T
E
L
KO
M
NIK
A
,
V
ol
.
17
,
No.
5,
O
c
tob
er
20
1
9,
p
p.
2
68
4
~
2696
IS
S
N: 1
69
3
-
6
93
0
,
accr
ed
ited
F
irst
Gr
ad
e b
y K
em
en
r
istekdikti,
Decr
ee
No: 2
1/E/
K
P
T
/20
18
DOI:
10.12928/TE
LK
OM
N
IK
A
.v
1
7
i
5
.
13292
◼
26
84
Rec
ei
v
ed
J
an
u
ary
5
, 2
01
9
;
Rev
i
s
ed
Ma
y
2
8
, 2
01
9
;
A
c
c
ep
te
d
J
un
e
12
,
20
1
9
A
d
v
a
n
ced
w
atermarkin
g t
ec
hn
i
qu
e
to
i
mp
ro
v
e
me
d
ical ima
ges’ se
c
urit
y
M
edia
A
n
u
g
er
ah
A
yu
*
,
T
edd
y
M
ant
o
r
o
,
I M
ade
A
lan
P
r
i
ya
t
n
a
Dep
a
rtm
e
n
t
o
f
Co
m
p
u
t
e
r Sc
i
e
n
c
e
,
Fa
c
u
l
ty
o
f
E
n
g
i
n
e
e
ri
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
Sa
m
p
o
e
rn
a
U
n
i
v
e
rs
i
ty
,
J
a
k
a
rt
a
,
In
d
o
n
e
s
i
a
*C
o
rre
s
p
o
n
d
i
n
g
a
u
th
o
r,
e
-
m
a
i
l
:
m
e
d
i
a
.a
y
u
@s
a
m
p
o
e
rn
a
u
n
i
v
e
rs
i
ty
.a
c
.i
d
Ab
strac
t
Ad
v
a
n
c
e
s
i
n
i
m
a
g
i
n
g
te
c
h
n
o
l
o
g
y
h
a
v
e
m
a
d
e
m
e
d
i
c
a
l
i
m
a
g
e
s
b
e
c
o
m
e
o
n
e
o
f
th
e
i
m
p
o
rta
n
t
s
o
u
r
c
e
s
fo
r
i
n
fo
rm
a
ti
o
n
i
n
s
u
p
p
o
rti
n
g
a
c
c
u
ra
te
d
i
a
g
n
o
s
e
s
a
n
d
tre
a
tm
e
n
t
d
e
c
i
s
i
o
n
s
b
y
d
o
c
to
r
s
fo
r
th
e
i
r
p
a
ti
e
n
ts
.
Howe
v
e
r,
th
e
v
u
l
n
e
ra
b
i
l
i
ty
o
f
m
e
d
i
c
a
l
i
m
a
g
e
s
’
s
e
c
u
ri
ty
i
s
h
i
g
h
.
Th
e
i
m
a
g
e
s
c
a
n
b
e
e
a
s
i
l
y
‘
a
tt
a
c
k
e
d
’
,
whi
c
h
a
l
te
re
d
t
h
e
i
r
i
n
f
o
rm
a
ti
o
n
t
h
a
t
c
a
n
l
e
a
d
t
o
i
n
c
o
rr
e
c
t
d
i
a
g
n
o
s
e
s
o
r
tre
a
tm
e
n
t.
In
o
rd
e
r
t
o
m
a
k
e
th
e
i
m
a
g
e
s
l
e
s
s
v
u
l
n
e
ra
b
l
e
fro
m
o
u
t
s
i
d
e
a
tt
a
c
k
s
,
t
h
i
s
s
tu
d
y
p
ro
p
o
s
e
s
to
s
e
c
u
re
t
h
e
m
b
y
a
d
v
a
n
c
i
n
g
th
e
wat
e
rm
a
rk
i
n
g
u
s
i
n
g
d
u
a
l
-
l
a
y
e
r
fra
g
i
l
e
te
c
h
n
i
q
u
e
.
It
i
s
e
x
p
e
c
t
e
d
t
h
a
t
th
i
s
d
u
a
l
-
l
a
y
e
r
fra
g
i
l
e
w
a
te
rm
a
rk
i
n
g
wil
l
g
u
a
ra
n
te
e
th
e
i
n
te
g
r
i
ty
,
a
u
th
e
n
t
i
c
i
ty
,
a
n
d
c
o
n
fi
d
e
n
t
i
a
l
i
ty
o
f
p
a
t
i
e
n
t’
s
a
n
d
a
n
y
o
th
e
r
i
m
p
o
r
ta
n
t
i
n
fo
rm
a
t
i
o
n
a
n
d
a
l
s
o
th
e
p
i
x
e
l
d
a
ta
o
f
t
h
e
m
e
d
i
c
a
l
i
m
a
g
e
s
.
Th
e
wo
rk
i
n
t
h
i
s
s
tu
d
y
i
m
p
l
e
m
e
n
ts
two
L
SB
s
o
f
i
m
a
g
e
wh
e
r
e
th
e
ro
l
e
o
f
t
h
e
f
i
rs
t
L
SB
i
s
a
s
a
ta
m
p
e
r
d
e
t
e
c
to
r
,
a
n
d
t
h
e
s
e
c
o
n
d
L
SB
i
s
u
s
e
d
t
o
s
to
re
p
a
ti
e
n
t’
s
a
n
d
a
n
y
o
th
e
r
i
m
p
o
rta
n
t
i
n
f
o
rm
a
ti
o
n
.
M
e
d
i
c
a
l
i
m
a
g
e
s
o
f
f
o
u
r
d
e
a
d
l
i
e
s
t
d
i
s
e
a
s
e
s
i
n
In
d
o
n
e
s
i
a
w
e
re
u
s
e
d
to
te
s
t
th
e
p
r
o
p
o
s
e
d
wat
e
rm
a
r
k
i
n
g
te
c
h
n
i
q
u
e
.
Res
u
l
t
s
fr
o
m
th
e
c
o
n
d
u
c
t
e
d
t
e
s
t
s
s
h
o
w th
a
t
th
e
p
ro
p
o
s
e
d
te
c
h
n
i
q
u
e
a
b
l
e
t
o
g
e
n
e
r
a
te
a
wa
t
e
rm
a
rk
e
d
i
m
a
g
e
t
h
a
t
h
a
s
n
o
n
o
t
i
c
e
a
b
l
e
c
h
a
n
g
e
s
c
o
m
p
a
re
d
t
o
i
t
s
o
ri
g
i
n
a
l
i
m
a
g
e
,
wit
h
PSNR
v
a
l
u
e
m
o
re
th
a
n
4
4
d
B
a
n
d
SSIM
v
a
l
u
e
o
f
a
l
m
o
s
t
1
,
whe
re
th
e
ta
m
p
e
r
d
e
t
e
c
t
o
r
c
a
n
c
o
rr
e
c
t
l
y
d
e
te
c
t
a
n
d
l
o
c
a
l
i
z
e
a
n
y
t
a
m
p
e
r
i
n
g
o
n
th
e
wa
te
rm
a
r
k
e
d
i
m
a
g
e
.
F
u
rth
e
rm
o
re
,
th
e
p
ro
p
o
s
e
d
t
e
c
h
n
i
q
u
e
h
a
s
s
h
o
w
n
t
o
h
a
v
e
a
h
i
g
h
e
r
l
e
v
e
l
o
f
s
e
c
u
ri
ty
o
n
m
e
d
i
c
a
l
i
m
a
g
e
s
,
c
o
m
p
a
re
d
t
o
DICO
M
s
ta
n
d
a
r
d
a
n
d
s
ta
n
d
a
r
d
wat
e
rm
a
rk
i
n
g
m
e
th
o
d
.
Key
w
ords
:
DICO
M
,
fra
g
i
l
e
wa
te
rm
a
rk
i
n
g
,
L
SB,
m
e
d
i
c
a
l
i
m
a
g
e
,
ta
m
p
e
r
d
e
te
c
ti
o
n
Copy
righ
t
©
2
0
1
9
Uni
v
e
rsi
t
a
s
Ahm
a
d
D
a
hl
a
n.
All
rig
ht
s
r
e
s
e
rve
d
.
1.
Int
r
o
d
u
ctio
n
T
he
Dev
e
l
op
m
en
t
i
n
m
ed
i
c
al
i
m
ag
i
ng
t
ec
hn
o
l
o
g
y
h
a
s
m
ad
e
m
ed
i
c
al
i
m
ag
es
c
an
be
us
ed
t
o
d
o
e
arl
y
d
ete
c
t
i
o
n
of
v
ario
us
d
i
s
ea
s
es
,
i
n
c
l
ud
i
ng
di
s
e
as
es
r
eg
arded
as
de
ad
l
i
es
t
di
s
ea
s
es
.
Da
ta
f
r
o
m
W
o
r
l
d
Heal
t
h
O
r
ga
n
i
z
ati
on
(
W
H
O
)
,
i
n
20
1
3,
s
tr
ok
e
an
d
he
art
di
s
ea
s
e
ar
e
the
de
ad
l
i
es
t
di
s
ea
s
e
i
n
I
n
do
ne
s
i
a.
A
r
ou
n
d
3
m
i
l
l
i
on
of
Ind
on
es
i
an
wer
e
di
ag
no
s
ed
wi
th
he
art
di
s
ea
s
e
an
d
2
m
i
l
l
i
on
w
er
e
di
a
gn
os
e
d
w
i
th
s
tr
ok
e.
O
n
wom
an
s
i
de
,
i
n
the
s
am
e
y
e
ar,
c
er
v
i
c
a
l
c
an
c
er
an
d
bre
as
t
c
an
c
er
are
the
d
ea
dl
i
es
t
di
s
ea
s
e
i
n
Ind
o
ne
s
i
a
w
i
t
h
98
,69
2
c
a
s
es
f
or
c
erv
i
c
al
c
an
c
er
an
d
61
,
68
2
c
as
es
f
or
breas
t
c
an
c
er
[1]
.
In
l
i
ne
wi
th
t
ha
t
the
hi
gh
n
um
be
r
of
c
erv
i
c
a
l
c
an
c
er
c
as
es
ha
s
m
ad
e
In
do
ne
s
i
a
be
c
om
e
the
c
o
un
tr
y
wi
th
the
s
ec
on
d
hi
gh
es
t
i
n
the
w
or
l
d
f
or
c
erv
i
c
al
c
an
c
er [2
].
E
arl
y
de
tec
ti
on
of
the
d
i
s
e
as
es
uti
l
i
z
i
ng
m
ed
i
c
al
i
m
ag
i
ng
t
ec
hn
o
l
o
g
y
i
s
ex
pe
c
ted
to
be
ab
l
e
to
m
i
ni
m
i
z
e
th
e
nu
m
be
r
of
d
ea
ths
c
au
s
e
d
b
y
tho
s
e
di
s
ea
s
es
.
Me
d
i
c
al
i
m
ag
i
ng
tec
h
no
l
og
y
i
s
an
ad
v
an
c
e
d t
ec
hn
o
l
o
g
y
tha
t
ha
s
a
f
un
c
t
i
on
to
prod
u
c
e a
v
i
s
ua
l
r
e
pres
en
t
ati
on
o
f
hu
m
an
bo
d
y
whi
c
h
i
s
us
e
d
to
d
o
m
ore
ac
c
urate
di
ag
n
os
es
an
d
a
pp
r
op
r
i
ate
tr
e
atm
en
t
de
c
i
s
i
on
.
In
a
dd
i
ti
on
,
m
ed
i
c
al
i
m
ag
e
c
a
n
be
s
en
d
o
v
er
th
e
i
nt
ernet
(
w
e
l
l
k
no
w
n
as
t
el
em
ed
i
c
i
ne
)
whi
c
h
a
l
l
o
w
do
c
tors
,
ph
y
s
i
c
i
an
s
,
an
d
an
y
ot
he
r
he
a
l
th
c
are
prof
es
s
i
on
a
l
s
d
o
an
e
v
a
l
u
ati
on
,
d
i
a
gn
os
e
an
d
tr
e
atm
en
t
f
r
o
m
a
di
s
tan
c
e
[3]
.
T
he
r
e
are
v
ario
us
t
y
pe
s
or
m
od
al
i
ti
es
of
m
ed
i
c
al
i
m
ag
es
a
v
ai
l
ab
l
e,
s
uc
h
as
r
ad
i
o
graph
y
,
c
om
pu
ted
to
m
og
r
ap
h
y
,
m
ag
ne
ti
c
r
es
o
na
nc
e
i
m
ag
i
n
g,
ul
tr
as
ou
n
d
i
m
ag
i
ng
[4]
.
W
h
ereas
,
the
d
i
f
f
erenc
e
b
et
w
ee
n
e
ac
h
m
od
al
i
t
y
i
s
on
t
he
t
y
p
es
of
en
ergi
es
an
d
ac
qu
i
s
i
ti
on
tec
hn
o
l
og
y
us
ed
to
prod
u
c
e
m
ed
i
c
al
i
m
ag
e.
Ma
gn
e
ti
c
Res
o
na
nc
e
I
m
ag
i
ng
a
nd
C
om
pu
ted
T
o
m
og
r
ap
h
y
(
CT
)
m
od
al
i
ti
es
are
t
he
m
os
t
c
om
m
on
us
ed
t
o
d
i
a
gn
os
e
s
tr
ok
e,
he
art
di
s
e
as
e,
c
erv
i
c
al
an
d b
r
ea
s
t c
an
c
er
[
5
,
6].
T
he
us
e
of
m
ed
i
c
al
i
m
ag
i
ng
te
c
hn
ol
og
y
c
on
ti
n
u
es
i
nc
r
ea
s
i
ng
i
n
ev
er
y
y
e
ar.
F
i
g
ure
1
s
ho
ws
the
us
ag
e
of
CT
s
an
d
MRIs
de
v
i
c
es
i
nc
r
ea
s
i
ng
f
r
om
the
l
as
t
30
y
e
ars
i
n
O
r
ga
n
i
z
at
i
on
f
or
E
c
on
om
i
c
Co
-
O
pe
r
at
i
on
a
nd
De
v
e
l
o
pm
en
t
(
O
E
CD)
c
ou
ntri
es
[
7].
A
c
c
ordi
n
g
to
M
i
n
i
s
tr
y
of
Heal
t
h
I
nd
o
ne
s
i
a,
g
ov
ernor
s
w
i
l
l
s
u
pp
ort
i
n
i
nc
r
ea
s
i
n
g
the
nu
m
be
r
of
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
A
dv
an
c
e
d wa
termark
i
ng
t
e
c
hn
i
q
ue
t
o i
mp
r
ov
e
me
d
i
c
al
i
ma
ge
s
’
…
(
Me
di
a A
nu
ge
r
ah
A
y
u
)
2685
he
a
l
th
de
v
i
c
es
i
nc
l
u
di
n
g
C
T
s
an
d
MR
Is
d
ev
i
c
es
t
wi
c
e
or
ev
en
thre
e
t
i
m
es
m
ore
t
o
th
e
pre
v
i
o
us
nu
m
be
r
w
i
thi
n
y
ea
r
20
1
5
–
2
01
9 [
8].
Inc
r
ea
s
e
i
n
the
nu
m
be
r
of
m
ed
i
c
al
i
m
ag
i
ng
t
ec
hn
o
l
o
g
y
c
an
l
ea
d
to
m
ore
an
d
ne
w
c
y
be
r
-
threats
s
i
nc
e
i
t
i
s
tr
an
s
f
erabl
e
o
v
er
the
i
nte
r
n
et
.
T
he
r
ef
ore,
i
t
i
s
a
m
us
t
to
s
ec
ure
the
m
ed
i
c
al
i
m
ag
es
f
r
o
m
an
y
k
i
nd
of
att
ac
k
,
s
i
nc
e
an
y
c
ha
ng
e
s
on
m
ed
i
c
al
i
m
ag
es
m
a
y
l
ea
d
to
i
nc
orr
ec
t
di
a
gn
os
es
an
d
tr
ea
tm
en
t
de
c
i
s
i
o
n
s
f
or
the
pa
ti
e
nt.
In
m
ed
i
c
al
i
m
ag
i
ng
tec
hn
ol
o
g
y
arena,
D
i
g
i
ta
l
Im
ag
i
ng
an
d
Com
m
un
i
c
ati
on
i
n
M
ed
i
c
i
ne
(
DICO
M)
i
s
us
ed
as
a
n
i
nt
ernat
i
o
na
l
s
tan
da
r
d
f
or
tr
an
s
f
err
i
ng
,
s
tori
ng
,
r
etri
e
v
i
ng
m
ed
i
c
a
l
i
m
ag
e
i
nf
orm
ati
on
.
In
s
h
ort,
DICO
M
i
m
ag
e
f
or
m
at
c
on
tai
ni
ng
t
w
o
i
m
po
r
tan
t
pa
r
ts
:
T
he
he
ad
er
f
i
l
e
whi
c
h
s
av
e
i
m
po
r
tan
t
i
nf
orm
ati
on
s
uc
h
a
s
pa
ti
en
t
i
nf
orm
ati
on
,
s
tu
d
y
i
nf
or
m
ati
on
,
an
d
the
p
i
x
el
d
ata
.
DICO
M
S
t
an
d
ard
i
m
pl
em
en
t
s
ev
era
l
s
ec
urit
y
s
c
h
em
es
on
bo
th,
the
c
o
nf
i
de
n
ti
a
l
da
t
a
a
nd
i
m
ag
es
as
s
tat
ed
i
n
P
S
3
.
1
5
of
th
e
DI
CO
M
S
tan
da
r
d,
w
h
i
c
h a
b
ou
t
S
ec
urit
y
S
y
s
tem
Ma
na
g
em
en
t P
r
of
i
l
es
[
9].
DICO
M
S
t
an
d
ard
c
urr
en
t
l
y
i
m
pl
e
m
en
ts
di
g
i
ta
l
s
i
gn
at
ure
to
gu
arant
ee
t
he
au
th
en
ti
c
i
t
y
an
d
i
nt
eg
r
i
t
y
of
the
pi
x
e
l
da
ta
an
d
s
a
v
ed
i
n
grou
p
tag
(
F
F
F
A
,FF
F
A
)
.
F
urth
erm
ore,
DICO
M
S
tan
da
r
d
i
m
pl
em
en
ts
en
c
r
y
pti
o
n s
uc
h
as
A
E
S
a
nd
tr
i
p
l
e D
E
S
o
n s
e
l
ec
te
d
DICO
M
Data
S
et,
on
e
of
the
m
i
s
the
E
l
ec
tr
on
i
c
P
ati
en
t
Rec
ord
(
E
P
R)
.
M
oreov
er,
DICO
M
S
tan
da
r
d
al
s
o
s
pe
c
i
f
i
es
an
on
y
m
i
z
at
i
on
s
c
he
m
e
b
y
de
l
eti
ng
or
c
ha
ng
i
ng
an
y
c
on
f
i
de
n
ti
a
l
DICO
M
Dat
a
S
et
i
n
order
t
o
al
l
o
w
the
DICO
M
f
i
l
e
up
l
o
ad
ed
p
ub
l
i
c
l
y
f
or
c
ert
ai
n
p
urpos
es
s
uc
h
as
r
es
e
arc
h,
pres
en
tat
i
on
,
s
tud
y
,
etc
.
[9]
.
T
he
l
i
m
i
tat
i
on
s
of
th
e
s
ec
ur
i
t
y
s
c
he
m
es
i
m
pl
em
en
ted
b
y
DICO
M
S
tan
da
r
ds
are
i
nte
grit
y
a
nd
a
uth
e
nti
c
i
t
y
ar
e
no
t
ad
dres
s
ed
f
or
s
el
ec
ted
DICO
M
Data
S
e
t
an
d
t
he
an
on
y
m
i
z
ed
DICO
M f
i
l
e c
ou
l
d
no
t
be
us
ed
f
or f
urther d
i
ag
no
s
es
a
n
y
m
or
e,
s
i
nc
e t
h
e i
nf
orm
ati
on
ab
ou
t
pa
t
i
e
nt
was
r
em
ov
ed
pe
r
m
an
en
t
l
y
F
urtherm
ore,
un
au
t
ho
r
i
z
e
d
pe
o
pl
e
c
ou
l
d
e
di
t
pa
r
t
l
y
or
c
om
pl
ete
l
y
al
ter
th
e
pi
x
e
l
da
t
a
of
the
m
ed
i
c
al
i
m
ag
e
an
d
c
an
no
t
l
oc
al
i
z
e
of
al
tered
p
i
x
el
.
T
ho
s
e
be
c
om
e
the
m
aj
or l
i
m
i
tat
i
on
s
i
n t
he
DICO
M
S
tan
da
r
d [
10
,
11
].
K
no
wi
n
g
th
e
l
i
m
i
tat
i
o
ns
of
c
urr
en
t
s
ec
urit
y
s
c
he
m
es
,
the
w
ork
pres
en
te
d
i
n
th
i
s
pa
pe
r
tr
i
es
to
f
i
nd
an
d
de
v
e
l
o
p
a
s
ec
urit
y
s
c
he
m
e
f
or
m
ed
i
c
al
i
m
ag
es
.
T
hi
s
s
tud
y
propo
s
es
a
s
ec
urin
g
tec
hn
i
qu
e
w
h
i
c
h
ba
s
ed
o
n
du
al
-
l
a
y
er
f
r
ag
i
l
e
i
m
ag
e
w
ate
r
m
ar
k
i
ng
to
e
ns
ure
the
i
n
teg
r
i
t
y
,
c
on
f
i
de
nt
i
a
l
i
t
y
an
d
au
the
nti
c
i
t
y
of
the
m
ed
i
c
a
l
i
m
ag
es
.
T
he
r
em
ai
ni
ng
of
the
pa
p
er
i
s
org
an
i
z
ed
as
f
ol
l
o
w
s
:
s
ec
ti
o
n
2
di
s
c
u
s
s
es
the
l
i
tera
ture
r
e
v
i
e
w
o
n
di
g
i
ta
l
i
m
ag
e
w
at
erm
ar
k
i
ng
f
o
r
m
ed
i
c
al
i
m
ag
es
;
the
n
s
ec
ti
on
3
e
x
pl
ai
ns
ab
o
ut
t
he
pro
po
s
e
d
m
eth
od
;
f
ol
l
o
wed
b
y
s
e
c
ti
on
4
whi
c
h
de
s
c
r
i
be
s
th
e
m
eth
od
o
l
o
g
y
us
ed
;
s
ec
t
i
on
5
i
l
l
us
tr
at
es
the
ex
am
i
na
ti
o
n
r
es
u
l
t
of
ou
r
pro
po
s
ed
m
eth
od
; a
nd
f
i
na
l
l
y
,
s
ec
ti
on
6
pres
en
ts
th
e c
o
nc
l
us
i
on
s
of
th
e s
tu
dy
.
F
i
gu
r
e
1.
N
um
be
r
of
C
T
/MRI
de
v
i
c
es
i
n O
E
CD
c
o
un
tr
i
es
i
n
19
8
0
-
20
1
4
2.
Dig
it
al
W
at
er
ma
r
king
f
o
r
M
edical
Image
s
Di
g
i
ta
l
w
at
erm
ar
k
i
ng
i
s
a
tec
hn
i
qu
e
f
or
em
be
dd
i
ng
da
ta
or
i
nf
orm
ati
on
c
al
l
e
d
‘
w
at
erm
ar
k
’
i
nto
a
m
ul
ti
m
ed
i
a
ob
j
ec
t
w
h
i
c
h
c
a
l
l
ed
th
e
‘
c
o
v
er
’
s
uc
h
as
i
m
ag
e,
v
i
de
o,
a
ud
i
o
or
ev
e
n
on
l
y
a
tex
t.
A
c
c
ordi
n
g
to
i
ts
f
un
c
ti
o
na
l
i
t
i
es
,
d
i
g
i
t
al
w
at
erm
ar
k
i
ng
c
an
be
c
l
a
s
s
i
f
i
ed
i
nt
o
t
w
o
c
ate
go
r
i
es
:
r
ob
us
t
an
d
f
r
ag
i
l
e
wate
r
m
ar
k
s
.
Robu
s
t
w
at
erm
ar
k
s
are
go
od
f
or
au
t
h
en
ti
c
a
ti
o
n
a
nd
c
ha
r
ac
teri
z
i
n
g
r
es
i
s
ta
nt
to
a
c
om
m
on
s
i
gn
a
l
proc
es
s
i
ng
s
uc
h
as
c
om
pres
s
i
on
.
O
n
the
ot
he
r
ha
nd
,
a
f
r
ag
i
l
e
wate
r
m
ar
k
wi
l
l
no
t
s
urv
i
v
e
f
r
om
s
i
gn
al
proc
e
s
s
i
ng
,
ho
we
v
er
i
t
i
s
g
oo
d
to
be
us
ed
f
or
da
ta
i
nt
eg
r
i
t
y
[1
2].
T
he
r
ef
ore,
thi
s
s
tud
y
de
v
e
l
o
ps
a
f
r
ag
i
l
e
wate
r
m
ar
k
i
ng
to
en
s
uri
ng
the
i
nte
grit
y
of
the
m
ed
i
c
al
i
m
ag
e
w
h
i
c
h
al
s
o
f
ol
l
o
w
e
d
b
y
en
s
urin
g
i
ts
a
u
the
nt
i
c
i
t
y
a
nd
c
on
f
i
de
nt
i
a
l
i
t
y
.
Mo
r
eo
v
er,
i
n
thi
s
s
tu
d
y
,
Le
as
t
S
i
gn
i
f
i
c
an
c
e
B
i
t
(
LS
B
)
m
od
i
f
i
c
ati
on
,
a
s
p
ati
al
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
M
NIK
A
V
ol
.
17
,
No
.
5,
O
c
tob
er 20
19
:
26
8
4
-
26
96
2686
do
m
ai
n
w
a
term
ar
k
i
ng
s
c
he
m
e
i
s
i
m
pl
em
en
ted
to
d
e
v
e
l
op
a
f
r
ag
i
l
e
wate
r
m
ar
k
i
ng
f
or
m
ed
i
c
al
i
m
ag
es
.
W
he
r
ea
s
,
the
LS
B
m
od
i
f
i
c
ati
on
i
m
ag
e
wat
erm
ar
k
i
ng
w
ork
s
b
y
c
ha
n
gi
n
g
the
l
ea
s
t
s
i
gn
i
f
i
c
an
t
b
i
t
(
L
S
B
)
of
th
e
c
ov
er
i
m
ag
e
w
i
t
h
th
e
wate
r
m
ar
k
bi
ts
[13
].
E
m
be
d
the
wate
r
m
ar
k
i
nto
the
l
as
t
t
w
o
s
i
g
ni
f
i
c
an
t
b
i
ts
f
or ex
am
pl
e (2
nd
a
nd
1
st
LS
B
of
c
o
v
er
i
m
ag
e) f
or eac
h
pi
x
el
v
al
ue
,
th
e
wate
r
m
ar
k
m
o
s
tl
y
no
t
go
i
ng
to
de
tec
t
ab
l
e
b
y
hu
m
an
e
y
es
[14
].
F
or
ex
am
pl
e,
i
f
the
i
m
ag
e
pi
x
el
i
s
18
0
whi
c
h
ha
s
b
i
na
r
y
v
al
ue
of
10
11
0
10
0
an
d
th
e
w
at
e
r
m
ar
k
bi
ts
0,
the
v
a
l
u
e
of
th
e
pi
x
e
l
wi
l
l
be
the
s
am
e,
18
0
.
S
am
e
pi
x
el
v
a
l
u
e
wi
th
wate
r
m
ar
k
bi
ts
1,
th
e
p
i
x
el
v
al
ue
wi
l
l
be
10
11
01
01
whi
c
h
i
s
18
1
i
n
d
ec
i
m
al
.
T
hu
s
,
hu
m
an
e
y
es
ar
e
l
es
s
s
en
s
i
t
i
v
e
w
i
t
h
t
w
o
c
o
l
ors
tha
t
ha
v
e
v
a
l
ue
18
0
an
d
18
1
i
n
gra
y
c
ol
ors
b
et
w
e
en
bl
ac
k
i
s
0
an
d
w
h
i
te
i
s
2
55
[
15
].
Di
g
i
ta
l
w
ate
r
m
ar
k
tec
hn
i
qu
es
ha
v
e
be
e
n
prop
os
ed
f
or
s
ec
urin
g
m
ed
i
c
al
i
m
ag
es
.
[16
]
propos
e
d
a
r
ob
us
t
an
d
i
m
pe
r
c
ep
ti
b
l
e
d
ua
l
w
ate
r
m
ar
k
i
ng
.
In
th
ei
r
prop
os
ed
m
eth
od
h
y
b
r
i
d
err
or
c
orr
ec
ti
ng
c
od
es
(
c
om
bi
na
t
i
on
of
B
CH
&
r
ep
et
i
t
i
on
c
o
de
)
are
ut
i
l
i
z
ed
to
e
nc
od
e
t
he
w
at
erm
ar
k
da
ta
be
f
ore
em
be
dd
i
ng
i
t
i
nto
th
e
c
o
v
er
i
m
ag
e.
S
i
nc
e
the
y
us
e
an
oth
er
i
m
ag
e
as
the
c
o
v
er
i
m
ag
e,
the
i
r
m
eth
od
r
eq
ui
r
es
c
ov
er
i
m
ag
e
w
h
i
c
h
c
an
ho
l
d
a
l
l
the
w
at
erm
ar
k
da
ta
(
the
m
ed
i
c
a
l
i
m
ag
e).
T
he
[17
,
18
]
prop
os
ed
m
eth
od
s
wi
th
th
e
i
d
e
a
of
c
o
m
bi
ni
ng
d
i
g
i
ta
l
w
at
erm
ar
k
i
ng
w
i
th
c
o
m
pres
s
i
on
or/and
c
r
y
p
tog
r
ap
h
y
a
l
g
orit
hm
.
B
oth
m
eth
od
s
uti
l
i
z
e
th
e
e
nc
r
y
pti
on
a
nd
c
o
m
pres
s
i
on
on
w
at
erm
ar
k
da
ta
(
E
P
R
a
nd
bi
ts
of
RO
I)
an
d
do
the
proc
es
s
be
f
o
r
e
i
ns
ert
i
ng
i
t
i
nto
LS
B
of
the
c
o
v
er
i
m
ag
e.
T
he
di
f
f
erenc
e
be
t
wee
n
the
m
i
s
[17
]’
s
m
eth
od
u
s
e
Regi
on
of
Int
eres
t
(
RO
I)
as
the
c
o
v
er
i
m
ag
e,
thu
s
m
a
y
c
au
s
e
de
gradat
i
o
n
o
n
RO
I’
s
qu
a
l
i
t
y
.
[18
]
’
s
m
eth
od
on
t
he
oth
er
h
an
d,
us
e
Regi
on
of
Non
-
I
nte
r
es
t
(
RO
NI)
as
th
e
c
o
v
er
i
m
ag
e
w
h
i
c
h
m
ea
n
s
the
RO
NI
pa
r
t s
h
ou
l
d b
e l
arge e
no
u
gh
t
o h
o
l
d
al
l
t
he
wate
r
m
ar
k
.
Ins
tea
d
of
j
us
t
hi
di
n
g
E
P
R
or
i
m
ag
e’
s
bi
ts
i
nto
c
o
v
er
i
m
ag
e
(
s
am
e
i
m
ag
e
or
oth
er
un
r
el
ate
d
i
m
ag
e),
r
es
ea
r
c
he
r
s
al
s
o
de
v
e
l
op
m
eth
od
f
or
tam
pe
r
de
tec
ti
on
an
d
l
oc
a
l
i
z
at
i
on
to
k
no
w
w
he
t
he
r
the
i
m
ag
e
ha
s
be
e
n
tam
pe
r
ed
an
d
i
ts
l
oc
ati
on
.
[1
9
,
20
]
propo
s
ed
a
tam
pe
r
de
tec
t
i
on
an
d
l
oc
a
l
i
z
at
i
on
o
f
i
m
ag
e
ba
s
ed
o
n
w
at
erm
a
r
k
i
ng
bl
oc
k
.
How
e
v
er,
t
he
i
r
m
eth
od
s
on
l
y
f
oc
us
w
he
t
he
r
the
i
m
ag
e
h
as
be
en
t
am
pe
r
ed
or
no
t,
an
d
th
e
y
i
g
no
r
e
t
he
DICO
M
tag
s
s
uc
h
as
E
P
R
,
i
m
ag
e
prop
erti
es
,
s
tud
y
an
d
s
er
i
es
proper
ti
es
,
etc
.
[
21
]
pro
po
s
ed
a
du
a
l
-
l
a
y
er
wate
r
m
ar
k
i
ng
f
or
m
ed
i
c
al
i
m
ag
es
.
T
he
m
eth
od
em
be
d
s
th
e
E
P
R
da
ta
i
nt
o 2
nd
LS
B
of
RO
NI
pa
r
t
an
d
bl
oc
k
-
ba
s
ed
i
m
ag
e
o
n
1
st
L
S
B
of
a
m
ed
i
c
al
i
m
ag
e
f
or
tam
pe
r
de
tec
ti
on
an
d
l
oc
a
l
i
z
a
ti
o
n.
O
the
r
pro
po
s
ed
m
eth
od
b
y
[6]
do
th
e
c
om
bi
ni
ng
of
c
r
y
p
t
og
r
ap
h
y
a
nd
wate
r
m
ar
k
i
ng
t
o
prov
i
d
e
c
on
f
i
de
nt
i
a
l
i
t
y
,
au
t
he
nti
c
i
t
y
an
d
i
n
teg
r
i
t
y
of
th
e
m
ed
i
c
a
l
i
m
ag
e.
B
oth
[12
]’
s
an
d
[2
1]’
s
m
eth
od
s
em
be
d
the
E
P
R
d
ata
i
nt
o
RO
NI
pa
r
t
of
the
i
m
ag
es
,
whi
c
h
m
ea
ns
the
i
r
m
eth
od
r
e
al
l
y
de
p
en
d
on
the
s
i
z
e
of
th
e RO
NI
w
ho
c
an
h
ol
d
al
l
wate
r
m
ar
k
bi
ts
.
T
he
s
tud
y
pres
e
nte
d
i
n
th
i
s
pa
pe
r
prop
os
es
a
tec
h
ni
q
ue
to
s
ec
ure
m
ed
i
c
al
i
m
ag
es
whi
c
h
ba
s
e
d
on
d
ua
l
-
l
a
y
e
r
fr
ag
i
l
e
d
i
g
i
ta
l
w
at
erm
ar
k
i
ng
.
T
he
propos
ed
t
ec
hn
i
q
ue
us
es
ha
s
h
v
a
l
ue
o
f
DICO
M
t
ag
s
t
o
e
ns
ure
t
he
i
nt
eg
r
i
t
y
of
th
e
DICO
M
t
ag
s
,
en
c
r
y
p
ts
t
he
E
P
R
to
pro
v
i
de
c
on
f
i
de
nt
i
a
l
i
t
y
an
d
au
th
en
t
i
c
i
t
y
,
an
d
the
n
em
be
ds
i
t
i
n
th
e
2
nd
L
S
B
of
th
e
m
ed
i
c
al
i
m
ag
e
an
d
l
as
tl
y
c
al
c
u
l
at
es
the
ha
s
h
v
al
u
e
of
8x
8
no
n
-
ov
erla
pp
e
d
bl
oc
k
,
an
d
c
r
ea
tes
an
i
d
f
or
ea
c
h
bl
oc
k
whi
c
h
t
he
n
em
be
ds
i
t
i
nt
o
1
st
L
S
B
to
pro
v
i
d
e
i
nte
gri
t
y
of
th
e
m
ed
i
c
al
i
m
ag
e.
T
he
propos
ed
tec
hn
i
qu
e
h
as
b
ee
n
tes
t
ed
on
CT
a
nd
MRI
m
ed
i
c
al
i
m
ag
es
of
he
art
d
i
s
ea
s
es
,
s
tr
ok
es
,
c
erv
i
c
al
c
an
c
ers
an
d
breas
t
c
an
c
ers
w
h
i
c
h
are
the
d
ea
d
l
i
es
t
di
s
ea
s
e
i
n
Ind
on
es
i
a.
T
he
de
tai
l
of
propos
e
d t
ec
h
ni
q
ue
i
s
di
s
c
us
s
ed
i
n
s
ec
ti
o
n 3
.
3.
T
h
e P
r
o
p
o
se
d
S
e
curit
y T
ec
h
n
iqu
e
T
he
r
e
are
t
wo
c
ore
proc
es
s
es
tha
t
ne
ed
to
be
pe
r
f
or
m
ed
i
n
ou
r
pr
op
os
e
d
tec
h
ni
q
ue
,
i
.e.
wate
r
m
ar
k
e
m
be
dd
i
n
g
pr
oc
es
s
an
d
wate
r
m
ar
k
ex
tr
ac
ti
on
proc
es
s
.
T
hi
s
s
ec
ti
on
di
s
c
us
s
es
the
proc
es
s
es
i
n d
eta
i
l
s
.
3.1.
W
ate
r
ma
r
k
E
mb
edd
in
g
P
r
o
c
es
s
T
he
wate
r
m
ar
k
e
m
be
dd
i
n
g
proc
es
s
m
ai
nl
y
ha
s
t
wo
m
ai
n
s
ub
-
proc
es
s
es
,
whi
c
h
are
the
wat
erm
ar
k
ge
ne
r
ati
on
an
d
w
at
erm
ar
k
i
ns
erti
on
.
T
he
f
l
o
w
of
t
he
proc
es
s
i
s
s
um
m
a
r
i
z
ed
i
n
a
di
a
gram
w
hi
c
h
pres
e
nte
d
i
n
Fig
ur
e 2
.
In
ou
r
s
tu
d
y
,
t
w
o
wate
r
m
ar
k
s
are us
ed
an
d e
m
be
dd
ed
i
nto
i
ts
c
orr
es
po
nd
i
ng
m
ed
i
c
a
l
i
m
ag
e:
c
on
c
ate
na
t
i
o
n
of
E
P
R
an
d
h
as
h
of
DICO
M
D
at
a
S
et
(
DI
CO
M
T
ag
s
)
an
d
t
he
c
o
nc
ate
na
t
i
on
of
i
d
an
d
h
as
h
v
a
l
u
e
of
ea
c
h
no
n
-
o
v
er
l
ap
pe
d
b
l
oc
k
.
Her
e
i
s
m
ore
de
ta
i
l
ab
o
ut
t
he
wate
r
m
ar
k
ge
ne
r
ati
on
:
a.
S
ep
arate
the
E
P
R
ta
gs
a
nd
i
m
pl
em
en
t
A
E
S
25
6
en
c
r
y
p
ti
o
n
a
l
g
orit
hm
to
en
c
r
y
p
t
(
E
nc
Ct)
us
ed
to
gu
arant
ee
t
he
c
on
f
i
de
nt
i
a
l
i
t
y
a
nd
au
th
en
t
i
c
i
t
y
o
f
th
e E
P
R.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
A
dv
an
c
e
d wa
termark
i
ng
t
e
c
hn
i
q
ue
t
o i
mp
r
ov
e
me
d
i
c
al
i
ma
ge
s
’
…
(
Me
di
a A
nu
ge
r
ah
A
y
u
)
2687
b.
Cal
c
u
l
ate
t
he
h
as
h
of
DIC
O
M
tag
s
(
Has
hDt
whi
c
h
i
s
us
ed
to
g
ua
r
an
tee
t
he
i
nt
eg
r
i
t
y
of
i
t
us
i
ng
S
H
A
2
56
a
l
g
orit
hm
.
c.
Conc
ate
na
t
i
on
of
th
e H
as
h
Dt an
d E
nc
Ct to
be
c
om
e t
h
e 1
st
wate
r
m
ar
k
: Con
DC.
d.
A
f
ter
em
be
dd
i
n
g
th
e
Co
nD
C
i
nt
o
2
nd
L
S
B
of
the
c
o
v
er
i
m
ag
e,
prep
are
8x
8
no
n
-
o
v
erl
ap
p
ed
bl
oc
k
s
i
m
ag
e
c
on
c
ate
na
t
e
wi
th
t
he
r
o
w
&
c
o
l
um
n
as
an
i
d
of
ea
c
h
no
n
-
o
v
er
l
ap
p
ed
b
l
oc
k
s
.
T
he
n
c
al
c
ul
ate
th
e
h
as
h
of
i
t
us
i
n
g
S
H
A
25
6
a
l
g
orit
hm
.
T
hi
s
w
at
erm
ar
k
da
ta
ac
t
as
tam
pe
r
de
tec
tor
an
d
ab
l
e t
o s
ho
w t
he
ta
m
pe
r
’
s
l
oc
a
ti
o
n.
F
i
gu
r
e
2.
W
ate
r
m
ar
k
ge
ne
r
ati
o
n a
nd
i
ns
erti
on
of
th
e p
r
op
os
ed
te
c
hn
i
q
ue
T
he
ne
x
t
s
tep
i
s
wate
r
m
ar
k
i
ns
erti
o
n
(
em
be
dd
i
n
g)
pro
c
es
s
.
E
m
be
d
the
Co
nDC
i
nto
2
nd
LS
B
an
d
c
o
nc
ate
na
t
i
on
of
i
d
an
d
h
as
h
of
ea
c
h
bl
oc
k
i
nto
i
ts
c
orr
es
po
n
di
ng
b
l
oc
k
s
.
T
he
de
tai
l
of
i
ns
erti
on
m
eth
od
di
s
c
us
s
es
i
n t
he
f
ol
l
o
wi
ng
m
an
ne
r
:
a.
E
x
pa
n
d t
h
e l
en
g
th
of
Co
nD
C to
f
i
t th
e s
i
z
e
of
th
e c
o
v
er
i
m
ag
e
b.
T
o
i
nc
r
ea
s
e
the
s
ec
urit
y
,
i
ns
ert
r
an
do
m
l
y
ea
c
h
b
i
t
of
ConDC
i
nto
2
nd
LS
B
of
i
m
ag
e
us
i
ng
the
f
ol
l
o
wi
ng
eq
u
at
i
on
:
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
M
NIK
A
V
ol
.
17
,
No
.
5,
O
c
tob
er 20
19
:
26
8
4
-
26
96
2688
f
(
x
)
=
k
x
m
od
n
+
1
(
1)
whereas
,
k
i
s
th
e
s
ec
r
et
prim
e
k
e
y
i
n
w
h
i
c
h
k
∈
[1,
n
];
x
i
s
t
he
b
i
t
po
s
i
t
i
o
n
of
C
on
DC
wi
t
h
x
∈
[1,
l
e
ng
th
of
Con
DC]
;
an
d
n
i
s
t
he
tot
a
l
nu
m
be
r
of
pi
x
e
l
s
a
v
a
i
l
a
bl
e
f
or
wate
r
m
ar
k
em
be
dd
i
ng
.
c.
T
he
n
m
ov
e
f
or
s
ec
on
d
w
at
erm
ar
k
s
.
S
i
nc
e
W
e
us
e
S
HA
25
6
on
ea
c
h
b
l
oc
k
,
w
h
i
c
h
ge
ne
r
a
te
64
he
x
-
b
y
tes
c
h
arac
ters
,
t
he
n
W
e
ne
ed
to
c
h
oo
s
e
r
an
do
m
l
y
1
6
h
ex
-
b
y
t
es
c
ha
r
ac
ters
to
f
i
t
the
t
ota
l
b
i
ts
a
v
ai
l
a
bl
e
on
8
x
8 n
on
-
o
v
er
l
ap
pe
d
bl
oc
k
s
u
s
i
ng
f
orm
ul
a:
f
(
x
)
=
k
x
m
od
n
+
1
(
2)
whereas
,
k
i
s
the
s
ec
r
et
pri
m
e k
e
y
i
n
w
h
i
c
h
k
∈
[1
,
n]
;
x
i
s
th
e
c
h
arac
ters
po
s
i
ti
on
of
the
ha
s
h
r
es
ul
t
wi
th
x
∈
[
1,
1
6];
an
d
n i
s
th
e t
o
tal
nu
m
be
r
of
pi
x
el
s
a
v
a
i
l
a
bl
e f
or waterm
ar
k
em
be
dd
i
ng
.
d.
E
m
be
d
r
an
do
m
l
y
the
s
e
l
ec
t
ed
c
ha
r
ac
ters
i
n
to
1s
t
L
S
B
of
c
orr
es
po
nd
i
n
g
bl
oc
k
us
i
ng
f
ol
l
o
wi
n
g
f
or
m
ul
a:
f
(
x
)
=
k
x
m
od
n
+
1
(
3)
whereas
,
k
i
s
the
s
ec
r
et
prim
e
k
e
y
eq
ua
l
to
prev
i
o
us
k
us
ed
i
n
pr
ev
i
ou
s
s
te
p;
x
i
s
th
e
c
ha
r
ac
ters
po
s
i
t
i
on
of
the
c
on
c
ate
na
te
d
s
tr
i
n
g
i
n
bi
na
r
y
x
∈
[1,
6
4];
an
d
n
i
s
th
e
t
ota
l
nu
m
be
r
of
pi
x
el
s
a
v
a
i
l
ab
l
e f
or
w
ate
r
m
ar
k
e
m
be
dd
i
n
g.
3.2.
W
ate
r
ma
r
k
E
xtr
ac
t
io
n
P
r
o
c
es
s
T
he
m
ai
n
ob
j
ec
ti
v
e
of
wate
r
m
ar
k
i
ng
ex
tr
ac
ti
on
pr
oc
es
s
i
s
to
c
he
c
k
the
i
n
teg
r
i
t
y
,
c
on
f
i
de
nt
i
a
l
i
t
y
an
d
a
uth
en
t
i
c
i
t
y
of
th
e
i
m
ag
e.
Det
ai
l
s
o
n
th
i
s
proc
es
s
are
pres
en
t
ed
i
n
F
i
g
ure
3.
A
s
de
p
i
c
ted
i
n
F
i
g
ure
3,
t
h
ere
are
t
w
o
p
ha
s
es
to
c
he
c
k
the
da
ta
i
nte
grit
y
,
f
i
r
s
t
c
he
c
k
w
he
th
er
the
i
m
ag
e
h
as
be
e
n
tam
pe
r
ed
or
no
t,
th
en
c
he
c
k
w
h
et
he
r
the
DICO
M
T
ag
s
ha
s
b
ee
n
t
am
pe
r
ed
or not.
Her
e
i
s
m
ore de
tai
l
ab
ou
t th
e
w
at
erm
ar
k
ex
tr
ac
ti
on
:
a.
Di
v
i
d
e t
h
e m
ed
i
c
al
i
m
ag
e i
n
to
8x
8
no
n
-
o
v
er
l
ap
pe
d
bl
oc
k
s
an
d e
x
tr
ac
t 1
s
t
LS
B
of
e
ac
h b
l
oc
k
.
b.
Cal
c
u
l
ate
ha
s
h
of
ea
c
h
b
l
oc
k
the
n
c
o
m
pa
r
e
the
c
al
c
ul
at
i
on
r
es
u
l
t
wi
th
t
he
ex
t
r
ac
ted
ha
s
h
v
a
l
ue
.
If
the
v
al
ue
i
s
no
t
the
s
am
e,
thi
s
m
e
an
s
the
i
m
ag
e
ha
s
be
en
al
t
ered.
T
he
no
n
-
o
v
er
l
ap
pe
d
b
l
oc
k
s
c
an
s
h
o
w
the
l
oc
ati
on
of
ta
m
pe
r
ed
i
m
ag
e.
O
the
r
wi
s
e,
c
on
ti
n
ue
to
the
n
ex
t s
tep
s
.
c.
E
x
tr
ac
t
th
e
2
nd
L
S
B
of
the
c
ov
er
i
m
ag
e
to
ge
t
the
c
on
c
ate
na
t
ed
s
tr
i
ng
of
DICO
M
T
ag
s
ha
s
h
v
a
l
ue
an
d
th
e
en
c
r
y
p
te
d E
P
R.
d.
Cal
c
u
l
ate
the
ha
s
h
of
the
DICO
M
ta
gs
the
n
c
om
pa
r
e
w
i
t
h
th
e
ex
tr
ac
ted
D
I
CO
M
tag
s
.
If
the
v
al
ue
s
are
s
am
e,
th
en
c
on
ti
n
ue
t
o
the
ne
x
t
s
tep
s
,
ot
he
r
wi
s
e
th
e
D
ICO
M
tag
s
ha
v
e
be
en
al
tered
.
e.
La
s
tl
y
, a
f
ter c
he
c
k
i
ng
t
he
m
ed
i
c
a
l
i
m
ag
e
a
nd
t
he
D
ICO
M t
ag
s
,
de
c
r
y
p
t th
e e
nc
r
y
pt
ed
E
P
R.
4.
M
eth
o
d
o
log
y
T
he
s
c
he
m
ati
c
f
l
ow
of
thi
s
s
tud
y
i
s
f
i
r
s
tl
y
b
y
r
ev
i
e
wi
n
g
prev
i
ou
s
w
ork
s
an
d
i
de
n
ti
f
y
i
n
g
the
s
tu
d
y
po
s
i
t
i
o
n
a
nd
c
o
n
tr
i
bu
t
i
on
to
the
top
i
c
of
s
ec
urin
g
m
ed
i
c
al
i
m
ag
es
.
T
he
n,
de
s
i
gn
an
d
i
m
pl
em
en
t
ati
on
of
th
e
prop
os
ed
t
ec
hn
i
qu
e,
f
ol
l
o
wed
b
y
c
o
l
l
ec
ti
n
g
the
s
tud
y
m
ate
r
i
al
s
an
d
d
ata
c
ol
l
ec
t
i
on
. L
as
t
l
y
,
an
al
y
z
e t
he
d
ata
an
d
dra
w
a c
on
c
l
us
i
on
.
4.1.
S
o
u
r
c
e of
M
ateri
als
Me
d
i
c
al
i
m
ag
es
whi
c
h
r
el
at
ed
wi
th
f
ou
r
de
a
dl
i
es
t
di
s
e
as
es
i
n
I
nd
o
ne
s
i
a
ar
e
us
e
d
i
n
t
hi
s
s
tud
y
to
t
es
t
the
pro
po
s
e
d
wate
r
m
ar
k
i
ng
tec
hn
i
qu
e
.
T
he
an
on
y
m
i
z
ed
m
ed
i
c
al
i
m
ag
es
are
tak
en
f
r
o
m
op
en
w
eb
s
i
te.
T
he
i
m
ag
e
f
or
m
at
f
r
o
m
o
f
the
m
ed
i
c
al
i
m
ag
e
orig
i
n
al
l
y
i
s
i
n
“
dc
m
”
f
or
m
at
(
DICO
M
f
orm
at).
T
he
s
i
z
e
of
m
ed
i
c
al
i
m
ag
e
of
CT
s
c
an
i
s
ab
o
ut
51
2x
5
12
-
p
i
x
el
m
atri
x
an
d
the
MRI
i
s
ab
o
ut
25
6x
2
56
-
pi
x
el
i
m
ag
e
wi
th
1
6
a
l
l
oc
ate
d
b
i
ts
a
nd
the
i
m
ag
e
i
n
m
on
oc
hrom
e
m
od
e.
T
ho
s
e s
i
z
es
are t
he
c
o
m
m
on
s
i
z
e o
f
m
ed
i
c
al
i
m
ag
e f
or CT
s
c
an
an
d
MRI
m
od
al
i
ti
es
[2
2]
:
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
A
dv
an
c
e
d wa
termark
i
ng
t
e
c
hn
i
q
ue
t
o i
mp
r
ov
e
me
d
i
c
al
i
ma
ge
s
’
…
(
Me
di
a A
nu
ge
r
ah
A
y
u
)
2689
F
i
gu
r
e
3.
W
ate
r
m
ar
k
ex
tr
ac
ti
on
of
th
e
propos
e
d t
ec
h
ni
qu
e
4.2.
P
e
r
f
o
r
man
c
e
A
n
al
ys
i
s on
Image Q
u
ali
t
y
(
Imp
e
r
ce
p
t
ibilit
y
A
n
al
y
s
is)
In
t
hi
s
r
es
e
arc
h
wor
k
,
the
p
r
op
os
ed
s
c
he
m
ati
c
us
e
of
LS
B
wate
r
m
ar
k
i
ng
tec
h
ni
q
ue
i
s
to
em
be
d
the
w
ate
r
m
ar
k
to
the
c
ov
er
i
m
ag
e.
A
s
a
r
es
u
l
t,
an
i
m
ag
e
qu
a
l
i
t
y
d
eg
r
ad
ati
on
w
i
l
l
oc
c
ur.
T
he
go
o
d
qu
a
l
i
t
y
of
i
m
ag
e
wate
r
m
ar
k
i
ng
tec
h
ni
qu
e
i
s
tha
t
the
i
m
ag
e
c
an
ho
l
d
as
m
uc
h
da
ta
as
po
s
s
i
bl
e,
an
d
t
he
i
m
ag
e
q
ua
l
i
t
y
d
eg
r
a
da
t
i
on
l
ev
el
i
s
l
o
w
[
23
].
T
he
r
e
are
v
ari
ou
s
tec
hn
i
q
ue
s
t
o
ev
a
l
u
ate
the
d
i
f
f
erenc
e
b
et
w
ee
n
an
orig
i
n
al
i
m
ag
e
a
nd
a
wate
r
m
ar
k
ed
i
m
ag
e
whet
h
er
qu
a
l
i
t
ati
v
e
l
y
or
q
ua
nt
i
ta
ti
v
e
l
y
.
A
m
on
g
tho
s
e
tec
h
ni
q
ue
s
,
the
m
o
s
t
an
d
c
om
m
on
l
y
us
ed
c
r
i
ter
i
a
an
d
b
ec
om
e
the
s
tan
d
ard
ev
a
l
u
ati
on
of
i
m
ag
e
qu
al
i
t
y
m
ea
s
urem
en
ts
are
Me
an
S
qu
ared
E
r
r
or
(
MS
E
)
a
nd
t
he
P
ea
k
S
i
g
na
l
t
o
No
i
s
e
Rat
i
o
(
P
S
N
R)
f
or
qu
an
ti
t
ati
v
e
m
eth
od
an
d
S
tr
uc
tura
l
S
i
m
i
l
arit
y
I
nd
ex
Me
tr
i
x
(
S
S
I
M) f
or qua
l
i
tat
i
v
e m
eth
od.
T
he
go
al
of
MS
E
i
s
to
m
ea
s
ure
or
pro
v
i
de
qu
an
t
i
tat
i
v
e
s
c
ores
of
t
w
o
s
i
gn
al
s
/
i
m
ag
e
s
whi
c
h
d
es
c
r
i
be
ho
w
s
i
m
i
l
ar
the
s
i
gn
al
/
i
m
ag
e
to
ea
c
h
oth
er,
th
e
de
gree
of
err
or
or
di
s
torti
o
n
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
M
NIK
A
V
ol
.
17
,
No
.
5,
O
c
tob
er 20
19
:
26
8
4
-
26
96
2690
am
on
g
the
m
.
MS
E
c
an
be
ob
tai
ne
d
b
y
c
a
l
c
ul
ati
ng
th
e
av
er
ag
e
of
s
qu
ared
i
nt
e
n
s
i
t
y
of
t
w
o
or
m
ore (
i
np
uts
)
i
m
ag
es
as
s
ho
w
n
i
n t
he
f
orm
ul
a o
f
MS
E
be
l
o
w
:
=
1
∑
∑
(
,
−
,
′
)
2
−
1
=
0
−
1
=
0
(
4)
where
N
M
i
s
th
e
i
m
ag
e
s
i
z
e,
x
j
,k
i
nd
i
c
ate
s
t
he
j
k
-
th
pi
x
el
v
a
l
u
e
of
wate
r
m
ar
k
ed
i
m
ag
e
an
d
x
’
j
,k
i
nd
i
c
ate
s
the
j
k
-
th
pi
x
el
v
a
l
u
e
of
the
ori
gi
na
l
i
m
ag
e.
T
he
s
m
al
l
er
th
e
MS
E
v
a
l
ue
i
nd
i
c
ate
s
the
d
eg
r
a
da
t
i
on
l
e
v
e
l
of
th
e
wate
r
m
ar
k
ed
i
m
ag
e i
s
l
o
w.
P
ea
k
S
i
gn
al
t
o
Noi
s
e
R
ati
o
(
P
S
NR)
i
s
a
m
ath
em
ati
c
al
wa
y
to
m
ea
s
ure
the
i
m
ag
e
qu
a
l
i
t
y
ba
s
ed
on
th
e
pi
x
e
l
d
i
f
f
erent
b
et
w
e
en
the
t
wo
i
m
ag
es
:
th
e
or
i
g
i
na
l
i
m
ag
e
a
n
d
w
a
term
ar
k
e
d
i
m
ag
e.
MS
E
ne
ed
to
c
a
l
c
ul
at
ed
f
i
r
s
t
be
f
ore
c
al
c
ul
at
i
ng
t
he
P
S
NR
whi
c
h
c
a
n
be
c
al
c
u
l
at
ed
b
y
f
ol
l
o
wi
ng
f
or
m
ul
a:
=
10
10
2
(
5)
whereas
,
L
i
s
th
e
r
an
g
e
of
i
m
ag
e
pi
x
el
i
nte
ns
i
t
y
.
A
s
ex
am
pl
e,
f
or
i
m
ag
e
of
8
bi
t/p
i
x
el
of
gra
y
s
c
al
e
i
m
ag
e,
L=
2
8
-
1
=
25
5.
T
he
hi
g
he
r
v
a
l
ue
of
P
S
NR
i
nd
i
c
ate
s
t
he
be
t
ter
qu
al
i
t
y
of
wate
r
m
ar
k
ed
i
m
ag
e.
B
oth
MS
E
an
d
P
S
R
v
a
l
ue
s
are
s
i
m
pl
e
an
d
ea
s
y
t
o
c
al
c
ul
at
e,
ho
wev
er,
the
y
are
no
t
v
er
y
wel
l
m
atc
he
d
t
o
p
erc
ei
v
ed
v
i
s
ua
l
q
ua
l
i
t
y
.
F
or
th
at
r
e
as
on
,
i
n
20
04
,
S
tr
uc
t
ural
S
i
m
i
l
arit
y
Ind
ex
Me
tr
i
x
wer
e
prop
os
ed
b
y
[
24
].
S
S
IM
i
s
on
e
of
H
um
a
n
V
i
s
ua
l
S
y
s
t
em
(
HV
S
)
m
eth
o
d
t
o
c
he
c
k
i
m
ag
e
qu
a
l
i
t
y
an
d
S
S
IM
u
s
ed
to
c
om
pa
r
e
t
w
o
i
m
ag
es
qu
a
l
i
t
y
b
y
m
ea
s
urin
g
t
he
i
r
s
i
m
i
l
ari
t
y
.
T
hree
as
pe
c
t
s
are
c
al
c
ul
at
ed
to
d
ete
r
m
i
ne
the
s
i
m
i
l
ari
t
y
of
two
i
m
ag
es
:
l
um
i
na
nc
e,
c
on
tr
as
t
an
d
s
tr
uc
ture
of
th
e
i
m
ag
es
.
T
he
l
um
i
na
nc
e
f
un
c
ti
o
ns
l
(x
y
)
f
or
r
ef
erenc
e
i
m
ag
e
x
(
i
n
th
i
s
c
as
e
i
s
the
or
i
gi
na
l
i
m
ag
e) an
d t
es
t
i
m
ag
e
y
(
th
e
w
a
term
ar
k
ed
i
m
ag
e) i
s
:
(
,
)
=
2
+
1
2
+
2
+
1
(
6)
where μ
x
an
d
μ
y
are
the
m
ea
n
v
al
ue
s
of
x
an
d
y
,
an
d C
1 i
s
a
s
t
ab
i
l
i
z
i
ng
c
o
ns
tan
t.
T
he
s
tr
uc
ture c
om
pa
r
i
s
on
fu
nc
ti
o
n s
(x
y
)
of
S
S
IM
i
s
e
x
pres
s
ed
as
:
(
,
)
=
+
3
+
3
(
7)
where
σ
xy
i
s
the
c
orr
el
at
i
on
be
t
w
e
en
x
an
d
y
a
nd
C3
i
s
c
on
s
tan
t
s
tab
i
l
i
z
er.
T
he
n,
th
e
S
S
IM
i
nd
ex
c
an
be
w
r
i
tte
n a
s
:
(
,
)
=
[
(
,
)
]
.
[
(
,
)
]
.
[
(
,
)
]
(
8)
i
n [
24
],
α =
β
=
γ
=
1
an
d C
3 =
C2/2
. T
ho
s
e,
S
S
IM(x
,
y
)
f
i
na
l
l
y
c
a
n b
e
ex
pres
s
ed
a
s
:
(
,
)
=
(
2
+
1
)
(
2
+
2
)
(
2
+
2
+
1
)
(
2
+
2
+
2
)
(
9)
S
S
I
M
i
s
a
d
ec
i
m
al
v
al
u
e
be
t
w
e
en
-
1
to
1.
W
he
r
ea
s
,
-
1
f
or
no
n
-
i
de
nti
c
a
l
da
tas
et
an
d
the
o
pp
os
i
te,
1
i
f
al
l
pi
x
e
l
v
a
l
ue
s
are
i
d
en
t
i
c
al
.
I
n
th
i
s
s
tud
y
,
th
e
S
S
I
M
v
al
u
e
c
al
c
ul
ate
d
us
i
n
g
p
y
t
ho
n
l
i
brar
y
c
a
l
l
ed
“
s
c
i
k
i
t
-
i
m
ag
e” l
i
brar
y
.
4.3.
P
e
r
f
o
r
man
c
e
A
n
al
ys
i
s on
T
amp
er
Dete
ctio
n
(
F
r
agilit
y
A
n
al
ys
is)
S
e
v
era
l
m
al
i
c
i
ou
s
att
ac
k
s
s
uc
h
as
i
m
ag
e
c
r
op
pi
n
g,
c
op
y
-
p
as
te
att
ac
k
s
,
c
ol
l
ag
e
att
ac
k
an
d
c
o
ns
tan
t
-
a
v
er
ag
e
att
a
c
k
are
c
om
m
on
l
y
us
ed
t
o
t
he
f
r
ag
i
l
i
t
y
of
w
at
erm
ar
k
i
ng
s
c
he
m
e
[25
].
W
h
ereas
,
i
m
ag
e
c
r
op
pi
n
g
an
d
c
op
y
-
pa
s
te
a
tta
c
k
s
w
ere
do
ne
b
y
c
r
op
or
c
op
y
-
p
as
te
s
om
e
pa
r
t
of
the
w
at
erm
ar
k
ed
i
m
ag
e
r
an
do
m
l
y
.
Co
l
l
a
ge
at
ta
c
k
i
n
thi
s
s
tud
y
do
ne
b
y
c
o
p
y
i
n
g
t
he
au
th
en
t
i
c
ate
d
b
l
oc
k
of
w
at
erm
ar
k
ed
i
m
ag
e
i
nt
o
an
ot
he
r
p
art
of
the
i
m
ag
e.
Co
ns
tan
t
-
a
v
era
ge
att
ac
k
do
ne
b
y
m
od
i
f
y
i
n
g
t
he
p
art
of
t
he
i
m
ag
e
an
d
k
ee
p
the
av
erag
e
i
nte
ns
i
t
y
of
the
i
m
ag
e
the
s
am
e
wi
th
the
ori
gi
na
l
on
e.
F
r
a
gi
l
e
di
gi
t
al
wate
r
m
ark
i
ng
i
n
whi
c
h
de
pe
n
d
o
n
av
era
ge
i
m
ag
e
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
A
dv
an
c
e
d wa
termark
i
ng
t
e
c
hn
i
q
ue
t
o i
mp
r
ov
e
me
d
i
c
al
i
ma
ge
s
’
…
(
Me
di
a A
nu
ge
r
ah
A
y
u
)
2691
i
nte
ns
i
t
y
wi
l
l
n
ot
r
es
i
s
t
to
t
hi
s
k
i
nd
of
att
ac
k
.
T
he
s
e
att
ac
k
s
are
c
ho
s
en
i
n
ord
er
to
c
om
pa
r
e
the
pr
op
os
e
d
s
c
he
m
e
w
i
th
oth
er
f
r
ag
i
l
e
di
gi
t
al
wat
erm
ar
k
i
ng
w
h
i
c
h
pro
po
s
ed
b
y
[2
6]
a
nd
[
27
].
Cal
c
u
l
ate
d
tam
pe
r
ed
d
ete
c
ti
on
r
at
e
c
an
b
e
us
ed
to
e
v
al
u
ate
the
q
ua
l
i
t
y
of
the
ta
m
pe
r
de
tec
ti
o
n
m
eth
od
.
T
a
m
pe
r
ed
de
tec
t
i
on
r
at
e
c
al
c
u
l
at
es
the
n
um
be
r
of
ta
m
pe
r
ed
bl
oc
k
s
de
tec
ted
i
n
the
i
m
ag
e.
Hi
gh
er
r
at
e
of
ta
m
pe
r
ed
bl
oc
k
i
nd
i
c
ate
s
a
go
od
t
am
pe
r
de
tec
ti
o
n
m
eth
o
d.
Her
e
i
s
the
f
orm
ul
a t
o c
al
c
u
l
at
e t
h
e
ta
m
pe
r
ed
de
t
ec
ti
o
n rate
(
T
DR)
:
=
100%
(
10
)
5
. Re
sult
s
a
nd
D
isc
u
s
sio
n
T
hi
s
s
ec
ti
on
pres
e
nts
the
r
es
ul
ts
of
s
tud
y
c
on
d
uc
ted
a
nd
d
i
s
c
us
s
es
the
i
nte
r
pret
a
ti
on
of
the
r
es
u
l
ts
an
d i
ts
i
m
pl
i
c
ati
on
s
.
5.1.
M
edical
Im
ages
Dat
a
se
t
T
he
propos
ed
tec
hn
i
q
ue
was
de
v
e
l
o
pe
d
a
nd
i
m
pl
em
en
te
d
us
i
n
g
p
y
t
ho
n
progr
a
m
m
i
ng
l
an
gu
a
ge
.
I
t
w
as
t
he
n
tes
te
d
on
m
ed
i
c
a
l
i
m
ag
es
w
h
i
c
h
r
el
at
ed
t
o
th
e
f
ou
r
de
ad
l
i
e
s
t
di
s
ea
s
es
i
n
Ind
o
ne
s
i
a
.
F
i
gu
r
e
4
s
ho
w
s
al
l
m
ed
i
c
al
i
m
ag
es
da
tas
et
us
ed
i
n
th
i
s
r
es
ea
r
c
h
wor
k
w
i
th
de
t
ai
l
propert
i
es
s
tat
e
d i
n T
ab
l
e
1
.
T
ab
l
e 1
.
P
r
op
erti
es
of
Me
d
i
c
al
Im
ag
es
Data
s
et
No
N
a
m
e
M
o
d
a
li
t
y
S
iz
e
B
it
D
e
p
t
h
C
o
lor
Ty
p
e
1
B
r
e
a
s
t
_
M
R
I
M
R
I
2
5
6
x
2
5
6
16
M
o
n
o
c
h
r
o
m
e
2
C
e
r
v
i
x
_
M
R
I
M
R
I
5
1
2
x
5
1
2
16
M
o
n
o
c
h
r
o
m
e
3
B
r
a
in_
M
R
I
M
R
I
2
5
6
x
2
5
6
16
M
o
n
o
c
h
r
o
m
e
4
H
e
a
r
t
_
C
T
CT
5
1
2
x
5
1
2
16
M
o
n
o
c
h
r
o
m
e
5
B
r
e
a
s
t
_
C
T
CT
5
1
2
x
5
1
2
16
M
o
n
o
c
h
r
o
m
e
6
B
r
a
in_
C
T
CT
5
1
2
x
5
1
2
16
M
o
n
o
c
h
r
o
m
e
(
a)
(
b)
(
c
)
(
d)
(
e)
(
f
)
F
i
gu
r
e
4.
Im
ag
e d
ata
s
et:
(
a
)
breas
t (MRI)
, (b)
c
er
v
i
x
(
M
RI)
, (c
)
brai
n (M
RI)
,
(
d) hea
r
t (C
T
)
, (e)
breas
t (C
T
)
, (
f
)
brai
n (CT
)
S
i
nc
e
al
l
i
m
ag
es
are
tak
en
f
r
o
m
o
pe
n
w
e
bs
i
te
n
am
ed
as
htt
p:/
/
www.c
a
nc
erim
ag
i
ng
a
r
c
hi
v
e.
ne
t/
,
th
i
s
m
ea
ns
the
i
m
ag
es
are
an
on
y
m
i
z
ed
m
ed
i
c
al
i
m
ag
es
i
n
whi
c
h
p
oi
nt
t
he
pa
ti
e
nt
i
nf
or
m
ati
on
or
ot
he
r
c
o
nf
i
de
nti
a
l
tag
s
ha
s
be
e
n
r
em
ov
ed
or
c
ha
ng
e
d
wi
th
r
an
d
om
v
al
u
e
as
s
ho
wn
i
n
F
i
gu
r
e
5.
F
or
tha
t
r
e
as
on
,
b
ef
ore
us
i
ng
t
he
i
m
ag
es
f
or
f
urther
proc
es
s
,
r
ati
o
na
l
du
m
m
y
v
al
u
es
are
i
ns
erte
d
i
nto
r
em
ov
e
d
DICO
M
ta
gs
.
T
hi
s
pro
c
es
s
of
pu
tti
ng
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
M
NIK
A
V
ol
.
17
,
No
.
5,
O
c
tob
er 20
19
:
26
8
4
-
26
96
2692
du
m
m
y
v
al
ue
s
k
no
w
n
as
de
-
i
d
en
t
i
f
i
er
as
s
h
o
w
n
i
n
F
i
gu
r
e
6.
A
f
ter
th
e
de
-
i
de
nti
f
i
er
proc
es
s
,
a
DICO
M
f
i
l
e
w
i
t
h
c
om
pl
ete
D
ICO
M
t
ag
s
i
s
r
ea
d
y
to
us
e
f
or
tes
ti
ng
t
he
p
erf
or
m
an
c
e
of
the
pro
po
s
e
d t
ec
h
ni
q
ue
.
F
i
gu
r
e
5.
A
no
n
y
m
i
z
e
d DIC
O
M
tags
F
i
gu
r
e
6.
D
e
-
i
d
en
t
i
f
i
er DIC
O
M
tags
5.2.
P
e
r
f
o
r
man
c
e
A
n
al
ys
i
s on
Image Q
u
ali
t
y
(
Imp
e
r
ce
p
t
ibilit
y
A
n
al
y
s
is)
P
erc
ep
t
ua
l
i
m
pe
r
c
ep
ti
bi
l
i
t
y
i
s
o
ne
of
th
e
m
os
t
s
i
gn
i
f
i
c
an
t
i
n
de
x
es
i
n
th
e
wate
r
m
ark
i
ng
pe
r
f
or
m
an
c
e
an
a
l
y
s
i
s
.
T
hi
s
m
ea
ns
tha
t
h
um
an
e
y
es
s
ho
u
l
d
no
t
b
e
ab
l
e
to
de
t
ec
t
the
em
be
dd
ed
wate
r
m
ar
k
i
n
the
c
ov
er
i
m
ag
e.
F
i
gu
r
e
7
gi
v
es
t
he
wat
erm
ar
k
ed
i
m
ag
es
ge
n
erate
d
b
y
pro
po
s
ed
wate
r
m
ar
k
i
ng
s
c
he
m
es
on
m
ed
i
c
al
i
m
ag
es
.
W
e
c
a
n
s
ee
th
at
th
e
w
at
erm
ar
k
ed
i
m
ag
es
i
n
F
i
gu
r
e
7
are
a
l
m
os
t
the
s
a
m
e
as
the
orig
i
n
al
i
m
ag
e
i
n
F
i
g
ure
4.
It
s
ho
w
s
the
pro
po
s
ed
s
c
he
m
e
prov
i
de
s
a
s
at
i
s
f
ac
tor
y
wat
erm
ar
k
i
m
pe
r
c
ep
ti
bi
l
i
t
y
.
T
he
s
m
al
l
di
f
f
er
en
c
e
be
t
ween
the
i
m
ag
es
c
an
be
s
ee
n f
r
o
m
th
ei
r
M
S
E
, P
S
NR a
nd
S
S
IM
v
a
l
ue
s
as
d
es
c
r
i
be
s
i
n
T
ab
l
e 2
.
T
ab
l
e
2
s
ho
w
s
th
e
c
om
pa
r
i
s
on
of
L
S
B
s
ta
nd
ard
w
a
ter
m
ar
k
an
d
propos
e
d
wate
r
m
ark
i
ng
tec
hn
i
qu
e
i
n
t
erm
of
MS
E
,
P
S
NR
an
d
S
S
IM
v
al
u
es
.
S
t
an
d
ard
wate
r
m
ar
k
i
ng
i
n
th
i
s
s
tat
e
i
s
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
A
dv
an
c
e
d wa
termark
i
ng
t
e
c
hn
i
q
ue
t
o i
mp
r
ov
e
me
d
i
c
al
i
ma
ge
s
’
…
(
Me
di
a A
nu
ge
r
ah
A
y
u
)
2693
where
t
he
wate
r
m
ar
k
bi
t
e
m
be
dd
ed
i
nto
the
LS
B
of
c
ov
er
i
m
ag
e
wi
th
ou
t
r
an
d
o
m
i
z
at
i
on
an
d
the
propos
e
d
wate
r
m
ar
k
i
ng
tec
h
ni
q
ue
r
an
d
om
i
z
ed
t
he
i
ns
erti
on
of
the
wate
r
m
ar
k
bi
ts
.
B
oth
,
s
tan
da
r
d
an
d
prop
os
ed
w
a
t
erm
ar
k
i
ng
tec
hn
i
q
ue
s
ha
s
P
S
N
R
v
al
u
es
m
ore
t
ha
n
44
dB
,
a
nd
S
S
I
M
v
a
l
ue
s
a
l
m
os
t
1
w
h
ereas
1
i
s
the
m
ax
i
m
u
m
v
al
u
e
of
S
S
I
M)
the
s
e
pro
v
e
th
e
i
m
pe
r
c
ep
ti
b
i
l
i
t
y
of
the
w
at
erm
ar
k
ob
j
ec
ti
v
el
y
.
F
r
o
m
thi
s
tab
l
e
,
the
n
ge
n
erate
gra
ph
s
l
i
k
e
s
ho
w
n
i
n
F
i
gu
r
e
8
an
d
F
i
gu
r
e
9.
F
r
om
the
s
e
f
i
gu
r
e
s
,
P
S
NR
an
d
S
S
I
M
v
a
l
ue
s
o
f
propos
e
d
t
ec
hn
i
qu
e
c
on
s
i
s
ten
tl
y
s
ho
w
hi
g
he
r
v
a
l
u
es
tha
n
t
he
s
ta
nd
ard
w
a
term
ar
k
i
ng
.
T
hi
s
m
ea
ns
tha
t
r
an
do
m
i
z
ed
th
e
i
ns
erti
on
of
wate
r
m
ar
k
bi
ts
gi
v
es
h
i
g
he
r
v
a
l
ue
of
P
S
NR
a
nd
S
S
I
M
w
h
i
c
h
i
nd
i
c
ate
the
pro
p
os
ed
tec
hn
i
qu
e
produc
es
b
ett
er
wate
r
m
ar
k
ed
i
m
ag
e q
ua
l
i
t
y
c
om
pa
r
e wi
th
the
s
t
an
d
ard
w
a
term
ar
k
i
ng
te
c
hn
i
qu
e
.
(
a)
(
b)
(
c
)
(
d)
(
e)
(
f
)
F
i
gu
r
e
7.
W
ate
r
m
ar
k
ed
i
m
ag
e
: (a)
bre
as
t (MRI)
, (b)
c
er
v
i
x
(
MR
I)
, (c
)
brai
n (
MRI),
(
d) hea
r
t (C
T
)
, (e)
breas
t (C
T
)
, (
f
)
brai
n (CT
)
T
ab
l
e 2
.
W
ate
r
m
ar
k
ed
I
m
ag
es
Q
ua
l
i
t
y
C
om
pa
r
i
s
on
M
e
d
ic
a
l
I
m
a
g
e
s
S
t
a
n
d
a
r
d
W
a
t
e
r
m
a
r
k
ing
P
r
o
p
o
s
e
d
S
c
h
e
m
e
M
S
E
P
S
N
R
S
S
I
M
M
S
E
P
S
N
R
S
S
I
M
c
t
_
b
r
a
in
2
.
0
9
5
2
5
6
8
1
4
4
.
9
1
8
4
3
1
0
1
0
.
9
9
9
8
8
0
7
0
2
.
0
2
3
4
3
3
6
9
4
5
.
0
6
9
9
1
3
8
5
0
.
9
9
9
8
7
8
7
4
m
r
_
b
r
a
in
2
.
1
8
0
2
8
2
5
9
4
4
.
7
4
5
6
7
5
7
3
0
.
9
9
8
9
6
8
3
6
2
.
1
8
1
3
8
1
2
3
4
4
.
7
4
3
4
8
7
9
0
0
.
9
9
9
0
3
7
5
7
c
t
_
b
r
e
a
s
t
2
.
1
8
9
6
7
4
3
8
4
4
.
7
2
7
0
0
8
2
4
0
.
9
9
9
5
4
0
8
4
2
.
1
1
2
2
8
1
8
0
4
4
.
8
8
3
2
8
5
0
4
0
.
9
9
9
6
3
5
3
7
m
r
_
b
r
e
a
s
t
2
.
3
2
8
2
9
2
8
5
4
4
.
4
6
0
4
2
7
5
7
0
.
9
9
9
1
2
2
6
2
2
.
1
9
1
2
0
7
8
9
4
4
.
7
2
3
9
6
7
7
9
0
.
9
9
9
4
7
8
2
4
c
t
_
h
e
a
r
t
2
.
4
9
8
2
6
4
3
1
4
4
.
1
5
4
4
1
9
7
7
0
.
9
9
9
8
2
7
7
5
2
.
5
0
1
8
0
4
3
5
4
4
.
1
4
8
2
7
0
1
7
0
.
9
9
9
8
1
6
2
1
m
r
_
c
e
r
v
ix
2
.
4
3
0
2
0
6
3
0
4
4
.
2
7
4
3
7
2
1
9
0
.
9
9
9
3
5
9
0
2
2
.
3
9
3
4
4
7
8
8
4
4
.
3
4
0
5
6
3
8
7
0
.
9
9
9
3
7
4
6
2
F
i
gu
r
e
8.
P
S
N
R
c
om
pa
r
i
s
on
F
i
gu
r
e
9.
S
S
I
M
c
om
pa
r
i
s
on
Evaluation Warning : The document was created with Spire.PDF for Python.