
.
Meh
ed
i H
a
s
an
)
1333
Fig
u
r
e
1
.
Sch
e
m
atic
ar
ch
itec
tu
r
e
o
f
th
e
p
r
o
p
o
s
ed
h
y
b
r
id
m
ed
ical
i
m
a
g
e
en
h
a
n
ce
m
e
n
t
3
.
1
.
Dis
cr
et
e
LT
f
o
r
edg
e
en
ha
nce
m
ent
T
h
e
L
ap
lacia
n
o
p
er
ato
r
is
a
s
ec
o
n
d
-
o
r
d
er
d
if
f
er
en
tial
f
ilter
co
m
m
o
n
l
y
u
s
ed
i
n
i
m
a
g
e
p
r
o
ce
s
s
i
n
g
to
h
ig
h
li
g
h
t
r
eg
io
n
s
o
f
r
ap
id
in
te
n
s
it
y
c
h
an
g
e,
t
h
er
eb
y
s
h
ar
p
en
in
g
ed
g
es
a
n
d
f
in
e
d
etail
s
[
1
4
]
,
[
1
5
]
.
I
n
m
ed
ical
i
m
a
g
in
g
,
it i
s
p
ar
ticu
lar
l
y
u
s
e
f
u
l f
o
r
e
m
p
h
asiz
in
g
an
a
to
m
ical
b
o
u
n
d
ar
ies in
m
o
d
ali
ties
s
u
c
h
as C
T
an
d
X
-
r
a
y
.
T
h
e
2
D
L
ap
lacia
n
o
p
er
ato
r
is
d
ef
in
ed
as [
1
6
]
.
∇
2
(
,
)
=
2
(
,
)
2
+
2
(
,
)
2
(
1
)
w
h
er
e
(
,
)
i
s
t
h
e
p
ix
el
i
n
ten
s
it
y
at
co
-
o
r
d
in
ates
(
,
)
.
T
h
e
L
ap
lacia
n
o
p
er
ato
r
(
1
)
,
a
s
ec
o
n
d
-
o
r
d
er
d
if
f
er
en
tia
l
o
p
er
ato
r
w
id
el
y
u
s
ed
in
i
m
a
g
e
p
r
o
ce
s
s
in
g
f
o
r
ed
g
e
d
etec
tio
n
an
d
f
ea
t
u
r
e
en
h
a
n
ce
m
en
t
[
1
7
]
.
3
.
1
.
1
.
L
a
pla
cia
n
t
e
m
pla
t
es
T
h
e
L
ap
lacia
n
te
m
p
lates
ca
n
b
e
r
ep
r
esen
ted
as
co
n
v
o
l
u
tio
n
k
er
n
els
(
f
ilter
s
)
ap
p
lied
to
th
e
i
m
ag
e
.
T
h
ese
te
m
p
lates d
ef
in
e
h
o
w
n
eig
h
b
o
u
r
i
n
g
p
ix
el
v
al
u
es c
o
n
tr
ib
u
te
to
th
e
ca
lc
u
latio
n
o
f
th
e
L
ap
lacia
n
.
−
B
asic L
ap
lacia
n
k
er
n
e
l [
1
8
]
-
[
2
0
]
:
(
0
1
0
1
−
4
1
0
1
0
)
−
Mo
d
if
ied
k
er
n
el
w
i
th
d
ia
g
o
n
al
s
[
2
0
]
:
(
1
1
1
1
−
8
1
1
1
1
)
−
Oth
er
v
ar
iatio
n
s
(
e.
g
.
,
r
o
tatio
n
o
r
f
lip
p
ed
v
er
s
io
n
s
)
[
2
1
]
,
[
2
2
]
:
(
0
−
1
0
−
1
4
−
1
0
−
1
0
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
4
,
A
u
g
u
s
t
20
26
:
1
331
-
1
3
4
1
1334
(
−
1
−
1
−
1
−
1
8
−
1
−
1
−
1
−
1
)
T
h
ese
te
m
p
lates
d
i
f
f
er
s
li
g
h
tl
y
i
n
h
o
w
t
h
e
y
w
ei
g
h
t
d
iag
o
n
al
co
n
tr
ib
u
tio
n
s
o
r
ce
n
tr
e
p
i
x
el
in
te
n
s
i
t
y
,
b
u
t
all
ef
f
ec
tiv
e
l
y
p
er
f
o
r
m
th
e
L
ap
lac
ian
o
p
er
atio
n
.
3
.
2
.
G
ra
dient
-
ba
s
ed
edg
e
det
ec
t
i
o
n
SO
I
n
t
h
is
s
ec
tio
n
w
e
i
n
tr
o
d
u
ce
th
e
SO
m
et
h
o
d
f
o
r
ed
g
e
d
etec
tio
n
.
T
h
e
SO
co
n
s
i
s
ts
o
f
t
w
o
3
×3
co
n
v
o
lu
tio
n
k
er
n
el
s
o
n
e
f
o
r
d
etec
tin
g
tr
an
s
v
er
s
e
(
v
er
tical)
ed
g
es
an
d
o
n
e
f
o
r
lo
n
g
it
u
d
i
n
al
(
h
o
r
izo
n
tal)
ed
g
es
.
T
h
ese
k
er
n
els
ar
e
co
n
v
o
lv
ed
w
it
h
t
h
e
i
m
a
g
e
to
co
m
p
u
te
g
r
ad
ien
t
ap
p
r
o
x
i
m
atio
n
s
in
t
h
e
x
-
a
n
d
y
-
d
ir
ec
tio
n
s
.
L
et
r
ep
r
esen
t t
h
e
o
r
ig
i
n
al
g
r
a
y
s
ca
le
i
m
a
g
e.
T
h
e
g
r
ad
ien
t r
es
p
o
n
s
es a
r
e
co
m
p
u
ted
as [
1
8
]
:
(
)
=
(
−
1
0
+
1
−
2
0
+
2
−
1
0
+
1
)
×
(
)
=
(
+
1
+
2
+
1
0
0
0
−
1
−
2
+
1
)
×
T
h
e
g
r
ad
ien
t
m
a
g
n
itu
d
e
=
√
2
+
2
…
is
u
s
ed
to
p
r
o
d
u
ce
th
e
f
i
n
al
e
d
g
e
-
e
n
h
a
n
ce
d
i
m
ag
e,
w
h
ic
h
i
s
cr
itical
f
o
r
id
en
ti
f
y
in
g
s
tr
u
ct
u
r
al
f
ea
t
u
r
es in
MU
R
A
a
n
d
C
T
d
atasets
[
1
9
]
-
[
2
3
]
.
3
.
3
.
HE
f
o
r
co
ntr
a
s
t
enha
nc
e
m
e
nt
T
o
a
d
d
r
ess
lo
w
co
n
tr
ast,
HE
r
ed
is
tr
ib
u
tes
p
i
x
el
i
n
ten
s
itie
s
to
ac
h
iev
e
a
n
ea
r
-
u
n
i
f
o
r
m
p
r
o
b
ab
ilit
y
d
is
tr
ib
u
t
io
n
.
Fo
r
an
8
-
b
it
i
m
a
g
e
w
ith
=
256
in
ten
s
it
y
le
v
els,
th
e
m
ap
p
in
g
f
u
n
ct
io
n
(
)
is
d
er
iv
ed
f
r
o
m
th
e
cu
m
u
lati
v
e
d
is
tr
ib
u
tio
n
f
u
n
ctio
n
(
C
DF)
[
2
4
]
,
[
2
5
]
:
=
(
)
=
(
−
1
)
∑
=
0
(
)
(
2
)
w
h
er
e
(
)
is
th
e
p
r
o
b
ab
ilit
y
o
f
i
n
ten
s
i
t
y
le
v
el
.
T
h
is
en
s
u
r
e
s
th
at
th
e
d
y
n
a
m
ic
r
a
n
g
e
o
f
t
h
e
i
m
ag
e
is
f
u
ll
y
u
tili
ze
d
,
m
a
k
i
n
g
s
u
b
tle
p
ath
o
l
o
g
ical
d
etails
m
o
r
e
v
is
ib
le
[
2
6
]
.
3
.
4
.
ML
m
o
del f
o
r
m
edica
l i
m
a
g
e
enha
nce
m
ent
T
h
e
lim
ita
tio
n
s
o
f
class
ical
i
m
ag
e
p
r
o
ce
s
s
i
n
g
tec
h
n
iq
u
es
i
n
ad
ap
tin
g
to
d
iv
er
s
e
m
ed
ica
l
i
m
ag
i
n
g
m
o
d
ali
ties
a
n
d
p
atien
t
-
s
p
ec
i
f
i
c
v
ar
iatio
n
s
n
ec
e
s
s
ita
te
ad
v
a
n
ce
d
,
d
ata
-
d
r
iv
e
n
ap
p
r
o
ac
h
es.
ML
,
p
ar
ticu
lar
l
y
C
NNs,
o
f
f
er
s
a
p
o
w
er
f
u
l
s
o
lu
tio
n
b
y
lear
n
in
g
o
p
ti
m
al
en
h
an
ce
m
e
n
t
p
ar
a
m
eter
s
f
r
o
m
lar
g
e
d
ata
s
ets,
co
m
p
le
m
e
n
ti
n
g
th
e
ed
g
e
a
n
d
co
n
tr
ast
i
m
p
r
o
v
e
m
en
t
s
ac
h
iev
ed
b
y
LT
,
SO
,
a
n
d
HE
.
T
h
is
s
ec
tio
n
d
etail
s
t
h
e
in
te
g
r
atio
n
o
f
a
C
NN
i
n
to
th
e
f
r
a
m
e
w
o
r
k
,
tailo
r
ed
to
en
h
an
c
e
MU
R
A
s
k
e
letal
an
d
th
e
ca
n
ce
r
im
a
g
i
n
g
ar
ch
iv
e
(
T
C
I
A
)
lu
n
g
i
m
ag
e
s
,
w
it
h
a
f
o
cu
s
o
n
i
m
p
r
o
v
i
n
g
d
ia
g
n
o
s
tic
p
r
ec
is
io
n
[
2
7
]
,
[
2
8
]
.
3
.
4
.
1
.
C
NN
-
ba
s
ed
i
m
a
g
e
enh
a
nce
m
e
nt
A
C
NN
w
a
s
e
m
p
lo
y
ed
to
en
h
an
ce
C
T
s
ca
n
i
m
a
g
es
b
y
lear
n
in
g
a
n
o
n
-
li
n
ea
r
m
ap
p
in
g
ℱ
:
→
b
et
w
ee
n
th
e
i
n
p
u
t lo
w
-
q
u
a
lit
y
i
m
a
g
e
an
d
th
e
e
n
h
a
n
ce
d
o
u
tp
u
t
.
T
h
e
n
et
w
o
r
k
co
n
s
is
t
s
o
f
t
h
r
ee
s
tag
es:
−
F
eat
u
r
e
ex
tr
ac
tio
n
u
s
i
n
g
co
n
v
o
lu
tio
n
,
=
∑
,
,
∙
+
,
+
,
,
,
(
3
)
−
N
on
-
l
in
ea
r
tr
an
s
f
o
r
m
atio
n
w
i
t
h
r
ec
tif
ied
li
n
ea
r
u
n
it
(
R
e
L
U
)
(
)
=
ma
x
(
0
,
)
,
an
d
−
R
ec
o
n
s
tr
u
ctio
n
v
ia
f
i
n
al
co
n
v
o
lu
tio
n
al
la
y
er
s
.
I
n
th
e
ab
s
en
ce
o
f
p
air
ed
g
r
o
u
n
d
tr
u
th
i
m
a
g
es,
e
n
h
an
ce
m
e
n
t q
u
alit
y
w
a
s
g
u
id
ed
b
y
p
er
ce
p
tu
al
m
etr
ic
s
(
e.
g
.
,
s
tr
u
ct
u
r
al
s
i
m
ilar
it
y
in
d
ex
m
ea
s
u
r
e
(
SS
I
M
)
an
d
n
atu
r
al
i
m
a
g
e
q
u
al
it
y
e
v
al
u
at
o
r
(
NI
QE
)
)
as
th
e
o
p
tim
izatio
n
o
b
j
ec
tiv
e.
T
h
e
en
h
a
n
ce
m
en
t
p
r
o
ce
s
s
f
o
llo
w
s
:
n
o
r
m
alize
,
ex
tr
ac
t
f
ea
t
u
r
es
,
ap
p
ly
m
u
ltip
l
e
co
n
v
o
lu
tio
n
–
R
e
L
U
la
y
er
s
,
r
e
co
n
s
tr
u
ct
,
an
d
p
o
s
t
-
p
r
o
ce
s
s
to
m
a
in
ta
in
v
alid
in
ten
s
it
y
r
an
g
e
s
w
h
ich
i
s
illu
s
tr
ated
in
Fig
u
r
e
2
.
T
h
is
a
p
p
r
o
ac
h
im
p
r
o
v
es
s
h
ar
p
n
e
s
s
,
co
n
tr
ast,
an
d
ed
g
e
d
etail
w
h
il
e
r
ed
u
cin
g
n
o
is
e
in
m
ed
ic
al
i
m
ag
er
y
[
2
9
]
,
[
3
0
]
.
T
h
e
Fig
u
r
e
3
s
h
o
w
s
n
et
w
o
r
k
e
m
p
lo
y
s
3
×3
co
n
v
o
l
u
tio
n
a
l
k
er
n
el
s
th
r
o
u
g
h
o
u
t
t
h
e
ar
ch
i
t
ec
tu
r
e
to
ef
f
ec
tiv
e
l
y
ca
p
t
u
r
e
lo
ca
l
s
p
ati
al
f
ea
t
u
r
es
w
h
ile
p
r
eser
v
in
g
r
eso
lu
tio
n
.
R
e
L
U
ac
ti
v
atio
n
f
u
n
ctio
n
s
ar
e
ap
p
lied
af
ter
ea
ch
co
n
v
o
lu
tio
n
la
y
er
to
in
tr
o
d
u
ce
n
o
n
-
li
n
ea
r
it
y
an
d
e
n
h
a
n
ce
f
ea
tu
r
e
lear
n
in
g
ca
p
ab
ilit
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
C
o
mp
a
r
a
tive
ev
a
lu
a
tio
n
o
f c
la
s
s
ica
l a
n
d
ma
ch
in
e
le
a
r
n
in
g
m
eth
o
d
s
fo
r
med
ica
l
…
(
Md
.
Meh
ed
i H
a
s
an
)
1335
Fig
u
r
e
2
.
B
lo
ck
d
iag
r
a
m
o
f
t
h
e
p
r
o
p
o
s
ed
C
NN
-
b
ase
d
C
T
i
m
ag
e
en
h
an
ce
m
e
n
t a
p
p
r
o
ac
h
Fig
u
r
e
3
.
C
NN
ar
ch
itect
u
r
e
f
o
r
i
m
ag
e
e
n
h
a
n
ce
m
e
n
t
4.
E
x
peri
m
e
nta
l
s
et
up
,
re
s
ult
s
,
a
nd
dis
cu
s
s
io
n
4
.
1
.
H
a
rdwa
re
a
nd
s
o
f
t
w
a
re
env
iro
n
m
e
nt
C
las
s
ical
i
m
a
g
e
p
r
o
ce
s
s
i
n
g
was
i
m
p
le
m
e
n
ted
i
n
M
A
T
L
A
B
R
2
0
2
3
b
,
w
h
ile
th
e
d
ee
p
lear
n
in
g
(
D
L
)
f
r
a
m
e
w
o
r
k
w
as
d
ev
elo
p
ed
u
s
i
n
g
P
y
T
o
r
ch
2
.
1
.
E
x
p
er
i
m
en
ts
w
er
e
ex
ec
u
ted
o
n
a
h
ig
h
-
p
er
f
o
r
m
a
n
ce
w
o
r
k
s
tatio
n
eq
u
ip
p
ed
w
it
h
a
n
I
n
tel
C
o
r
e
i9
-
1
3
9
0
0
K
C
P
U,
an
NVI
DI
A
R
T
X
4
0
9
0
GP
U
(
2
4
GB
VR
A
M)
,
a
n
d
1
2
8
GB
o
f
DDR5
R
A
M,
r
u
n
n
i
n
g
o
n
Ub
u
n
tu
2
2
.
0
4
L
T
S.
4
.
1
.
1
.
I
m
ple
m
ent
a
t
io
n
d
et
a
il
s
C
las
s
ical
m
et
h
o
d
s
:
L
T
u
s
ed
a
3
×3
k
er
n
el
[
−1
,
−1
,
−1
;
−1
,
8
,
−1
;
−1
,
−1
,
−1
]
.
E
d
g
e
d
etec
tio
n
u
t
ilized
s
tan
d
ar
d
SO
,
an
d
co
n
tr
ast
en
h
an
ce
m
e
n
t
w
a
s
p
er
f
o
r
m
ed
v
ia
a
d
ap
tiv
e
HE
(
clip
lim
it:
2
.
0
)
.
C
NN
a
r
ch
itec
tu
r
e:
a
r
esid
u
al
en
co
d
er
-
d
ec
o
d
er
-
C
N
N
(
R
E
D
-
C
NN)
ar
ch
itect
u
r
e
w
as
e
m
p
lo
y
ed
,
co
n
s
is
t
in
g
o
f
f
i
v
e
co
n
v
o
lu
tio
n
al
an
d
f
i
v
e
d
ec
o
n
v
o
l
u
tio
n
al
la
y
er
s
with
3
×3
k
er
n
els
a
n
d
c
h
an
n
el
d
ep
th
s
r
an
g
i
n
g
f
r
o
m
6
4
to
1
2
8
.
R
esid
u
al
s
h
o
r
tcu
t
s
w
er
e
i
n
teg
r
ated
to
f
ac
ilit
ate
g
r
ad
ien
t
f
lo
w
.
T
r
ain
in
g
p
r
o
to
co
l:
t
h
e
m
o
d
el
w
as
o
p
ti
m
ize
d
u
s
i
n
g
t
h
e
A
d
a
m
o
p
tim
izer
(
lear
n
i
n
g
r
at
e:
1
×
10
−
4
)
f
o
r
1
8
0
ep
o
ch
s
.
Data
w
as
s
p
lit
i
n
to
tr
ain
i
n
g
/
v
ali
d
atio
n
/
test
i
n
g
s
et
s
(
MU
R
A
:
8
0
/1
0
/1
0
;
T
C
I
A
:
7
0
/1
5
/1
5
)
[
1
6
]
.
B
atch
s
ize
an
d
ep
o
ch
s
:
a
b
atch
s
ize
o
f
1
6
w
as
u
s
ed
o
v
er
1
0
0
ep
o
ch
s
to
b
alan
ce
co
n
v
er
g
e
n
ce
s
p
ee
d
an
d
m
e
m
o
r
y
ef
f
ici
en
c
y
.
L
o
s
s
f
u
n
ctio
n
:
Si
n
ce
p
ai
r
ed
g
r
o
u
n
d
tr
u
t
h
i
m
a
g
es
w
er
e
u
n
a
v
ailab
le,
a
co
m
p
o
s
ite
lo
s
s
f
u
n
ct
io
n
w
as
u
til
i
ze
d
:
=
1
+
2
(
1
−
+
3
(
4
)
4
.
2
.
E
v
a
lua
t
i
o
n
m
et
rics a
nd
s
t
a
t
is
t
ica
l a
na
ly
s
is
Six
q
u
an
t
itati
v
e
m
etr
ics
wer
e
e
m
p
lo
y
ed
to
ev
al
u
ate
en
h
a
n
ce
m
en
t
p
er
f
o
r
m
a
n
ce
:
s
h
ar
p
n
e
s
s
,
in
f
o
r
m
atio
n
en
tr
o
p
y
,
r
o
o
t
m
ea
n
s
q
u
ar
e
(
R
MS)
co
n
tr
ast,
m
ea
n
g
r
ad
ien
t,
m
ea
n
b
r
ig
h
tn
e
s
s
,
a
n
d
ed
g
e
i
n
te
n
s
it
y
.
Du
e
to
t
h
e
u
n
p
air
ed
n
at
u
r
e
o
f
th
e
m
ed
ical
d
ataset
s
,
e
m
p
h
asi
s
w
as
p
lace
d
o
n
t
h
e
r
el
ativ
e
i
m
p
r
o
v
e
m
en
t
b
et
w
ee
n
r
a
w
an
d
p
r
o
ce
s
s
ed
s
tates.
Stati
s
tical
s
i
g
n
i
f
ican
ce
w
as
v
al
id
ated
u
s
i
n
g
t
h
e
n
o
n
-
p
ar
a
m
etr
ic
p
air
ed
W
ilco
x
o
n
s
ig
n
ed
-
r
an
k
test
(
=
0
.
05
)
ac
r
o
s
s
1
0
0
r
an
d
o
m
l
y
s
e
lect
ed
s
am
p
le
s
p
er
d
ataset
to
e
n
s
u
r
e
th
e
r
o
b
u
s
tn
es
s
o
f
t
h
e
r
ep
o
r
ted
m
et
r
ic
g
ain
s
[
1
7
]
.
4
.
3
.
Q
ua
ntit
a
t
iv
e
perf
o
r
m
a
n
ce
co
m
pa
riso
n
Fig
u
r
es
4
to
1
4
p
r
esen
t
v
is
u
al
co
m
p
ar
is
o
n
s
(
b
ef
o
r
e/af
ter
)
f
o
r
in
d
iv
id
u
al
a
n
d
co
m
b
i
n
ed
m
eth
o
d
s
o
n
r
ep
r
esen
tativ
e
l
u
n
g
C
T
(
HM
PV
-
r
elate
d
p
atter
n
s
)
an
d
MU
R
A
s
k
eleta
l
i
m
a
g
es.
LT
(
Fig
u
r
e
4
)
:
e
n
h
an
ce
s
h
i
g
h
-
f
r
eq
u
en
c
y
ed
g
e
s
b
u
t
p
r
o
d
u
ce
s
a
h
i
g
h
-
p
as
s
f
i
lter
ed
ap
p
ea
r
a
n
ce
(
b
lack
b
ac
k
g
r
o
u
n
d
,
w
h
ite
ed
g
e
s
)
.
Mu
ltip
licat
io
n
/
u
n
s
h
ar
p
m
a
s
k
i
n
g
(
Fi
g
u
r
e
5
)
:
s
h
ar
p
en
s
s
tr
u
ct
u
r
al
o
u
t
lin
e
s
;
u
s
ef
u
l
f
o
r
v
ir
u
s
m
e
m
b
r
an
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
4
,
A
u
g
u
s
t
20
26
:
1
331
-
1
3
4
1
1336
b
o
u
n
d
ar
ies
o
r
b
o
n
e
tr
ab
ec
u
lae.
SO
(
Fig
u
r
e
6
)
:
p
r
o
d
u
ce
s
co
h
er
en
t
g
r
ad
ien
t
m
ap
s
,
i
m
p
r
o
v
in
g
b
o
u
n
d
ar
y
d
ef
in
i
tio
n
w
it
h
n
o
is
e
s
u
p
p
r
es
s
io
n
.
Un
s
h
ar
p
m
ea
n
-
f
ilter
ed
(
Fig
u
r
e
7
)
:
u
n
s
h
ar
p
m
ea
n
f
ilt
er
in
g
s
h
ar
p
en
s
th
e
i
m
a
g
e
b
y
s
u
b
tr
ac
ti
n
g
a
b
lu
r
r
ed
v
er
s
io
n
(
o
b
tain
ed
u
s
in
g
a
5
×5
av
er
ag
e
f
ilter
)
f
r
o
m
t
h
e
o
r
ig
in
al
i
m
ag
e
to
en
h
a
n
ce
ed
g
e
d
etail
s
.
T
h
e
en
h
an
ce
d
ed
g
es
ar
e
th
e
n
ad
d
ed
b
ac
k
to
i
m
p
r
o
v
e
co
n
tr
a
s
t
an
d
o
v
er
all
v
is
u
al
clar
it
y
.
HE
(
Fig
u
r
e
8
)
:
d
r
a
m
atica
ll
y
i
m
p
r
o
v
es
g
lo
b
al
co
n
tr
a
s
t;
f
i
n
e
v
asc
u
lar
/le
s
io
n
d
etail
s
b
ec
o
m
e
m
o
r
e
v
is
ib
le.
Fig
u
r
e
9
s
h
o
w
s
th
e
ef
f
ec
ts
o
f
v
ar
io
u
s
i
m
ag
e
e
n
h
a
n
ce
m
en
t
tech
n
iq
u
es
o
n
a
lo
w
-
q
u
ali
t
y
C
T
s
ca
n
.
E
ac
h
m
et
h
o
d
L
ap
lace
f
ilter
i
n
g
,
m
u
ltip
lica
ti
o
n
,
So
b
el
ed
g
e
d
etec
tio
n
,
s
h
ar
p
en
in
g
,
an
d
HE
en
h
a
n
ce
s
ed
g
es,
co
n
tr
a
s
t,
o
r
s
tr
u
ct
u
r
al
d
etails
to
i
m
p
r
o
v
e
v
is
u
al
clar
it
y
a
n
d
d
iag
n
o
s
tic
q
u
ali
t
y
.
Fig
u
r
es
1
0
an
d
1
1
illu
s
tr
ate
t
h
e
tr
an
s
f
o
r
m
atio
n
o
f
b
o
n
e
i
m
a
g
es
f
r
o
m
t
h
e
MU
R
A
d
atas
e
t
f
o
llo
w
i
n
g
th
e
ap
p
licatio
n
o
f
th
e
in
te
g
r
ated
en
h
a
n
ce
m
en
t
p
ip
elin
e,
w
i
th
d
etailed
an
al
y
s
is
to
f
o
llo
w
.
C
NN
-
b
ased
i
m
a
g
e
en
h
a
n
ce
m
en
t
(
Fig
u
r
e
1
2
to
1
4
)
:
s
eq
u
en
tial
ap
p
licatio
n
(
L
ap
la
ce
→
So
b
el
→
u
n
s
h
ar
p
→
h
is
to
g
r
a
m
→
C
NN)
w
h
er
e
C
NN
-
b
a
s
ed
y
ield
s
b
alan
ce
d
en
h
a
n
ce
m
en
t s
h
ar
p
er
ed
g
es,
b
etter
co
n
tr
ast,
r
ed
u
ce
d
ar
tif
ac
ts
.
I
n
th
is
s
ec
tio
n
w
e
i
n
tr
o
d
u
ce
th
e
p
r
o
p
o
s
ed
ML
C
NN
-
b
ased
i
m
a
g
e
en
h
an
ce
m
e
n
t
m
et
h
o
d
.
Fig
u
r
e
4
.
C
o
m
p
ar
is
o
n
o
f
o
r
i
g
i
n
al
an
d
LT
C
T
s
ca
n
i
m
a
g
e
F
ig
u
r
e
5
.
C
o
m
p
ar
is
o
n
o
f
o
r
i
g
i
n
al
an
d
L
E
C
T
s
ca
n
o
f
HM
P
V
v
ir
u
s
Fig
u
r
e
6
.
C
o
m
p
ar
is
o
n
o
f
t
h
e
o
r
ig
in
a
l i
m
a
g
e
w
it
h
t
h
e
So
b
el
g
r
ad
ien
t (
w
ith
ad
d
ed
i
m
ag
e)
tr
an
s
f
o
r
m
ed
i
m
ag
e
Fig
u
r
e
7
.
C
o
m
p
ar
is
o
n
o
f
o
r
i
g
i
n
al
an
d
U
n
s
h
ar
p
m
ea
n
-
f
ilter
ed
C
T
s
ca
n
i
m
a
g
e
Fig
u
r
e
8
.
Fin
al
C
T
s
ca
n
i
m
a
g
e
af
ter
ad
ap
tiv
e
HE
Fig
u
r
e
9
.
P
r
o
ce
s
s
ed
lo
w
-
q
u
alit
y
C
T
s
ca
n
u
s
i
n
g
v
ar
io
u
s
i
m
a
g
e
en
h
a
n
ce
m
en
t te
c
h
n
iq
u
e
s
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
C
o
mp
a
r
a
tive
ev
a
lu
a
tio
n
o
f c
la
s
s
ica
l a
n
d
ma
ch
in
e
le
a
r
n
in
g
m
eth
o
d
s
fo
r
med
ica
l
…
(
Md
.
Meh
ed
i H
a
s
an
)
1337
Fig
u
r
e
1
0
.
(
On
th
e
r
ig
h
t)
p
r
o
ce
s
s
ed
i
m
a
g
e
o
f
t
h
e
s
k
eleto
n
o
f
th
e
r
ig
h
t lo
w
er
li
m
b
Fig
u
r
e
1
1
.
(
On
th
e
lef
t)
p
r
o
ce
s
s
ed
i
m
a
g
e
o
f
th
e
s
k
e
leto
n
o
f
t
h
e
r
ig
h
t
u
p
p
er
li
m
b
Fig
u
r
e
1
2
.
E
x
p
er
im
e
n
tal
r
es
u
lt
s
o
f
th
e
p
r
o
p
o
s
ed
C
NN
-
b
ased
en
h
a
n
ce
m
e
n
t o
n
a
l
u
n
g
C
T
s
ca
n
Fig
u
r
e
1
3
.
Or
ig
in
al
a
n
d
C
NN
-
en
h
a
n
ce
d
r
ig
h
t lo
w
er
u
p
p
er
li
m
b
s
k
ele
tal
r
ad
io
g
r
ap
h
f
r
o
m
MU
R
A
d
ataset
Fig
u
r
e
1
4
.
Or
ig
in
al
a
n
d
C
NN
-
en
h
a
n
ce
d
r
ig
h
t u
p
p
er
li
m
b
s
k
e
letal
r
ad
io
g
r
ap
h
f
r
o
m
MU
R
A
d
ataset
4
.
4
.
Q
ua
ntit
a
t
iv
e
r
esu
lt
s
T
ab
les
2
to
4
co
m
p
ar
e
b
ef
o
r
e/af
ter
m
etr
ics
f
o
r
th
r
ee
r
ep
r
esen
tati
v
e
i
m
ag
e
s
(
o
n
e
lu
n
g
C
T
,
tw
o
MU
R
A
s
k
eleta
l
v
ie
w
s
)
.
I
m
p
r
o
v
e
m
e
n
ts
ar
e
h
ig
h
li
g
h
ted
in
b
o
ld
w
h
er
e
>
10%
o
r
s
tatis
tical
l
y
s
i
g
n
i
f
ica
n
t
(
<
0
.
05
,
W
ilco
x
o
n
s
ig
n
ed
-
r
an
k
tes
t)
.
T
a
b
le
2
s
h
o
w
s
th
a
t
HE
an
d
C
NN
-
ba
s
ed
m
et
h
o
d
s
s
h
o
w
ed
th
e
m
o
s
t
co
n
s
is
ten
t
s
i
g
n
i
f
ican
t
i
m
p
r
o
v
e
m
en
ts
ac
r
o
s
s
s
h
ar
p
n
e
s
s
,
g
r
ad
ien
t,
an
d
ed
g
e
in
ten
s
it
y
(
<
0
.
0
1
in
m
o
s
t
ca
s
es).
So
b
el
ex
ce
lled
in
ed
g
e
m
etr
ics
b
u
t
r
ed
u
ce
d
en
tr
o
p
y
/b
r
i
g
h
tn
e
s
s
.
L
ap
lace
o
f
ten
a
m
p
li
f
ied
n
o
is
e
(
n
eg
ati
v
e
en
tr
o
p
y
ch
a
n
g
e,
<
0
.
0
5
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
4
,
A
u
g
u
s
t
20
26
:
1
331
-
1
3
4
1
1338
T
ab
les
3
an
d
4
d
em
o
n
s
tr
at
e
th
at
HE
an
d
So
b
el
ar
e
h
ig
h
l
y
ef
f
ec
ti
v
e
f
o
r
co
n
tr
ast
an
d
ed
g
e
en
h
a
n
ce
m
en
t
in
s
k
eleta
l
r
ad
io
g
r
ap
h
s
,
w
h
i
le
th
e
C
NN
-
b
a
s
ed
ap
p
r
o
ac
h
d
eliv
er
s
th
e
m
o
s
t
b
alan
ce
d
an
d
s
tatis
t
icall
y
s
i
g
n
if
ica
n
t i
m
p
r
o
v
e
m
en
ts
ac
r
o
s
s
m
u
ltip
le
q
u
a
lit
y
m
etr
ics.
T
ab
le
2
.
L
u
n
g
C
T
i
m
a
g
e
(
HM
P
V
-
r
elate
d
)
M
e
t
h
o
d
S
h
a
r
p
n
e
ss (Δ
)
En
t
r
o
p
y
(
Δ)
R
M
S
c
o
n
t
r
a
s
t
(
Δ)
M
e
a
n
g
r
a
d
i
e
n
t
(
Δ)
B
r
i
g
h
t
n
e
ss
(
Δ)
Ed
g
e
i
n
t
e
n
si
t
y
(
Δ)
W
i
l
c
o
x
o
n
p
-
v
a
l
u
e
L
a
p
l
a
c
e
5
9
8
.
5
8
→
6
5
8
.
9
3
(
+
1
0
%)
5
.
8
3
→
3
.
3
9
0
.
3
5
2
2
→
0
.
0
6
3
0
.
1
9
0
3
→
0
.
1
0
6
4
0
.
6
2
9
8
→
0
.
0
2
2
5
2
5
7
4
3
→
5
7
2
1
<
0
.
0
1
M
u
l
t
i
p
l
i
e
d
5
9
8
.
5
8
→
7
2
2
.
3
9
(
+
2
1
%)
5
.
8
3
→
5
.
6
9
0
.
3
5
2
2
→
0
.
3
7
2
(
+
6
%)
0
.
1
9
0
3
→
0
.
2
1
5
8
0
.
6
2
9
8
→
0
.
5
2
0
7
2
5
7
4
3
→
2
5
4
0
5
=
0
.
0
3
S
o
b
e
l
5
9
8
.
5
8
→
1
1
9
5
.
0
5
(
+
1
0
0
%)
5
.
8
3
→
4
.
3
2
0
.
3
5
2
2
→
0
.
1
0
9
0
.
1
9
0
3
→
0
.
2
1
2
3
0
.
6
2
9
8
→
0
.
0
4
7
6
2
5
7
4
3
→
3
2
1
4
0
(
+
2
5
%)
<
0
.
0
0
1
S
h
a
r
p
e
n
e
d
5
9
8
.
5
8
→
8
9
8
.
6
1
(
+
5
0
%)
5
.
8
3
→
5
.
8
6
0
.
3
5
2
2
→
0
.
3
5
9
0
.
1
9
0
3
→
0.
2
4
5
0
.
6
2
9
8
→
0
.
6
3
0
3
2
5
7
4
3
→
3
1
4
2
9
<
0
.
0
1
H
i
st
o
g
r
a
m
5
9
8
.
5
8
→
1
7
7
4
.
9
(
+
1
9
6
%)
5
.
8
3
→
5
.
5
8
0
.
3
5
2
2
→
0
.
3
3
4
0
.
1
9
0
3
→
0
.
2
9
7
9
(
+
5
7
%)
0
.
6
2
9
8
→
0
.
4
4
7
2
5
7
4
3
→
5
5
2
5
8
(
+
1
1
5
%)
<
0
.
0
0
1
M
L
(
C
N
N
-
b
a
se
d
)
5
9
8
.
5
8
→
1
3
7
5
.
6
6
(
+
1
3
0
%)
5
.
8
3
→
6
.
0
4
(
+
3
.
6
%)
0
.
3
5
2
2
→
0
.
3
7
7
0
.
1
903
→
0
.
3
6
5
6
(
+
9
2
%)
0
.
6
2
9
8
→
0
.
6
1
9
3
2
5
7
4
3
→
4
1
0
6
4
(
+
6
0
%)
<
0
.
0
0
1
T
ab
le
3
.
R
ig
h
t
lo
w
er
li
m
b
s
k
el
etal
i
m
a
g
e
(
MU
R
A
)
M
e
t
h
o
d
S
h
a
r
p
n
e
ss (Δ
)
En
t
r
o
p
y
(
Δ)
R
M
S
c
o
n
t
r
a
st
(
Δ)
M
e
a
n
g
r
a
d
i
e
n
t
(
Δ)
B
r
i
g
h
t
n
e
ss (Δ
)
Ed
g
e
i
n
t
e
n
si
t
y
(
Δ)
W
i
l
c
o
x
o
n
p
-
v
a
l
u
e
L
a
p
l
a
c
e
2
1
1
.
5
→
6
2
6
.
7
8
(
+
1
9
6
%)
5
.
1
9
→
2
.
4
2
0
.
0
8
9
6
→
0
.
6
6
1
(
+
6
3
7
%)
0
.
0
5
8
1
→
0
.
0
4
7
6
0
.
1
9
3
4
→
0
.
0
1
4
6
2
4
1
9
→
2
9
4
1
(
+
2
2
%)
<
0
.
0
1
M
u
l
t
i
p
l
i
e
d
2
1
1
.
5
→
2
3
1
.
7
8
(
+
1
0
%)
5
.
1
9
→
4
.
1
2
0
.
0
8
9
6
→
0
.
0
6
7
0
.
0
5
8
1
→
0
.
0
3
8
1
0
.
1
9
3
4
→
0
.
0
4
5
4
2
4
1
9
→
1
4
4
7
=
0
.
1
2
S
o
b
e
l
2
1
1
.
5
→
1
1
7
8
.
3
5
(
+
4
5
7
%)
5
.
1
9
→
2
.
7
1
0
.
0
8
9
6
→
0
.
0
6
7
1
0
.
0
5
8
1
→
0
.
0
5
9
7
0
.
1
9
3
4
→
0
.
0
1
6
1
2
4
1
9
→
3
5
8
2
(
+
4
8
%)
<
0
.
0
0
1
S
h
a
r
p
e
n
e
d
2
1
1
.
5
→
4
3
3
.
0
9
(
+
1
0
5
%)
5
.
1
9
→
5
.
6
4
(
+
9
%)
0
.
0
8
9
6
→
0
.
0
9
7
2
(
+
8
%)
0
.
0
5
8
1
→
0
.
0
7
6
1
0
.
1
9
3
4
→
0
.
1
9
3
5
2
4
1
9
→
2
7
9
0
<
0
.
0
5
H
i
st
o
g
r
a
m
2
1
1
.
5
→
8
5
8
.
9
(
+
3
0
6
%)
5
.
1
9
→
5
.
3
8
(
+
4
%)
0
.
0
8
9
6
→
0
.
0
9
6
4
(
+
8
%)
0
.
0
5
8
1
→
0
.
0
7
6
1
0
.
1
9
3
4
→
0
.
2
2
0
2
4
1
9
→
3
0
5
6
(
+
2
6
%)
<
0
.
0
0
1
M
L
(
C
N
N
-
b
a
se
d
)
2
1
1
.
5
→
2
6
3
.
5
1
(
+
2
5
%)
5
.
1
9
→
6
.
0
2
(
+
1
6
%)
0
.
0
8
9
6
→
0
.
1
0
8
4
(
+
2
1
%)
0
.
0
5
8
1
→
0
.
0
9
4
2
(
+
6
2
%)
0
.
1
9
3
4
→
0
.
1
9
0
2
2
4
1
9
→
4
0
1
0
(
+
6
6
%)
<
0
.
0
0
1
T
ab
le
4
.
R
ig
h
t
u
p
p
er
li
m
b
s
k
el
etal
i
m
a
g
e
(
MU
R
A
)
M
e
t
h
o
d
S
h
a
r
p
n
e
ss (Δ
)
En
t
r
o
p
y
(
Δ)
R
M
S
c
o
n
t
r
a
s
t
(
Δ)
M
e
a
n
g
r
a
d
i
e
n
t
(
Δ)
B
r
i
g
h
t
n
e
ss (Δ
)
Ed
g
e
i
n
t
e
n
si
t
y
(
Δ)
W
i
l
c
o
x
o
n
p
-
v
a
l
u
e
L
a
p
l
a
c
e
5
4
7
.
3
5
→
3
2
2
.
7
5
(
−
4
1
%)
5
.
2
9
→
1
.
5
4
0
.
1
5
1
1
→
0
.
0
5
8
7
0
.
0
7
2
5
→
0
.
0
4
8
4
0
.
1
4
1
9
→
0
.
0
1
0
6
1
8
5
2
→
3
4
7
9
(
+
8
8
%)
p
=
0
.
0
4
M
u
l
t
i
p
l
i
e
d
5
4
7
.
3
5
→
6
8
2
.
8
1
(
+
2
5
%)
5
.
2
9
→
3
.
5
0
0
.
1
5
1
1
→
0
.
1
4
3
2
0
.
0
7
2
5
→
0
.
0
6
0
6
0
.
1
4
1
9
→
0
.
0
4
5
1
1
8
5
2
→
1
8
1
7
p
=
0
.
1
8
S
o
b
e
l
5
4
7
.
3
5
→
1
1
9
9
.
2
6
(
+
1
1
9
%)
5
.
2
9
→
2
.
0
6
0
.
1
5
1
1
→
0
.
0
9
1
0
0
.
0
7
2
5
→
0
.
0
9
3
8
(
+
2
9
%)
0
.
1
4
1
9
→
0
.
0
1
8
1
1
8
5
2
→
3
6
2
8
(
+
9
6
%)
p
<
0
.
0
0
1
S
h
a
r
p
e
n
e
d
5
4
7
.
3
5
→
9
6
9
.
3
6
(
+
7
7
%)
5
.
2
9
→
5
.
6
6
(
+
7
%)
0
.
1
5
1
1
→
0
.
1
5
5
6
(
+
3
%)
0
.
0
7
2
5
→
0
.
0
8
6
8
0
.
1
4
1
9
→
0
.
1
4
7
7
1
8
5
2
→
2
0
3
0
p
<
0
.
0
1
H
i
st
o
g
r
a
m
5
4
7
.
3
5
→
7
2
4
.
7
9
(
+
3
2
%)
5
.
2
9
→
5
.
2
3
(
−
1
%)
0
.
1
5
1
1
→
0
.
2
9
1
8
(
+
9
3
%)
0
.
0
7
2
5
→
0
.
1
4
6
0
(
+
1
0
1
%)
0
.
1
4
1
9
→
0
.
5
1
6
6
(
+
2
6
4
%)
1
8
5
2
→
7
5
6
7
(
+
3
0
9
%)
p
<
0
.
0
0
1
M
L
(
C
N
N
-
b
a
se
d
)
5
4
7
.
3
5
→
6
7
6
.
1
7
(
+
2
3
%)
5
.
2
9
→
5
.
7
0
(
+
8
%)
0
.
1
5
1
1
→
0
.
1
5
6
9
(
+
4
%)
0
.
0
7
2
5
→
0
.
1
5
6
8
(
+
1
1
6
%)
0
.
1
4
1
9
→
0
.
1
4
5
4
1
8
5
2
→
2
5
9
8
(
+
4
0
%)
p
<
0
.
0
0
1
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
C
o
mp
a
r
a
tive
ev
a
lu
a
tio
n
o
f c
la
s
s
ica
l a
n
d
ma
ch
in
e
le
a
r
n
in
g
m
eth
o
d
s
fo
r
med
ica
l
…
(
Md
.
Meh
ed
i H
a
s
an
)
1339
4
.
5
.
Dis
cus
s
io
ns
T
h
e
co
m
p
ar
ati
v
e
e
v
alu
a
tio
n
o
f
t
h
e
s
ix
en
h
a
n
ce
m
e
n
t
tech
n
iq
u
es
ac
r
o
s
s
HM
P
V
C
T
s
ca
n
s
an
d
MU
R
A
s
k
eleta
l
r
ad
io
g
r
ap
h
s
r
ev
ea
ls
t
h
at
w
h
ile
cla
s
s
ical
o
p
er
ato
r
s
p
r
o
v
id
e
s
ig
n
if
ican
t
tar
g
e
ted
i
m
p
r
o
v
e
m
en
t
s
,
th
e
C
NN
-
b
ased
m
o
d
el
o
f
f
er
s
th
e
m
o
s
t
r
o
b
u
s
t
clin
ical
b
alan
ce
.
T
h
e
f
o
llo
w
in
g
T
ab
le
5
s
y
n
t
h
e
s
izes
th
e
q
u
a
n
titati
v
e
p
er
f
o
r
m
a
n
ce
,
th
e
s
tatis
tical
v
alid
it
y
o
f
t
h
e
i
m
p
r
o
v
e
m
en
ts
,
an
d
th
e
p
r
ac
tical
clin
ic
al
i
m
p
l
icatio
n
s
f
o
r
ea
ch
m
et
h
o
d
an
d
T
ab
le
6
p
r
esen
ts
a
s
y
s
te
m
a
tic
b
en
c
h
m
ar
k
i
n
g
o
f
m
etr
ics
li
k
e
m
ea
n
s
q
u
ar
ed
er
r
o
r
(
MSE
)
,
P
SNR
,
an
d
b
lin
d
/r
ef
er
en
ce
less
i
m
a
g
e
s
p
atial
q
u
alit
y
e
v
al
u
ato
r
(
B
R
I
S
QUE
)
.
T
h
ese
r
esu
lt
s
d
e
m
o
n
s
tr
ate
th
a
t
w
h
i
le
class
ical
m
e
th
o
d
s
ex
ce
l
in
s
p
e
cif
ic
f
il
ter
in
g
tas
k
s
,
th
e
C
NN
-
b
ased
ap
p
r
o
ac
h
p
r
o
v
id
es
a
m
o
r
e
b
alan
ce
d
p
r
o
f
ile
ac
r
o
s
s
p
er
ce
p
tu
al
an
d
n
o
is
e
-
r
e
d
u
ctio
n
m
etr
ics.
T
ab
le
5
.
C
o
m
p
ar
ativ
e
a
n
al
y
s
i
s
o
f
en
h
a
n
ce
m
e
n
t
m
e
th
o
d
s
an
d
clin
ical
u
tili
t
y
M
e
t
h
o
d
K
e
y
me
t
r
i
c
g
a
i
n
S
t
a
t
i
st
i
c
a
l
S
i
g
.
(
W
i
l
c
o
x
o
n
)
C
l
i
n
i
c
a
l
/
p
r
a
c
t
i
c
a
l
i
m
p
l
i
c
a
t
i
o
n
s
LT
S
h
a
r
p
n
e
ss:
+
1
0
%
to
+
1
9
6
%
<
0
.
0
5
$
B
e
st
f
o
r
h
i
g
h
-
f
r
e
q
u
e
n
c
y
sh
a
r
p
e
n
i
n
g
of
f
i
n
e
p
u
l
mo
n
a
r
y
v
e
ss
e
l
s,
but
r
e
q
u
i
r
e
s
p
r
e
-
smo
o
t
h
i
n
g
to
a
v
o
i
d
n
o
i
se
a
m
p
l
i
f
i
c
a
t
i
o
n
.
SO
S
h
a
r
p
n
e
ss:
up
to
+
4
5
7
%
<
0
.
0
0
1
O
p
t
i
mal
f
o
r
b
o
u
n
d
a
r
y
d
e
l
i
n
e
a
t
i
o
n
in
mu
sc
u
l
o
sk
e
l
e
t
a
l
i
mag
i
n
g
.
Esse
n
t
i
a
l
f
o
r
d
e
t
e
c
t
i
n
g
h
a
i
r
l
i
n
e
f
r
a
c
t
u
r
e
s
w
h
e
r
e
e
d
g
e
c
l
a
r
i
t
y
is
c
r
i
t
i
c
a
l
.
HE
Ed
g
e
i
n
t
e
n
s
i
t
y
:
+
3
0
9
%
<
0
.
0
0
1
S
u
p
e
r
i
o
r
f
o
r
g
l
o
b
a
l
c
o
n
t
r
a
s
t
s
t
r
e
t
c
h
i
n
g
in
"
w
a
sh
e
d
-
o
u
t
"
or
l
o
w
-
d
y
n
a
mi
c
-
r
a
n
g
e
sca
n
s.
I
d
e
a
l
f
o
r
r
a
p
i
d
i
n
i
t
i
a
l
d
i
a
g
n
o
st
i
c
scre
e
n
i
n
g
.
U
n
sh
a
r
p
m
a
sk
i
n
g
S
h
a
r
p
n
e
ss:
+
1
0
5
%
<
0.
01
Ef
f
e
c
t
i
v
e
at
r
e
d
u
c
i
n
g
"
h
a
z
e
"
in
d
i
g
i
t
a
l
r
a
d
i
o
g
r
a
p
h
s,
t
h
o
u
g
h
it
may
i
n
t
r
o
d
u
c
e
a
r
t
i
f
a
c
t
s
if
t
h
e
sh
a
r
p
e
n
i
n
g
f
a
c
t
o
r
is
n
o
t
t
u
n
e
d
p
e
r
mo
d
a
l
i
t
y
.
M
u
l
t
i
p
l
i
e
d
m
e
t
h
o
d
R
M
S
c
o
n
t
r
a
st
:
+
6
%
=
0
.
0
3
U
se
f
u
l
f
o
r
sp
e
c
i
f
i
c
f
e
a
t
u
r
e
a
m
p
l
i
f
i
c
a
t
i
o
n
(
e
.
g
.
,
v
i
r
u
s
me
mb
r
a
n
e
b
o
u
n
d
a
r
i
e
s)
b
u
t
l
a
c
k
s
t
h
e
r
o
b
u
st
n
e
ss
f
o
r
g
e
n
e
r
a
l
a
n
a
t
o
mi
c
a
l
v
i
s
u
a
l
i
z
a
t
i
o
n
.
ML
(
C
N
N
-
b
a
se
d
)
En
t
r
o
p
y
:
+
3
.
6
%
to
+
1
6
%
<
0
.
0
0
1
P
r
o
v
i
d
e
s
b
a
l
a
n
c
e
d
i
n
f
o
r
mat
i
o
n
p
r
e
se
r
v
a
t
i
o
n
.
U
n
l
i
k
e
o
t
h
e
r
s,
it
i
n
c
r
e
a
se
s
d
e
t
a
i
l
(
E
n
t
r
o
p
y
)
w
i
t
h
o
u
t
n
o
i
se
,
ma
k
i
n
g
it
t
h
e
mo
st
r
e
l
i
a
b
l
e
fo
r
AI
-
a
i
d
e
d
d
i
a
g
n
o
si
s.
T
ab
le
6
.
Qu
an
titati
v
e
co
m
p
ar
i
s
o
n
o
f
i
m
a
g
e
en
h
an
ce
m
e
n
t
m
e
th
o
d
s
v
ia
s
tan
d
ar
d
q
u
alit
y
m
e
t
r
ics
M
e
t
h
o
d
M
S
E
↓
P
S
N
R
(
d
B
)
↑
S
S
I
M
↑
N
I
Q
E
↓
B
R
I
S
Q
U
E
↓
L
a
p
l
a
c
e
0
.
4
1
3
3
3
.
8
4
0
.
1
7
4
3
5
.
3
7
6
1
.
1
M
u
l
t
i
p
l
i
e
d
i
m
a
g
e
0
.
0
7
9
3
4
3
1
1
.
0
0
4
9
0
.
5
2
3
4
1
4
.
9
3
2
9
6
0
.
0
9
1
4
S
o
b
e
l
0
.
4
2
7
3
3
.
6
9
0
.
2
2
8
5
5
.
0
1
5
8
.
3
S
h
a
r
p
e
n
e
d
0
.
0
6
6
8
1
1
.
7
5
0
.
4
9
2
9
4
.
3
8
5
7
.
5
3
H
i
st
o
g
r
a
m
0
.
0
7
5
5
1
1
.
2
2
0
.
5
3
9
4
4
.
9
2
5
4
.
2
9
5.
CO
NCLU
SI
O
N
T
h
is
s
t
u
d
y
s
y
s
te
m
atica
ll
y
e
v
al
u
ated
LT
,
SO
,
HE
,
an
d
a
C
N
N
-
b
ased
r
e
f
in
e
m
e
n
t a
p
p
r
o
ac
h
f
o
r
m
ed
ica
l
i
m
a
g
e
en
h
an
ce
m
e
n
t
o
n
MU
R
A
an
d
T
C
I
A
lu
n
g
d
atasets
.
C
l
ass
ical
m
eth
o
d
s
ex
ce
l
i
n
tar
g
e
ted
task
s
:
So
b
el
f
o
r
ed
g
e
d
elin
ea
t
io
n
,
HE
f
o
r
co
n
tr
ast
(
s
tatis
ticall
y
s
ig
n
i
f
ica
n
t
i
m
p
r
o
v
e
m
e
n
ts
,
<
0
.
0
0
1
)
.
T
h
e
C
NN
-
b
ased
p
r
o
v
id
es
b
alan
ce
d
,
ad
ap
tiv
e
e
n
h
a
n
ce
m
en
t
s
u
p
er
io
r
en
tr
o
p
y
p
r
eser
v
atio
n
a
n
d
o
v
er
all
m
etr
i
c
u
p
li
f
t.
No
m
et
h
o
d
u
n
i
v
er
s
al
l
y
d
o
m
i
n
ates
s
elec
t
io
n
d
ep
en
d
s
o
n
clin
ical
p
r
io
r
ity
(
ed
g
es
v
s
.
co
n
tr
ast
v
s
.
r
o
b
u
s
tn
ess
)
.
E
x
p
er
i
m
en
tal
r
esu
lt
s
co
n
f
ir
m
s
ig
n
i
f
ica
n
t
q
u
alit
y
g
ai
n
s
,
f
ac
ilit
ati
n
g
b
etter
v
is
u
aliza
t
io
n
o
f
HM
P
V
-
r
elate
d
lu
n
g
ab
n
o
r
m
al
ities
an
d
m
u
s
c
u
lo
s
k
eleta
l
s
tr
u
ct
u
r
e
s
.
Scien
t
if
icall
y
,
t
h
i
s
w
o
r
k
id
en
ti
f
ies
a
cr
itical
p
er
f
o
r
m
an
ce
tr
ad
e
-
o
f
f
:
clas
s
ical
m
et
h
o
d
s
ex
ce
l
in
co
m
p
u
tatio
n
al
ef
f
icie
n
c
y
a
n
d
lo
ca
lized
f
ea
tu
r
e
s
h
ar
p
e
n
i
n
g
,
w
h
er
ea
s
DL
ar
ch
itect
u
r
es
o
f
f
e
r
th
e
r
o
b
u
s
t
n
es
s
r
eq
u
ir
ed
f
o
r
cr
o
s
s
-
m
o
d
alit
y
i
m
ag
e
r
ec
o
n
s
tr
u
ct
io
n
.
Fu
t
u
r
e
r
esear
ch
w
ill
f
o
cu
s
o
n
th
e
r
ea
l
-
ti
m
e
h
y
b
r
id
d
ep
lo
y
m
en
t
o
f
th
e
i
n
te
g
r
ated
p
ip
elin
e
(
L
ap
lace
,
So
b
el
,
h
is
to
g
r
a
m
,
an
d
C
NN)
w
it
h
i
n
clin
ical
w
o
r
k
f
lo
w
s
to
ass
is
t
i
n
i
m
m
ed
iate
d
ia
g
n
o
s
i
s
.
W
e
also
aim
to
r
ep
lace
cu
r
r
en
t
r
esid
u
al
b
lo
c
k
s
w
i
th
d
i
f
f
u
s
io
n
m
o
d
els
a
n
d
v
i
s
io
n
tr
a
n
s
f
o
r
m
er
s
to
f
u
r
th
er
i
m
p
r
o
v
e
t
h
e
r
ec
o
n
s
tr
u
ctio
n
o
f
f
in
e
a
n
ato
m
i
ca
l d
etails
F
UNDIN
G
I
NF
O
RM
AT
I
O
N
Au
t
h
o
r
s
s
tate
n
o
f
u
n
d
i
n
g
i
n
v
o
l
v
ed
.
AUTHO
R
CO
NT
RIB
UT
I
O
NS ST
A
T
E
M
E
NT
T
h
is
j
o
u
r
n
al
u
s
e
s
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT
)
to
r
ec
o
g
n
ize
in
d
i
v
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
t
h
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
lla
b
o
r
atio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
4
,
A
u
g
u
s
t
20
26
:
1
331
-
1
3
4
1
1340
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
Md
.
Me
h
ed
i
Hasan
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Su
j
o
n
C
h
an
d
r
a
Su
tr
ad
h
ar
✓
✓
✓
✓
✓
✓
Z
a
n
n
at
u
l Fer
d
u
s
h
ie
✓
✓
✓
✓
✓
✓
✓
✓
R
ab
e
y
a
B
asri
✓
✓
✓
✓
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
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
ma
l
a
n
a
l
y
si
s
I
:
I
n
v
e
st
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
si
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
t
h
o
r
s
s
tate
n
o
co
n
f
lic
t o
f
i
n
t
er
est.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
s
tu
d
y
u
til
ized
t
w
o
clin
ica
l
d
atasets
:
HM
P
V
(
L
u
n
g
C
T
)
,
f
o
cu
s
ed
o
n
p
u
l
m
o
n
ar
y
i
n
f
ec
tio
n
s
,
an
d
MU
R
A
(
Sk
ele
tal
X
-
r
a
y
)
,
tar
g
etin
g
li
m
b
f
r
ac
t
u
r
es.
Data
w
a
s
p
ar
titi
o
n
ed
in
to
tr
ain
in
g
,
v
al
id
atio
n
,
an
d
t
esti
n
g
s
ets
(
7
0
/1
5
/1
5
f
o
r
HM
P
V
an
d
8
0
/1
0
/1
0
f
o
r
MU
R
A
)
to
en
s
u
r
e
r
o
b
u
s
t
ev
al
u
atio
n
.
P
r
ep
r
o
ce
s
s
in
g
in
v
o
l
v
ed
no
r
m
alizi
n
g
p
ix
el
i
n
ten
s
itie
s
to
[
0
,
1
]
,
r
esizin
g
to
2
5
6
×
256
o
r
5
1
2
×
512
,
an
d
a
p
p
ly
i
n
g
r
a
n
d
o
m
f
l
i
p
s
an
d
r
o
tatio
n
s
(
±
15°
)
f
o
r
d
ata
au
g
m
en
ta
t
io
n
.
RE
F
E
R
E
NC
E
S
[
1
]
Y
.
G
u
o
,
Y
.
Z
h
u
,
L
.
L
i
u
,
a
n
d
Q
.
H
u
a
n
g
,
"
R
e
se
a
r
c
h
r
e
v
i
e
w
o
f
sp
a
c
e
-
f
r
e
q
u
e
n
c
y
d
o
ma
i
n
i
mag
e
e
n
h
a
n
c
e
me
n
t
me
t
h
o
d
s,"
C
o
m
p
u
t
e
r
En
g
i
n
e
e
r
i
n
g
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s,
v
o
l
.
5
8
,
n
o
.
1
1
,
p
p
.
2
3
–
3
2
,
2
0
2
2
,
d
o
i
:
1
0
.
3
7
7
8
/
j
.
i
s
sn
.
1
0
0
2
-
8
3
3
1
.
2
1
1
2
-
0
2
8
0
.
[
2
]
R
.
S
z
e
l
i
sk
i
,
C
o
m
p
u
t
e
r
Vi
s
i
o
n
:
A
l
g
o
ri
t
h
m
s a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
2
n
d
e
d
.
C
h
a
m,
S
w
i
t
z
e
r
l
a
n
d
:
S
p
r
i
n
g
e
r
N
a
t
u
r
e
,
2
0
2
2
.
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
0
3
0
-
3
4
3
7
2
-
9.
[
3
]
J.
J.
Je
o
n
g
,
A
.
T
a
r
i
q
,
T
.
A
d
e
j
u
mo
,
H
.
T
r
i
v
e
d
i
,
J.
W
.
G
i
c
h
o
y
a
,
a
n
d
I
.
B
a
n
e
r
j
e
e
,
“
S
y
s
t
e
mat
i
c
R
e
v
i
e
w
o
f
G
e
n
e
r
a
t
i
v
e
A
d
v
e
r
sari
a
l
N
e
t
w
o
r
k
s
(
G
A
N
s
)
f
o
r
M
e
d
i
c
a
l
I
mag
e
C
l
a
ssi
f
i
c
a
t
i
o
n
a
n
d
S
e
g
me
n
t
a
t
i
o
n
,
”
J
o
u
rn
a
l
o
f
D
i
g
i
t
a
l
I
m
a
g
i
n
g
,
v
o
l
.
3
5
,
n
o
.
2
,
p
p
.
1
3
7
–
1
5
2
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
0
2
7
8
-
0
2
1
-
0
0
5
5
6
-
w.
[
4
]
A
.
S
a
r
mad
i
,
Z
.
S
.
R
a
z
a
v
i
,
A
.
J.
v
a
n
W
i
j
n
e
n
,
a
n
d
M
.
S
o
l
t
a
n
i
,
“
C
o
mp
a
r
a
t
i
v
e
a
n
a
l
y
si
s
o
f
v
i
si
o
n
t
r
a
n
sf
o
r
me
r
s
a
n
d
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
i
n
o
st
e
o
p
o
r
o
si
s
d
e
t
e
c
t
i
o
n
f
r
o
m
X
-
r
a
y
i
m
a
g
e
s,”
S
c
i
e
n
t
i
f
i
c
R
e
p
o
rt
s
,
v
o
l
.
1
4
,
n
o
.
1
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
3
8
/
s4
1
5
9
8
-
024
-
6
9
1
1
9
-
7.
[
5
]
D
.
R
.
S
a
r
v
a
m
a
n
g
a
l
a
a
n
d
R
.
V
.
K
u
l
k
a
r
n
i
,
“
C
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
i
n
me
d
i
c
a
l
i
m
a
g
e
u
n
d
e
r
st
a
n
d
i
n
g
:
a
s
u
r
v
e
y
,
”
Ev
o
l
u
t
i
o
n
a
r
y
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
1
5
,
n
o
.
1
,
p
p
.
1
–
2
2
,
M
a
r
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
2
0
6
5
-
0
2
0
-
0
0
5
4
0
-
3.
[
6
]
S
.
T
a
k
a
h
a
s
h
i
e
t
a
l
.
,
“
C
o
mp
a
r
i
so
n
o
f
V
i
s
i
o
n
T
r
a
n
sf
o
r
me
r
s
a
n
d
C
o
n
v
o
l
u
t
i
o
n
a
l
N
e
u
r
a
l
N
e
t
w
o
r
k
s
i
n
M
e
d
i
c
a
l
I
mag
e
A
n
a
l
y
si
s:
A
S
y
st
e
mat
i
c
R
e
v
i
e
w
,
”
J
o
u
r
n
a
l
o
f
Me
d
i
c
a
l
S
y
st
e
m
s
,
v
o
l
.
4
8
,
n
o
.
1
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
9
1
6
-
0
2
4
-
0
2
1
0
5
-
8.
[
7
]
Z
.
L
i
-
p
i
n
g
a
n
d
H
.
L
i
a
n
-
q
i
n
g
,
“
M
e
d
i
c
a
l
I
mag
e
En
h
a
n
c
i
n
g
M
e
t
h
o
d
B
a
se
d
o
n
L
o
c
a
l
l
y
R
e
d
i
st
r
i
b
u
t
e
d
H
i
s
t
o
g
r
a
ms,”
J
o
u
r
n
a
l
o
f
O
p
t
o
e
l
e
c
t
r
o
n
i
c
s·
l
a
ser
,
v
o
l
.
5
,
p
p
.
3
9
9
–
4
0
2
,
2
0
1
0
.
[
8
]
N
a
g
e
sh
a
a
n
d
G
.
H
.
K
u
mar,
“
A
l
e
v
e
l
c
r
o
ssi
n
g
e
n
h
a
n
c
e
me
n
t
s
c
h
e
me
f
o
r
c
h
e
st
r
a
d
i
o
g
r
a
p
h
i
m
a
g
e
s,”
C
o
m
p
u
t
e
r
s
i
n
Bi
o
l
o
g
y
a
n
d
Me
d
i
c
i
n
e
,
v
o
l
.
3
7
,
n
o
.
1
0
,
p
p
.
1
4
5
5
–
1
4
6
0
,
O
c
t
.
2
0
0
7
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
mp
b
i
o
me
d
.
2
0
0
7
.
0
1
.
0
0
2
.
[
9
]
P
.
Y
a
n
,
R
e
se
a
rc
h
a
n
d
I
m
p
l
e
m
e
n
t
a
t
i
o
n
o
f
L
u
n
g
C
T
I
m
a
g
e
En
h
a
n
c
e
m
e
n
t
A
l
g
o
ri
t
h
m
.
S
h
e
n
y
a
n
g
A
e
r
o
sp
a
c
e
U
n
i
v
e
r
si
t
y
,
2
0
0
9
.
[
1
0
]
J.
W
u
e
t
a
l
.
,
“
A
p
p
l
i
c
a
t
i
o
n
o
f
i
m
a
g
e
e
n
h
a
n
c
e
me
n
t
t
e
c
h
n
o
l
o
g
y
i
n
d
i
g
i
t
a
l
X
-
r
a
y
me
d
i
c
a
l
i
mag
e
”
(
i
n
C
h
i
n
e
se
)
,
C
h
i
n
a
M
e
d
i
c
a
l
Eq
u
i
p
m
e
n
t
,
v
o
l
.
9
,
n
o
.
5
,
p
p
.
6
0
–
6
2
,
2
0
1
2
.
[
1
1
]
B
.
Z
h
a
n
g
,
B
.
G
a
o
,
S
.
L
i
a
n
g
,
X
.
L
i
,
a
n
d
H
.
W
a
n
g
,
“
A
c
l
a
ssi
f
i
c
a
t
i
o
n
a
l
g
o
r
i
t
h
m
b
a
se
d
o
n
i
mp
r
o
v
e
d
me
t
a
l
e
a
r
n
i
n
g
a
n
d
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
f
o
r
f
e
w
-
sh
o
t
me
d
i
c
a
l
i
m
a
g
e
s,”
I
ET I
m
a
g
e
Pr
o
c
e
ssi
n
g
,
v
o
l
.
1
7
,
n
o
.
1
2
,
p
p
.
3
5
8
9
–
3
5
9
8
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
4
9
/
i
p
r
2
.
1
2
8
8
9
.
[
1
2
]
O
.
L
a
o
u
s
y
,
A
.
A
r
a
u
j
o
,
G
.
C
h
a
ss
a
g
n
o
n
,
N
.
P
a
r
a
g
i
o
s,
M
.
-
P
.
R
e
v
e
l
,
a
n
d
M
.
V
a
k
a
l
o
p
o
u
l
o
u
,
“
C
e
r
t
i
f
i
c
a
t
i
o
n
o
f
D
e
e
p
L
e
a
r
n
i
n
g
M
o
d
e
l
s
f
o
r
M
e
d
i
c
a
l
I
mag
e
S
e
g
me
n
t
a
t
i
o
n
,
”
L
e
c
t
u
re
N
o
t
e
s
i
n
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
.
S
p
ri
n
g
e
r
N
a
t
u
re
S
w
i
t
z
e
rl
a
n
d
,
p
p
.
6
1
1
–
6
2
1
,
2
0
2
3
.
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
0
3
1
-
4
3
9
0
1
-
8
_
5
8
.
[
1
3
]
J.
H
o
,
A
.
Ja
i
n
,
a
n
d
P
.
A
b
b
e
e
l
,
“
D
e
n
o
i
si
n
g
D
i
f
f
u
s
i
o
n
P
r
o
b
a
b
i
l
i
s
t
i
c
M
o
d
e
l
s,”
3
4
t
h
C
o
n
f
e
re
n
c
e
o
n
N
e
u
r
a
l
I
n
f
o
rm
a
t
i
o
n
Pro
c
e
ss
i
n
g
S
y
s
t
e
m
s (
N
e
u
rI
PS
2
0
2
0
)
,
V
a
n
c
o
u
v
e
r
,
C
a
n
a
d
a
,
2
0
2
0
,
d
o
i
:
1
0
.
4
8
5
5
0
/
A
R
X
I
V
.
2
0
0
6
.
1
1
2
3
9
.
[
1
4
]
Z
.
W
a
n
g
a
n
d
M
.
J
.
C
.
S
a
mo
n
t
e
,
“
R
e
se
a
r
c
h
o
n
l
o
w
-
d
o
se
C
T
i
mag
e
e
n
h
a
n
c
e
me
n
t
a
l
g
o
r
i
t
h
m
b
a
se
d
o
n
d
e
e
p
l
e
a
r
n
i
n
g
,
”
2
0
2
4
5
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
Vi
s
i
o
n
,
I
m
a
g
e
a
n
d
D
e
e
p
L
e
a
r
n
i
n
g
(
C
VI
D
L
)
.
I
EE
E
,
p
p
.
6
1
9
–
6
2
4
,
A
p
r
.
1
9
,
2
0
2
4
.
d
o
i
:
1
0
.
1
1
0
9
/
c
v
i
d
l
6
2
1
4
7
.
2
0
2
4
.
1
0
6
0
4
0
8
9
.
[
1
5
]
X
.
M
e
n
g
,
K
.
S
u
n
,
J.
X
u
,
X
.
H
e
,
a
n
d
D
.
S
h
e
n
,
“
M
u
l
t
i
-
M
o
d
a
l
M
o
d
a
l
i
t
y
-
M
a
sk
e
d
D
i
f
f
u
si
o
n
N
e
t
w
o
r
k
f
o
r
B
r
a
i
n
M
R
I
S
y
n
t
h
e
si
s
W
i
t
h
R
a
n
d
o
m
M
o
d
a
l
i
t
y
M
i
ss
i
n
g
,
”
I
E
EE
T
r
a
n
s.
Me
d
.
I
m
a
g
i
n
g
,
v
o
l
.
4
3
,
n
o
.
7
,
p
p
.
2
5
8
7
–
2
5
9
8
,
Ju
l
.
2
0
2
4
,
d
o
i
:
10
.
1
1
0
9
/
t
m
i
.
2
0
2
4
.
3
3
6
8
6
6
4
.
[
1
6
]
P
.
R
a
j
p
u
r
k
a
r
e
t
a
l
.
,
“
M
U
R
A
:
L
a
r
g
e
D
a
t
a
se
t
f
o
r
A
b
n
o
r
mal
i
t
y
D
e
t
e
c
t
i
o
n
i
n
M
u
sc
u
l
o
sk
e
l
e
t
a
l
R
a
d
i
o
g
r
a
p
h
s
,
”
M
a
y
2
0
1
8
,
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
:
/
/
a
r
x
i
v
.
o
r
g
/
a
b
s
/
1
7
1
2
.
0
6
9
5
7
[
1
7
]
K
.
C
l
a
r
k
e
t
a
l
.
,
“
T
h
e
c
a
n
c
e
r
i
mag
i
n
g
a
r
c
h
i
v
e
(
T
C
I
A
)
:
M
a
i
n
t
a
i
n
i
n
g
a
n
d
o
p
e
r
a
t
i
n
g
a
p
u
b
l
i
c
i
n
f
o
r
mat
i
o
n
r
e
p
o
s
i
t
o
r
y
,
”
J
o
u
r
n
a
l
o
f
D
i
g
i
t
a
l
I
m
a
g
i
n
g
,
v
o
l
.
2
6
,
n
o
.
6
,
p
p
.
1
0
4
5
–
1
0
5
7
,
2
0
1
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
2
7
8
-
0
1
3
-
9
6
2
2
-
7.
[
1
8
]
R
.
C
.
G
o
n
z
a
l
e
z
a
n
d
R
.
E.
W
o
o
d
s,
D
i
g
i
t
a
l
I
m
a
g
e
Pr
o
c
e
ss
i
n
g
,
4
t
h
e
d
.
U
p
p
e
r
S
a
d
d
l
e
R
i
v
e
r
,
2
0
1
8
.
[
1
9
]
X
.
R
e
n
a
n
d
S
.
L
a
i
,
“
M
e
d
i
c
a
l
I
ma
g
e
En
h
a
n
c
e
me
n
t
B
a
se
d
o
n
L
a
p
l
a
c
e
T
r
a
n
sf
o
r
m,
S
o
b
e
l
O
p
e
r
a
t
o
r
a
n
d
H
i
s
t
o
g
r
a
m
E
q
u
a
l
i
z
a
t
i
o
n
,
”
Evaluation Warning : The document was created with Spire.PDF for Python.