T
E
L
KO
M
N
I
KA
T
e
lec
om
m
u
n
icat
ion
,
Com
p
u
t
i
n
g,
E
lec
t
r
on
ics
an
d
Cont
r
ol
Vol.
18
,
No.
3
,
J
une
2020
,
pp.
1650
~
1657
I
S
S
N:
1693
-
6930
,
a
c
c
r
e
dit
e
d
F
ir
s
t
G
r
a
de
by
Ke
me
nr
is
tekdikti
,
De
c
r
e
e
No:
21/E
/KP
T
/2018
DO
I
:
10.
12928/
T
E
L
KO
M
NI
KA
.
v18i3.
14900
1650
Jou
r
n
al
h
omepage
:
ht
tp:
//
jour
nal.
uad
.
ac
.
id/
index
.
php/T
E
L
K
OM
N
I
K
A
A
r
e
vi
e
w
on
r
e
gi
on
o
f
i
n
t
e
r
e
st
-
b
ase
d
h
yb
r
id
m
e
d
ic
al
i
m
ag
e
c
o
m
p
r
e
ssi
on
al
gor
ith
m
s
S
u
h
ail
a
Ab
Az
iz
1
,
S
u
r
ian
i
M
oh
d
S
a
m
2
,
Nor
li
z
a
M
oh
am
e
d
3
,
S
alwan
i
M
oh
d
Daud
4
,
S
it
i
Z
aleha
A
b
d
u
l
Ras
h
id
5
,
Haf
iza
Abas
6
,
M
u
h
am
m
ad
F
at
h
i
Yus
of
7,
Rud
z
i
d
at
u
l
Akm
am
Dz
iyau
d
d
i
n
8
1,
2,
3,
4
,
6,
7
,
8
Raza
k
Facu
l
t
y
o
f
T
ech
n
o
l
o
g
y
a
n
d
In
f
o
rma
t
i
c
s
,
U
n
i
v
er
s
i
t
i
T
e
k
n
o
l
o
g
i
Mal
a
y
s
i
a
,
Mal
ay
s
i
a
5
A
zman
H
a
s
h
i
m
In
t
ern
a
t
i
o
n
a
l
Bu
s
i
n
e
s
s
Sc
h
o
o
l
,
U
n
i
v
ers
i
t
i
T
e
k
n
o
l
o
g
i
Mal
a
y
s
i
a
,
Mal
ay
s
i
a
Ar
t
icle
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
ived
J
ul
28
,
2019
R
e
vis
e
d
J
a
n
19
,
2020
Ac
c
e
pted
F
e
b
25
,
2020
D
i
g
i
t
al
med
i
cal
i
mag
e
s
h
a
v
e
b
ec
o
me
a
v
i
t
a
l
res
o
u
rce
t
h
at
s
u
p
p
o
rt
s
d
eci
s
i
o
n
-
ma
k
i
n
g
a
n
d
t
rea
t
men
t
p
r
o
ced
u
res
i
n
h
eal
t
h
care
fac
i
l
i
t
i
es
.
T
h
e
me
d
i
ca
l
i
ma
g
e
co
n
s
u
mes
l
arg
e
s
i
ze
s
o
f
m
emo
r
y
,
an
d
t
h
e
s
i
ze
k
eep
s
o
n
g
ro
w
t
h
d
u
e
t
o
t
h
e
t
r
en
d
o
f
me
d
i
ca
l
i
ma
g
e
t
e
ch
n
o
l
o
g
y
.
T
h
e
t
ech
n
o
l
o
g
y
o
f
t
el
emed
i
ci
n
e
en
co
u
rag
e
s
t
h
e
med
i
ca
l
p
ract
i
t
i
o
n
er
t
o
s
h
are
t
h
e
med
i
cal
i
ma
g
e
t
o
s
u
p
p
o
r
t
k
n
o
w
l
ed
g
e
s
h
ar
i
n
g
t
o
d
i
ag
n
o
s
e
an
d
an
al
y
s
e
t
h
e
i
mag
e.
T
h
e
h
ea
l
t
h
care
s
y
s
t
em
n
ee
d
s
t
o
en
s
u
re
d
i
s
t
ri
b
u
t
es
t
h
e
med
i
cal
i
ma
g
e
accu
rat
e
l
y
w
i
t
h
zero
l
o
s
s
o
f
i
n
f
o
rmat
i
o
n
,
fas
t
an
d
s
ecu
re.
Imag
e
co
m
p
res
s
i
o
n
i
s
b
e
n
efi
c
i
al
i
n
en
s
u
r
i
n
g
t
h
a
t
ach
i
ev
e
t
h
e
g
o
a
l
o
f
s
h
ari
n
g
t
h
i
s
d
at
a.
T
h
e
re
g
i
o
n
o
f
i
n
t
eres
t
-
b
a
s
ed
h
y
b
r
i
d
med
i
cal
co
m
p
res
s
i
o
n
al
g
o
r
i
t
h
m
p
l
a
y
s
t
h
e
p
art
s
t
o
red
u
ce
t
h
e
i
ma
g
e
s
i
ze
an
d
s
h
o
r
t
e
n
t
h
e
t
i
me
o
f
med
i
cal
i
mag
e
co
mp
re
s
s
i
o
n
p
ro
ce
s
s
.
V
ari
o
u
s
s
t
u
d
i
e
s
h
av
e
en
h
an
ce
d
b
y
co
mb
i
n
i
n
g
n
u
mer
o
u
s
t
ech
n
i
q
u
e
s
t
o
g
e
t
an
i
d
ea
l
res
u
l
t
.
T
h
i
s
p
a
p
er
rev
i
ew
s
t
h
e
p
rev
i
o
u
s
w
o
r
k
s
c
o
n
d
u
c
t
ed
o
n
a
reg
i
o
n
o
f
i
n
t
eres
t
-
b
a
s
ed
h
y
b
r
i
d
me
d
i
ca
l
i
mag
e
c
o
mp
re
s
s
i
o
n
al
g
o
r
i
t
h
ms
.
K
e
y
w
o
r
d
s
:
Hybr
id
M
e
dica
l
im
a
ge
c
ompr
e
s
s
ion
R
e
gion
of
int
e
r
e
s
t
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
ce
n
s
e
.
C
or
r
e
s
pon
din
g
A
u
th
or
:
S
ur
iani
M
ohd
S
a
m
,
R
a
z
a
k
F
a
c
ult
y
of
T
e
c
hnology
a
nd
I
n
f
or
matics
,
Unive
r
s
it
i
T
e
knologi
M
a
lays
ia,
Kua
la
L
umpur
,
M
a
lays
ia.
E
mail:
s
ur
iani
.
kl@ut
m.
my
1.
I
NT
RODU
C
T
I
ON
T
he
medic
a
l
im
a
ge
s
c
ontain
a
la
r
ge
a
mount
o
f
in
f
or
mation
that
is
ne
c
e
s
s
a
r
y
f
or
c
or
r
e
c
t
diagnos
is
by
doc
tor
s
a
nd
medic
a
l
pr
of
e
s
s
ionals
.
T
he
medic
a
l
im
a
ge
plays
e
s
s
e
nti
a
l
r
oles
to
a
s
s
i
s
t
the
medic
a
l
pr
a
c
ti
ti
one
r
in
e
a
s
ing
their
wor
k.
I
t
is
a
ls
o
the
mos
t
c
r
u
c
ial
a
s
s
e
t
of
the
he
a
lt
hc
a
r
e
s
ys
tem,
whic
h
is
us
e
d
c
omm
only
by
medic
a
l
pr
a
c
ti
ti
one
r
s
to
de
tec
t,
diagnos
e
,
tr
e
a
t,
a
nd
s
tudy
va
r
ious
types
of
dis
e
a
s
e
s
in
e
lec
tr
onic
f
or
m
.
I
t
is
a
ls
o
the
be
s
t
pr
a
c
ti
c
e
s
f
or
moni
tor
ing
the
pe
r
s
on’
s
he
a
lt
h
c
ondit
ion,
whic
h
r
e
mai
ns
us
e
d
wide
ly
nowa
da
ys
.
Ge
ne
r
a
ll
y,
e
ve
r
yone
knows
that
the
medic
a
l
im
a
ge
c
ons
umes
e
nor
mous
pr
opor
ti
ons
of
memo
r
y.
T
he
r
e
a
r
e
va
r
ious
type
of
modalit
ies
s
uc
h
a
s
X
-
R
a
y,
magne
ti
c
r
e
s
ona
nc
e
im
a
ging
(
M
R
I
)
,
s
ingl
e
-
photon
e
mi
s
s
ion
c
omput
e
d
tom
ogr
a
phy
(
S
P
E
C
T
)
,
c
om
puted
tom
ogr
a
phy
(
C
T
)
,
pos
it
r
on
e
mi
s
s
ion
tom
ogr
a
phy
(
P
E
T
)
a
nd
ult
r
a
s
ound
im
a
ging.
E
a
c
h
modal
it
y
whic
h
c
ons
umes
ve
r
y
lar
ge
of
memor
y
s
ize
a
nd
r
e
quir
e
s
a
mas
s
ive
volum
e
of
memor
y
s
tor
a
ge
s
pa
c
e
.
T
he
a
dmi
n
is
tr
a
tor
o
f
the
hos
pit
a
l
inf
or
mation
s
ys
tem
mus
t
e
ns
ur
e
the
a
va
il
a
bil
it
y
of
memor
y
s
pa
c
e
is
a
de
qua
te
to
mee
t
the
ne
e
ds
of
th
e
s
ys
tem
f
low.
T
his
s
tor
a
ge
oc
c
upies
mor
e
a
r
e
a
f
o
r
s
tor
ing
im
a
ge
s
a
long
ti
me
a
s
the
r
e
is
a
ne
e
d
to
ke
e
p
the
r
e
c
or
d
of
numer
ous
pa
ti
e
nts
[
1
,
2]
.
M
e
dica
l
pr
a
c
ti
ti
one
r
s
a
ls
o
ne
e
d
to
s
ha
r
e
da
ta,
e
s
pe
c
ially
medic
a
l
im
a
ge
s
,
to
s
ha
r
e
opini
ons
,
idea
,
a
nd
knowle
dge
to
diagnos
e
a
nd
a
na
lys
e
the
dis
e
a
s
e
.
T
he
y
ne
e
d
to
tr
a
ns
f
e
r
the
medic
a
l
im
a
ge
f
i
les
ove
r
the
ne
twor
k,
whic
h
r
e
qui
r
e
s
ve
r
y
high
ba
ndwidt
h.
T
he
dis
tr
ibut
ion
of
the
medic
a
l
im
a
ge
ne
e
ds
t
o
de
li
ve
r
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
A
r
e
v
iew
on
r
e
gion
of
int
e
r
e
s
t
-
bas
e
d
hy
br
id
me
dical
image
c
ompr
e
s
s
ion
algor
it
hms
(
Suhail
a
A
b
A
z
iz
)
1651
a
c
c
ur
a
tely,
pr
e
c
is
e
ly,
a
nd
s
a
f
e
ly
[
3]
.
He
nc
e
,
the
me
dica
l
im
a
ge
mus
t
be
c
omp
r
e
s
s
e
d
a
nd
s
e
nt
s
e
c
ur
e
ly
thr
ough
the
ne
twor
k
in
the
s
ys
tem.
T
hus
,
c
ompr
e
s
s
ion
ha
s
a
tt
a
ined
much
a
tt
e
nti
on
in
medic
a
l
im
a
ge
s
.
T
he
main
pu
r
pos
e
is
to
r
e
duc
e
the
volum
e
of
t
r
a
ns
mi
tt
e
d
da
ta
in
or
de
r
to
p
r
e
s
e
r
ve
ba
ndwidth
a
nd
wi
thout
s
igni
f
ica
nt
e
f
f
e
c
t
on
th
e
qua
li
ty
of
the
da
ta
by
r
e
movi
ng
a
ny
r
e
dunda
nt
in
f
or
matio
n
[
4]
.
M
e
dica
l
im
a
ge
c
ompr
e
s
s
ion
is
uti
li
s
e
d
in
a
ppli
c
a
ti
ons
to
pr
omot
e
f
a
s
ter
tr
a
ns
mi
s
s
ion
s
pe
e
d
a
nd
r
e
li
a
ble
da
ta
a
nd
r
e
duc
e
s
tor
a
ge
c
os
ts
a
nd
incr
e
a
s
e
tr
a
ns
mi
s
s
ion
s
pe
e
d
[
5]
.
C
omp
r
e
s
s
ion
of
medic
a
l
i
mage
s
is
e
s
s
e
nti
a
l
s
ince
the
e
f
f
icie
nt
s
tor
a
ge
a
nd
t
r
a
ns
mi
s
s
ion
of
da
ta
via
high
ba
ndwidth
digi
tal
c
omm
unica
ti
on
ne
t
wor
ks
is
c
r
uc
ial.
I
mage
c
ompr
e
s
s
ion
will
e
na
bl
e
pictur
e
a
r
c
hivi
ng
a
nd
c
omm
unica
ti
on
s
y
s
tem
(
P
AC
S
)
i
n
Hos
pit
a
l
I
nf
o
r
mation
S
ys
tem
to
r
e
duc
e
f
il
e
s
i
z
e
while
r
e
taining
r
e
leva
nt
diagnos
ti
c
inf
o
r
mation
in
their
s
tor
a
ge
r
e
quir
e
ments
.
He
r
e
the
c
ompr
e
s
s
ion
r
e
q
uir
e
ment
is
r
e
a
li
s
e
d
[
6]
.
I
n
or
de
r
to
de
c
r
e
a
s
e
the
volum
e
of
medic
a
l
i
mage
s
,
the
r
e
gion
of
int
e
r
e
s
t
-
ba
s
e
d
(
R
OI
-
ba
s
e
d
)
c
ompr
e
s
s
ion
a
lgor
it
hm
a
ppr
oa
c
h
wa
s
int
r
oduc
e
d.
T
he
pr
a
c
ti
c
a
li
ty
of
R
OI
is
s
igni
f
ica
nt
in
medic
inal
a
ppli
c
a
ti
ons
whe
r
e
c
e
r
tain
pa
r
ts
of
the
pictur
e
a
r
e
o
f
h
igher
s
ymbol
ic
s
igni
f
ica
nc
e
than
oth
e
r
s
[
7]
.
T
his
tec
hnique
c
a
n
incr
e
a
s
e
the
a
mount
of
th
e
s
tor
a
ge
s
pa
c
e
by
s
e
gmente
d
the
im
a
ge
int
o
R
OI
a
nd
non
-
R
OI
a
r
e
a
.
T
his
tec
hnique
pe
r
mi
ts
e
a
c
h
s
e
gment
c
ompr
e
s
s
e
d
the
im
a
ge
us
ing
the
a
ppr
opr
iate
tec
hnique
t
o
f
u
r
ther
their
potential
be
ne
f
it
f
or
op
ti
mal
r
e
s
ult
s
[
8]
.
T
his
pa
pe
r
a
im
s
to
r
e
view
r
e
c
e
nt
r
e
s
e
a
r
c
h
on
th
e
hybr
id
of
the
r
e
gion
of
int
e
r
e
s
t
-
ba
s
e
d
c
ompr
e
s
s
ion
medic
a
l
im
a
ge
a
lgo
r
it
hm
to
a
c
hieve
high
c
ompr
e
s
s
ion
r
a
ti
os
whe
r
e
a
s
maintaining
im
a
ge
qua
li
ty.
T
his
pa
pe
r
o
r
ga
nis
e
d
in
the
f
oll
owing
wa
y:
T
he
f
ir
s
t
s
e
c
ti
on
of
thi
s
pa
pe
r
wil
l
br
ief
ly
int
r
oduc
e
the
whole
idea
of
the
a
r
ti
c
le.
An
ove
r
view
of
the
medic
a
l
im
a
ge
c
ompr
e
s
s
ion
method
pr
e
s
e
nted
in
s
e
c
ti
on
2.
S
e
c
ti
on
3
will
c
ons
ider
the
s
our
c
e
s
a
nd
pr
a
c
ti
c
e
s
of
the
r
e
late
d
s
tudy
on
R
OI
-
ba
s
e
d
hybr
id
medic
a
l
im
a
ge
c
ompr
e
s
s
ion
tec
hniques
.
T
he
dis
c
us
s
ion
on
s
umm
a
r
is
ing
f
indi
ngs
f
r
om
p
r
e
vious
s
tudi
e
s
will
be
pr
e
s
e
nt
in
s
e
c
ti
on
4.
F
inally,
s
e
c
ti
on
5
gives
the
f
i
na
l
r
e
ma
r
ks
of
the
pa
pe
r
.
2.
M
E
DI
CA
L
I
M
AGE
COM
P
RE
S
S
I
ON
M
e
dica
l
im
a
ge
s
c
ontain
a
gr
e
a
t
de
a
l
o
f
in
f
or
mati
on
that
is
ne
c
e
s
s
a
r
y
f
or
diagnos
is
by
doc
tor
s
a
nd
medic
a
l
pr
of
e
s
s
ionals
.
I
t
is
a
pr
oc
e
s
s
of
r
e
duc
ing
im
a
ge
da
ta
ir
r
e
leva
nc
e
a
nd
r
e
dunda
nc
y
s
o
that
da
ta
c
a
n
pr
a
c
ti
c
a
ll
y
be
s
tor
e
d
o
r
t
r
a
ns
mi
tt
e
d.
I
t
mi
nim
is
e
s
the
s
ize
of
a
r
e
pr
e
s
e
ntation
doc
ument
in
bytes
without
de
gr
a
ding
the
im
a
ge
's
na
tur
e
.
I
mage
c
ompr
e
s
s
ion
include
s
a
f
or
wa
r
d
pr
oc
e
s
s
r
e
f
e
r
r
e
d
to
a
s
e
nc
o
ding
a
nd
a
r
e
ve
r
s
e
pr
oc
e
s
s
r
e
f
e
r
r
e
d
to
a
s
de
c
oding
.
T
he
c
om
pr
e
s
s
ion
pr
os
e
s
s
tar
t
onc
e
the
e
nc
ode
r
r
e
c
e
ived
the
im
a
ge
a
s
the
input
.
T
he
e
nc
ode
r
t
r
a
ns
f
or
med
the
inpu
t
i
mage
int
o
a
s
e
r
ies
o
f
binar
y
da
ta
c
a
ll
e
d
a
bit
-
s
tr
e
a
m.
T
he
de
c
ode
r
then
r
e
c
e
ives
a
nd
de
c
ode
s
the
e
nc
ode
d
bit
-
s
tr
e
a
m
to
f
or
m
the
de
c
ode
d
i
mage
,
a
ls
o
known
a
s
the
r
e
s
ult
ing
im
a
ge
.
F
igur
e
1
s
hows
the
ba
s
ic
f
l
ow
of
im
a
ge
c
ompr
e
s
s
ion
[
9]
.
F
igur
e
1.
B
a
s
ic
f
low
o
f
im
a
ge
c
omp
r
e
s
s
ion
[
9]
Ge
ne
r
a
ll
y,
ther
e
a
r
e
two
c
a
tegor
ies
of
im
a
ge
c
ompr
e
s
s
ion,
los
s
y
a
nd
los
s
le
s
s
[
10]
.
I
n
los
s
les
s
c
ompr
e
s
s
ion,
the
r
e
c
ove
r
e
d
da
ta
identica
l
to
the
or
igi
na
l
,
whe
r
e
a
s
the
r
e
c
ove
r
e
d
da
ta
is
a
c
los
e
r
e
pli
c
a
of
the
or
igi
na
l
wi
th
mi
nim
a
l
da
ta
los
s
i
n
the
c
a
s
e
of
los
s
y
c
omp
r
e
s
s
ion
[
11,
12
]
.
T
he
s
tudy
s
hows
t
ha
t
both
c
ompr
e
s
s
ions
ha
d
s
igni
f
ica
nt
a
dva
ntage
s
a
nd
di
s
a
d
va
ntage
s
in
the
f
ield
of
da
ta
c
ompr
e
s
s
ion
.
T
hus
,
th
e
hybr
id
im
a
ge
c
ompr
e
s
s
ion
tec
hniques
ha
ve
be
e
n
int
r
odu
c
e
d
b
y
manipulating
the
be
ne
f
it
s
o
f
thes
e
two
te
c
hniques
t
o
ge
t
the
maximu
m
im
pa
c
t
[
13]
.
T
he
r
e
f
or
e
,
the
r
e
gion
o
f
int
e
r
e
s
t
(
R
OI
)
is
pr
e
s
e
nt
to
r
e
s
olve
thi
s
is
s
ue
.
T
he
medic
a
l
im
a
ge
is
f
r
a
c
ti
on
int
o
the
s
mall
s
e
gment,
a
nd
e
a
c
h
s
e
gment
c
a
tegor
is
e
d
int
o
R
OI
or
non
-
R
OI
s
e
gment.
Ne
xt,
the
h
yb
r
id
tec
hnique
is
us
e
d
to
a
c
hi
e
ve
be
tt
e
r
r
e
s
ult
s
by
in
hy
br
id
the
los
s
y
a
nd
los
s
les
s
method
of
c
ompr
e
s
s
ion
to
t
he
a
ppr
opr
iate
on
the
s
e
g
ment
c
a
tegor
y
[
9,
14
]
.
T
he
R
OI
s
e
c
ti
on
is
c
ompr
e
s
s
e
d
by
tec
hniques
of
los
s
les
s
c
ompr
e
s
s
ion
while
the
non
-
R
OI
c
ompr
e
s
s
e
s
with
los
s
y
c
ompr
e
s
s
ion.
T
he
r
e
a
r
e
thr
e
e
pa
r
ts
o
f
the
im
a
ge
in
the
med
ica
l
im
a
ge
,
whic
h
is
R
OI
(
r
e
gion
o
f
int
e
r
e
s
t
)
,
non
-
R
OI
,
a
nd
ba
c
kgr
ound.
R
OI
is
the
mos
t
c
r
it
i
c
a
l
pa
r
t
o
f
the
im
a
ge
f
ound
in
ti
ny
r
e
gions
of
th
e
im
a
ge
.
Als
o
include
d
is
Non
-
R
OI
s
o
that
the
us
e
r
c
a
n
identi
f
y
the
mos
t
c
r
it
ica
l
pa
r
t
o
f
the
e
nti
r
e
i
mage
quickly.
Anothe
r
pa
r
t
of
the
im
a
ge
's
c
ontent
is
known
a
s
the
ba
c
kgr
o
und
a
nd
the
im
a
ge
's
mos
t
ignor
e
d
pa
r
t.
I
n
the
medi
c
a
l
f
ield,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
3
,
J
une
2020:
1650
-
1657
1652
thes
e
c
r
it
ica
l
pa
r
ts
ha
d
to
c
omp
r
e
s
s
without
a
ny
l
os
s
c
ompar
e
d
to
other
pa
r
ts
o
f
the
im
a
g
e
with
hig
h
-
qua
li
ty
c
ompr
e
s
s
ion
[
15]
.
F
ig
u
r
e
2
(
a
)
de
mons
tr
a
tes
a
n
e
xa
mpl
e
M
R
I
pictur
e
.
F
igur
e
2
(
b)
a
nd
F
igu
r
e
2
(
c
)
s
how
the
R
OI
r
e
gion
a
nd
non
-
R
OI
r
e
gion
o
f
the
s
a
mpl
e
i
mage
,
r
e
s
pe
c
ti
ve
ly.
(
a
)
(
b)
(
c
)
F
igur
e
2.
(
a
)
Or
igi
na
l
M
R
I
s
c
a
n
,
(
b
)
R
OI
pa
r
t
o
f
th
e
im
a
ge
,
(
c
)
Non
-
R
OI
pa
r
t
o
f
the
im
a
ge
[
16
]
M
a
ny
dif
f
e
r
e
nt
pa
r
a
mete
r
s
c
a
n
be
us
e
d
to
de
s
c
r
ibe
a
nd
c
ompar
e
the
pe
r
f
or
manc
e
of
a
c
ompr
e
s
s
ion
tec
hnique
s
uc
h
a
s
mea
n
s
qua
r
e
e
r
r
or
(
M
S
E
)
,
pe
a
k
s
ignal
to
nois
e
r
a
ti
o
(
P
S
NR
)
,
c
ompr
e
s
s
ion
r
a
t
io
(
C
R
)
,
a
nd
ti
me
c
ons
umpt
ion
to
c
ompr
e
s
s
the
im
a
ge
[
11,
17,
18]
.
T
he
c
omp
r
e
s
s
ion
r
a
ti
o
indi
c
a
tes
the
e
f
f
icie
nc
y
of
the
c
ompr
e
s
s
ion
a
lgor
it
hm.
T
he
higher
the
pe
r
c
e
ntage
of
c
ompr
e
s
s
ion
r
a
ti
o,
the
s
maller
the
s
ize
of
the
im
a
ge
is
c
ompr
e
s
s
e
d.
T
he
f
oll
owing
e
qua
ti
on
c
a
n
c
a
lcula
te
the
c
ompr
e
s
s
ion
r
a
ti
o:
=
(
1)
B
e
s
ide
s
,
the
e
xe
c
uti
on
ti
me
is
a
ls
o
c
r
uc
ial
in
de
ter
mi
ning
the
s
uc
c
e
s
s
f
ulnes
s
of
a
medic
a
l
im
a
ge
c
ompr
e
s
s
ion
a
lgor
it
hm.
T
he
M
S
E
a
nd
P
S
NR
is
t
he
wa
y
to
mea
s
ur
e
the
qua
li
ty
o
f
the
a
ppr
opr
iate
im
a
ge
.
T
he
M
S
E
de
f
ined
a
s
the
s
qua
r
e
of
the
pixel
va
lue
dif
f
e
r
e
nc
e
s
be
twe
e
n
the
two
im
a
ge
s
c
or
r
e
s
pondin
g
pixels
.
T
he
M
S
E
o
f
N
*
M
s
ize
im
a
ge
is
g
iven
by:
=
∑
,
[
1
(
,
)
−
2
(
,
)
]
2
(
∗
)
(
2)
I
−
th
e
in
p
u
t
im
a
g
e
s
M
,
m
−
nu
m
b
e
r
of
r
o
w
s
in
th
e
in
p
u
t
im
a
g
e
s
N
,
n
−
nu
m
b
e
r
of
co
l
u
m
ns
in
th
e
in
p
u
t
im
a
g
e
s
W
hil
e
the
P
S
NR
is
the
pa
r
a
mete
r
be
twe
e
n
the
c
ompr
e
s
s
e
d
im
a
ge
a
nd
the
or
igi
na
l
im
a
ge
to
c
a
lcula
te
the
maximum
e
r
r
o
r
.
T
o
mea
s
ur
e
the
P
S
NR
,
M
S
E
i
s
f
ir
s
t
mea
s
ur
e
d.
M
S
E
is
the
c
umul
a
ti
ve
dif
f
e
r
e
nc
e
be
twe
e
n
the
or
igi
na
l
i
mage
a
nd
the
c
omp
r
e
s
s
e
d
im
a
ge
.
A
low
M
S
E
va
lue
im
pr
ove
s
the
qua
li
ty
o
f
the
im
a
ge
a
nd
e
li
mi
na
tes
the
e
r
r
or
.
T
he
higher
P
S
NR
va
lue
im
p
li
e
s
a
higher
im
a
ge
qua
li
ty
[
9
]
.
T
he
f
oll
owing
is
us
e
d
f
or
c
a
lcula
ti
ng
the
va
lue
of
P
S
NR
:
=
10
10
(
2
)
(
3)
R
is
th
e
m
a
xim
u
m
fl
u
ctu
a
ti
o
n
in
th
e
in
p
u
t
im
a
g
e
d
a
ta
type
3.
VA
RI
OUS
ROI
-
B
ASE
D
HYB
RI
D
M
E
DI
CA
L
I
M
AGE
COM
P
RE
S
S
I
ON
AL
GO
RI
T
HMS
R
e
c
e
ntl
y
we
ha
ve
identif
ied
the
tec
hnique
of
c
ompr
e
s
s
ing
the
medic
a
l
im
a
ge
thor
oughly.
I
n
thi
s
s
e
gment,
we
wil
l
dis
c
us
s
the
a
na
lys
is
of
the
r
e
s
e
a
r
c
h
a
s
s
oc
iate
d
with
the
c
ompr
e
s
s
ion
of
R
OI
-
ba
s
e
d
hybr
id
medic
a
l
im
a
ge
.
Hybr
id
r
e
f
e
r
s
to
a
c
ombi
na
ti
on
of
two
o
r
mor
e
tec
hniques
that
c
ombi
ne
the
e
xc
e
ll
e
nt
pr
ope
r
ti
e
s
of
e
a
c
h
im
a
ge
c
ompr
e
s
s
ion
gr
oup
to
in
c
r
e
a
s
e
the
im
a
ge
qua
li
ty
[
19]
.
T
his
tec
hnique
incor
por
a
tes
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
A
r
e
v
iew
on
r
e
gion
of
int
e
r
e
s
t
-
bas
e
d
hy
br
id
me
dical
image
c
ompr
e
s
s
ion
algor
it
hms
(
Suhail
a
A
b
A
z
iz
)
1653
los
s
y
a
nd
los
s
l
e
s
s
c
ompr
e
s
s
ion
tec
hnique
to
a
c
h
ieve
a
c
ompr
e
s
s
ion
r
a
ti
o
of
high
qua
li
ty
while
r
e
taining
the
nor
m
o
f
the
r
e
c
ons
tr
uc
ted
im
a
ge
.
B
.
Ve
nka
tes
w
a
r
a
e
t
dis
c
ove
r
e
d
the
c
ombi
na
ti
on
of
both
los
s
les
s
a
nd
los
s
les
s
methods
a
s
their
c
ompr
e
s
s
ion
a
lgor
it
hm.
T
he
pr
incipa
l
a
im
of
the
s
e
medic
a
l
im
a
ge
c
ompr
e
s
s
ion
s
tr
a
tegie
s
is
to
s
tor
e
lar
ge
qua
nti
ti
e
s
of
medic
a
l
im
a
ge
s
[
20]
.
I
n
th
is
r
e
s
e
a
r
c
h,
the
M
R
I
medic
a
l
im
a
ge
is
c
ons
ider
e
d.
T
he
a
lgor
it
h
m
be
gins
with
the
im
a
ge
be
ing
s
e
gmente
d
to
R
OI
a
nd
non
-
R
OI
.
T
he
M
R
I
im
a
ge
s
a
r
e
div
ided
int
o
f
r
e
e
-
ha
nd
too
ls
us
ing
M
AT
L
AB
.
T
he
R
OI
s
e
gment
is
c
ompr
e
s
s
e
d
u
s
ing
HA
AR
wa
ve
let
(
HW
T
)
a
nd
the
Da
ube
c
hies
(
Db)
los
s
les
s
c
ompr
e
s
s
ion
tec
hniques
.
W
hil
e
the
non
-
R
OI
s
e
gme
nt
is
c
ompr
e
s
s
e
d
us
ing
los
s
y
c
ompr
e
s
s
ion
tec
hniqu
e
,
in
thi
s
c
a
s
e
,
the
r
e
s
e
a
r
c
he
r
us
ing
e
mbedde
d
z
e
r
ot
r
e
e
wa
v
e
let
(
E
Z
W
)
a
lgor
it
h
m
a
s
the
c
omp
r
e
s
s
ion
a
lgor
it
h
m.
T
he
s
e
s
tudi
e
s
will
be
s
im
ulate
d
with
the
I
ntel
P
e
nt
ium
I
3
pr
oc
e
s
s
or
a
nd
M
AT
L
AB
s
of
twa
r
e
.
T
he
output
im
a
ge
r
e
s
ult
s
a
r
e
e
va
luate
d
via
M
S
E
a
nd
P
S
NR
.
T
he
r
e
s
ult
s
ho
ws
that
thi
s
tec
hnique
of
f
e
r
e
d
a
be
tt
e
r
r
e
s
u
lt
whic
h
the
be
s
t
M
S
E
a
nd
the
P
S
NR
a
r
e
0
.
8546
a
nd
97.
6266
.
C
.
P
e
ter
De
va
dos
s
,
a
s
we
ll
a
s
B
.
S
a
nka
r
a
goma
thi
,
us
e
block
B
ur
r
ow
-
W
he
e
ler
tr
a
ns
f
or
m
-
move
to
f
r
ont
T
r
a
ns
f
or
m
(
B
W
T
-
M
T
F
)
with
Huf
f
man
e
nc
oding
a
nd
hyb
r
id
f
r
a
c
tal
e
nc
oding
to
e
f
f
icie
ntl
y
tr
a
ns
mi
t
med
ica
l
im
a
ge
s
.
[
21]
.
T
his
s
tudy
wa
s
c
onduc
ted
u
s
ing
a
n
ope
n
-
s
our
c
e
M
e
dP
ix
da
taba
s
e
us
ing
the
ult
r
a
s
ound,
M
R
I
,
C
T
s
c
a
n,
a
nd
XR
a
y
im
a
ge
.
T
his
a
lgor
it
h
m
ope
r
a
tes
with
the
los
s
les
s
c
ompr
e
s
s
ion
tec
hnique
f
or
the
R
OI
s
e
c
ti
on,
a
c
ombi
na
ti
on
of
B
W
T
-
M
T
F
a
nd
Huf
f
man
e
nc
oding
.
T
he
a
uthor
us
e
s
the
los
s
y
hyb
r
i
d
F
r
a
c
tal
E
nc
oding
f
o
r
the
non
-
R
OI
s
e
c
ti
on.
T
he
pr
op
os
e
d
c
ompr
e
s
s
ion
s
c
he
me
will
e
li
mi
na
te
thi
s
obs
tac
le
by
us
ing
r
e
gion
of
int
e
r
e
s
t
-
ba
s
e
d
c
ompr
e
s
s
ion,
r
e
s
ult
ing
in
los
s
les
s
c
ompr
e
s
s
ion
of
the
r
e
quir
e
d
a
r
e
a
whe
r
e
the
mos
t
c
r
it
ica
l
de
tails
a
nd
los
s
c
ompr
e
s
s
ion
a
r
e
s
tor
e
d
in
the
r
e
maining
r
e
gion.
T
he
pe
r
f
or
manc
e
of
the
c
o
mpr
e
s
s
ion
s
ys
tem
a
s
s
e
s
s
e
d
with
P
S
NR
,
C
R
,
s
a
vings
in
s
pa
c
e
a
nd
ti
me
c
ons
umpt
ion.
T
he
e
mpi
r
ica
l
outcome
indi
c
a
tes
that
the
pr
opos
e
d
method
of
f
e
r
s
be
tt
e
r
P
S
NR
outcome
s
than
c
onve
nti
ona
l
methods
.
M
os
t
outs
tanding
r
e
s
ult
s
obtaine
d
with
a
n
a
ve
r
a
g
e
va
lue
of
36
.
1
66
dB
.
f
or
ult
r
a
s
ound
im
a
ge
s
a
nd
magne
ti
c
r
e
s
ona
nc
e
im
a
ging
of
34
.
097
dB
.
T
he
opti
mum
C
R
f
ound
with
a
maximum
C
R
of
9
.
621.
Da
vid
Ye
e
e
t
a
l.
c
ompr
e
s
s
the
im
a
ge
us
ing
the
be
tt
e
r
por
table
gr
a
phics
(
B
P
G)
f
or
mat,
whic
h
ba
s
e
d
on
high
-
e
f
f
icie
nc
y
video
e
nc
oding
(
HE
VC
)
.
T
his
tec
h
nique
a
ls
o
us
e
s
br
a
in
s
c
a
nning
im
a
ge
s
f
r
o
m
D
I
C
OM
f
or
mat
of
M
R
I
to
c
ombi
ne
los
s
y
a
nd
los
s
les
s
c
ompr
e
s
s
ion
of
medic
a
l
im
a
ge
[
16]
.
F
i
r
s
t,
the
medic
a
l
im
a
ge
div
ides
int
o
two
pa
r
ts
:
R
OI
a
nd
non
-
R
OI
us
ing
the
a
lgor
it
hm
f
or
f
lood
f
il
li
ng
.
T
he
n
the
los
s
les
s
B
P
G
c
ompr
e
s
s
ion
a
l
gor
it
hm
a
ppli
e
d
to
the
R
OI
a
r
e
a
s
a
nd
the
los
s
B
P
G
a
ppli
e
d
to
r
e
gions
o
ther
than
the
R
OI
.
Ul
ti
mate
ly,
the
r
e
s
ult
ing
r
e
c
ons
tr
uc
ted
im
a
ge
s
a
r
e
c
ombi
ne
d
to
c
r
e
a
te
a
c
o
mpr
e
s
s
e
d
im
a
ge
c
ompl
e
te
a
s
the
e
va
luation
matr
i
x
f
or
the
a
lgor
it
hm,
Da
vid
Ye
e
e
t
a
l.
a
ppli
e
d
r
oot
mea
n
s
qu
a
r
e
e
r
r
or
(
R
M
S
E
)
.
T
he
c
ompr
e
s
s
ion
method's
pe
r
f
or
manc
e
s
ugge
s
ted
s
howing
a
n
im
pr
ove
ment
in
the
c
ompr
e
s
s
ion
r
a
te
be
twe
e
n
10%
a
nd
25%
c
ompar
e
d
to
s
tanda
r
d
im
a
ge
c
ompr
e
s
s
ion
tec
hniques
us
e
d
in
the
medic
a
l
indus
tr
y.
P.
E
be
n
S
oph
ia
a
nd
J
.
Anitha
ha
ve
to
pr
opos
e
the
c
ompr
e
s
s
ion
tec
hnique
of
M
R
I
im
a
ge
s
with
uti
li
s
e
d
R
OI
ba
s
e
d
c
ompr
e
s
s
ion,
whic
h
incr
e
a
s
e
s
the
c
o
mpr
e
s
s
ion
r
a
ti
o
c
ompa
r
e
d
to
the
block
-
ba
s
e
d
method
[
22
]
.
T
he
y
a
r
e
loca
ti
ng
R
OI
with
s
pli
t
a
nd
mer
ge
tec
hnique
a
nd
R
OI
mas
k
ge
ne
r
a
ti
on
.
T
he
R
OI
r
e
gion
is
c
ompr
e
s
s
e
d
us
ing
a
c
ombi
na
ti
on
of
los
s
les
s
c
ompr
e
s
s
i
on
a
lgor
it
hm;
r
un
-
length
c
oding
,
Huf
f
man
c
o
ding,
a
nd
a
r
it
hmetic
c
oding
.
T
he
non
-
R
OI
a
r
e
us
ing
los
s
y
c
ompr
e
s
s
ion,
whic
h
is
Ve
c
tor
Qua
nti
z
a
ti
on.
T
he
e
f
f
e
c
ti
ve
ne
s
s
of
thi
s
a
lgor
it
hm
is
mea
s
ur
e
d
us
ing
C
R
,
P
NSR
,
M
S
E
a
nd
ti
me
pe
r
f
or
manc
e
.
T
he
r
e
c
ons
tr
uc
t
medic
a
l
im
a
ge
s
hows
good
c
ompr
e
s
s
ion
r
a
ti
o
of
a
bout
4.
2
with
P
NSR
20.
76.
S
id
Ahme
d
E
lhanna
c
hi,
Na
c
r
a
B
e
na
mr
a
ne
,
a
nd
T
a
leb
-
Ahme
d
Abde
lm
a
li
k
identif
ied
a
n
e
f
f
icie
nt
e
mbedde
d
im
a
ge
e
nc
ode
r
ba
s
e
d
on
r
e
ve
r
s
ibl
e
dis
c
r
e
te
c
os
ine
tr
a
ns
f
or
mation
(
R
DC
T
)
de
s
igned
f
o
r
los
s
le
s
s
R
OI
c
oding
a
t
a
ve
r
y
high
c
ompr
e
s
s
ion
r
a
ti
o
[
23
]
.
T
he
R
OI
a
s
s
or
tm
e
nt
is
made
manua
ll
y
by
the
r
e
s
e
a
r
c
h
e
r
us
ing
the
va
r
ious
kind
of
medic
a
l
im
a
ge
s
including
M
R
I
,
X
-
r
a
ys
,
C
T
S
c
a
n
a
nd
ult
r
a
s
ound.
T
he
n,
the
non
-
R
O
I
r
e
gion
is
c
ompr
e
s
s
e
d
us
ing
the
s
e
t
pa
r
ti
ti
oning
in
hie
r
a
r
c
hica
l
tr
e
e
(
S
P
I
HT
)
tec
hnique
while
a
r
e
ve
r
s
ibl
e
D
C
T
with
the
los
s
les
s
e
mbedde
d
method
is
us
e
d
to
c
ompr
e
s
s
the
R
OI
r
e
gion.
F
inally,
a
n
e
ntr
opy
c
oding
ba
s
e
d
on
a
da
pti
v
e
a
r
it
hmetic
c
ode
r
a
ppli
e
d
to
the
f
inal
bit
s
tr
e
a
m.
T
his
r
e
s
e
a
r
c
h
is
e
va
luate
us
ing
P
NSR
a
nd
c
or
r
e
lation
c
o
e
f
f
icie
nt
(
C
oC
)
a
s
metr
ics
f
or
the
a
s
s
e
s
s
ment
of
tot
a
l
qua
li
ty
a
nd
bit
r
a
te
(
bit
pe
r
pixel,
bpp
)
f
or
the
e
va
luation
of
los
s
les
s
c
ompr
e
s
s
ion.
T
he
c
omp
r
e
s
s
ion
r
e
s
ult
s
s
how
that
the
pe
r
f
or
manc
e
s
of
the
pr
opos
e
d
ne
w
e
nc
ode
r
a
r
e
much
higher
than
thos
e
of
dif
f
e
r
e
nt
methods
o
f
c
omp
r
e
s
s
ion
of
the
late
s
t
ge
ne
r
a
ti
on
o
f
s
ti
ll
im
a
ge
s
.
M
a
npr
e
e
t
Ka
ur
a
nd
Vika
s
W
a
s
s
on
ha
ve
i
ntr
oduc
e
d
R
OI
-
ba
s
e
d
hybr
id
medic
a
l
im
a
ge
co
mpr
e
s
s
ion
[
24]
.
T
he
wo
r
k
s
tar
ts
with
medic
a
l
im
a
ge
s
pr
e
-
pr
oc
e
s
s
ing,
s
uc
h
a
s
X
-
R
a
y
a
nd
ult
r
a
s
ound,
to
e
li
mi
na
te
the
nois
e
of
a
n
im
a
ge
.
T
hus
s
e
gmenta
ti
on
is
a
ppli
e
d
to
divi
de
the
im
a
ge
int
o
two
unif
o
r
m
pa
r
ts
,
na
mely
R
OI
a
nd
non
-
R
OI
.
C
ompr
e
s
s
ion
f
inally
pe
r
f
or
med
.
T
he
a
utho
r
mana
ge
d
to
us
e
two
c
om
pr
e
s
s
ion
methods
in
the
r
e
s
e
a
r
c
h;
non
-
R
OI
los
s
f
r
a
c
tal
c
om
pr
e
s
s
ion
dur
ing
c
ontext
tr
e
e
we
ight
ing
los
s
les
s
f
or
the
R
OI
pa
r
t.
T
he
s
e
r
e
s
ult
s
c
ompar
e
with
the
s
c
a
lable
c
ompr
e
s
s
ion
tec
hniques
u
s
e
d
by
I
W
T
a
nd
R
B
C
us
ing
the
c
om
pr
e
s
s
ion
r
a
ti
o,
M
S
E
,
a
nd
P
S
NR
.
F
o
r
thi
s
te
c
hnique,
the
a
ve
r
a
ge
c
ompr
e
s
s
ion
r
a
ti
o
is
89
.
6005.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
3
,
J
une
2020:
1650
-
1657
1654
4.
DI
S
CU
S
S
I
ON
M
e
dica
l
im
a
ge
is
a
c
r
uc
ial
a
s
s
e
t
f
or
the
medic
a
l
pr
a
c
ti
ti
one
r
.
T
he
t
r
a
ns
f
e
r
o
f
im
a
ge
da
ta
to
p
r
int
e
d
or
opti
c
a
l
media
ha
s
be
c
ome
ins
uf
f
icie
nt
f
or
c
ur
r
e
nt
ne
e
ds
.
F
ur
ther
mor
e
,
the
wide
s
pr
e
a
d
dis
s
e
mi
na
ti
on
a
nd
a
c
c
e
s
s
ibi
li
ty
of
im
a
ge
da
ta
make
s
e
ns
e
f
or
a
n
e
f
f
icie
nt
medic
a
l
e
c
os
ys
tem
a
nd
e
xc
ha
nge
be
twe
e
n
he
a
lt
h
or
ga
nis
a
ti
ons
.
W
hil
e
a
r
c
hivi
ng
a
nd
a
c
c
e
s
s
ing
medic
a
l
im
a
ge
da
ta
is
a
r
a
ther
t
r
icky
tec
hnica
l
p
r
oblem,
s
uc
h
a
s
the
volu
me
o
f
im
a
ge
,
ti
me
c
ons
tr
a
int
a
nd
pr
ivac
y
r
e
gulations
to
ge
ne
r
a
te
publi
c
c
o
nf
idenc
e
in
c
e
ntr
a
li
s
e
d
s
tor
a
ge
of
c
onf
idential
da
ta,
he
a
lt
hc
a
r
e
or
ga
nis
a
ti
ons
incr
e
a
s
ingl
y
de
pe
nd
on
their
hyb
r
id
c
loud
e
nvir
onments
.
T
his
ha
s
ge
ne
r
a
ted
c
onti
nuous
r
e
que
s
ts
f
or
s
e
c
ur
e
tele
medic
ine
im
pleme
ntations
that
will
pr
ovide
c
onf
identialit
y,
a
uthenticity
,
a
nd
int
e
gr
it
y
f
or
the
medic
a
l
da
ta
tr
a
ns
mi
tt
e
d
in
the
he
a
lt
h
s
e
c
tor
.
I
t
is
c
r
uc
ial
to
e
ns
ur
e
that
a
ll
hos
pit
a
l
inf
or
matio
n
s
ys
tem
tr
a
ns
a
c
ti
ons
a
nd
c
omm
unica
ti
ons
a
r
e
c
a
r
r
ied
out
a
c
c
ur
a
tely
a
nd
without
los
s
of
da
ta.
T
h
e
c
o
mp
r
e
s
s
i
on
o
f
i
ma
ge
s
is
a
c
om
p
r
e
s
s
i
o
n
me
th
od
of
da
ta
,
e
n
c
o
di
ng
a
u
ni
qu
e
i
mag
e
wi
th
f
e
we
r
bi
ts
.
T
h
e
g
oa
l
o
f
i
ma
ge
c
om
p
r
e
s
s
io
n
is
t
o
r
e
d
uc
e
t
he
i
ma
ge
s
ize
f
r
o
m
t
he
o
r
ig
i
na
l
i
mag
e
,
a
n
d
t
he
c
om
p
r
e
s
s
e
d
i
ma
ge
m
us
t
be
s
i
mi
la
r
t
o
th
e
o
r
i
gi
na
l
im
a
ge
.
H
e
r
e
we
c
a
n
d
e
ve
lo
p
a
n
i
de
a
o
f
t
he
d
if
f
e
r
e
n
t
ty
pe
s
o
f
r
e
d
un
da
nc
ies
t
ha
t
n
e
e
d
to
ma
na
ge
wh
i
le
c
o
mp
r
e
s
s
i
ng
a
n
i
ma
ge
,
f
o
ll
ow
i
ng
d
if
f
e
r
e
n
t
c
o
mp
r
e
s
s
i
on
te
c
h
ni
qu
e
s
a
nd
t
he
i
r
a
pp
r
oa
c
h
to
c
om
p
r
e
s
s
in
g
a
n
i
ma
ge
.
E
a
c
h
me
th
od
ha
s
t
he
goa
l
of
a
c
h
ie
vi
ng
a
c
lea
r
ob
jec
t
ive
.
I
n
the
s
uc
c
e
s
s
o
f
a
n
y
m
e
d
ic
a
l
i
m
a
ge
c
o
mp
r
e
s
s
i
o
n
tec
hn
iq
ue
,
s
e
l
e
c
t
i
ng
the
t
ype
o
f
c
om
p
r
e
s
s
io
n
p
la
ys
a
n
e
s
s
e
nt
ia
l
r
o
le
.
I
n
ge
ne
r
a
l
pr
a
c
ti
c
e
of
im
a
ge
diagnos
ti
c
s
,
los
s
y
c
om
pr
e
s
s
ion
s
c
he
mes
r
a
r
e
ly
us
e
d
in
the
medic
a
l
im
a
ge
be
c
a
us
e
of
the
pos
s
ibl
e
los
s
o
f
us
e
f
ul
c
li
nica
l
inf
or
mation
a
nd
ope
r
a
ti
ons
s
uc
h
a
s
im
pr
ove
ment
c
a
n
lea
d
to
f
ur
ther
im
pa
ir
ments
in
los
s
y
c
ompr
e
s
s
ion.
T
he
r
e
f
o
r
e
,
ther
e
is
a
ne
e
d
f
o
r
e
f
f
icie
nt
s
c
he
mes
without
los
s
of
da
ta
f
or
medic
a
l
im
a
ging.
B
e
c
a
us
e
of
that,
mos
t
of
the
p
r
e
vious
s
tudy
r
e
s
e
a
r
c
h
a
pply
los
s
les
s
c
ompr
e
s
s
ion
tec
hnique
to
c
ompr
e
s
s
the
medic
a
l
da
ta.
I
t
is
be
c
a
us
e
the
los
s
les
s
a
lgor
it
hm
will
c
omp
r
e
s
s
the
im
a
ge
without
l
os
ing
a
ny
inf
or
mation.
How
e
ve
r
,
los
s
les
s
c
ompr
e
s
s
ion
ha
s
li
mi
ted
pe
r
f
or
manc
e
in
ter
ms
of
c
ompr
e
s
s
ion
r
a
ti
o,
while
the
s
ize
of
the
im
a
ge
ke
pt
on
gr
owth
dir
e
c
tl
y
p
r
op
o
r
ti
ona
l
to
the
ti
me
a
nd
tec
hnology
de
ve
lopm
e
nt
[
2
5]
.
H
a
n
dl
in
g
a
va
s
t
a
m
ou
nt
o
f
m
e
d
ica
l
im
a
ge
is
ve
r
y
c
ha
l
len
gi
ng
.
He
nc
e
t
he
s
tu
di
e
s
s
h
ow
t
ha
t
t
he
pr
a
c
ti
c
a
l
w
a
y
to
ge
t
th
e
m
a
x
im
um
c
o
mp
r
e
s
s
i
on
r
a
t
io
is
by
s
e
g
men
t
ize
d
t
he
i
mag
e
in
to
R
O
I
a
nd
no
n
-
R
O
I
a
r
e
a
[
26
]
.
I
t
is
i
mp
e
r
a
t
i
ve
t
ha
t
th
e
e
f
f
e
c
ti
ve
ne
s
s
o
f
r
e
s
ol
u
t
io
n
,
a
s
w
e
l
l
a
s
pe
r
c
e
p
tu
a
l
qua
l
it
y
,
be
r
e
s
to
r
e
d
w
hi
le
p
e
r
f
o
r
mi
ng
c
om
p
r
e
s
s
io
n
on
the
med
ic
a
l
i
ma
ge
s
.
T
h
e
s
pe
c
i
f
i
c
l
os
s
o
f
s
i
gn
i
f
i
c
a
nt
i
n
f
o
r
ma
ti
on
a
c
c
o
mp
a
n
ies
c
om
p
r
e
s
s
i
on
i
f
t
he
da
ta
is
m
a
s
s
i
ve
a
n
d
c
h
a
n
ne
l
c
a
pa
c
i
ty
is
h
ig
hl
y
inad
e
q
ua
te
f
o
r
t
r
a
ns
m
is
s
i
on
.
T
h
e
op
t
im
um
im
p
lem
e
n
ta
ti
o
n
t
im
e
b
y
s
pl
it
t
in
g
t
he
im
a
ge
in
t
o
a
s
m
a
l
le
r
s
ize
wi
l
l
r
e
du
c
e
th
e
p
owe
r
a
nd
t
im
e
r
e
q
ui
r
e
d
to
e
xe
c
u
te
t
he
im
a
g
e
i
n
t
he
h
os
p
i
ta
l
in
f
or
ma
t
io
n
s
ys
t
e
m
[
2
5]
.
M
a
ny
r
e
s
e
a
r
c
he
r
s
t
r
y
to
g
e
ne
r
a
te
h
yb
r
id
c
om
p
r
e
s
s
io
n
a
l
go
r
i
t
hm
b
y
c
om
b
in
in
g
va
r
io
us
c
om
p
r
e
s
s
io
n
t
e
c
h
n
iq
ue
s
to
e
nh
a
n
c
e
th
e
c
o
mp
r
e
s
s
i
on
a
l
go
r
it
hm
.
F
i
gu
r
e
3
s
h
ows
th
e
ge
n
e
r
a
l
b
lo
c
k
d
iag
r
a
m
o
f
R
O
I
-
ba
s
e
d
h
yb
r
i
d
me
di
c
a
l
i
ma
ge
c
o
mp
r
e
s
s
i
on
.
T
h
e
p
r
e
p
r
oc
e
s
s
i
ng
s
t
a
ge
o
f
t
he
R
OI
-
ba
s
e
d
h
yb
r
id
m
e
d
ica
l
i
ma
ge
c
om
p
r
e
s
s
io
n
p
r
o
c
e
s
s
is
th
e
s
e
g
me
nt
a
t
io
n
o
f
t
he
o
r
ig
in
a
l
i
ma
ge
.
T
he
i
ma
ge
is
s
u
b
-
d
iv
id
e
d
i
n
to
R
O
I
a
n
d
no
n
-
R
O
I
a
r
e
a
.
G
e
ne
r
a
l
ly
,
t
he
R
O
I
s
e
gm
e
n
t
i
m
pl
e
m
e
n
ts
t
he
l
os
s
les
s
a
lg
o
r
i
th
m
a
nd
no
n
-
R
O
I
c
o
mp
r
e
s
s
e
d
us
in
g
t
he
l
os
s
y
a
l
go
r
it
hm
.
T
he
s
e
lec
t
io
n
o
f
t
e
c
hn
iq
ue
s
a
nd
a
l
go
r
it
hm
s
f
o
r
c
om
p
r
e
s
s
im
a
ge
de
pe
nds
on
th
e
o
bj
e
c
ti
ve
o
f
t
he
r
e
s
e
a
r
c
h
c
o
nd
uc
te
d
b
y
t
he
a
pp
r
op
r
ia
te
r
e
s
e
a
r
c
h
e
r
[
2
2
]
.
T
h
e
n
t
he
r
e
s
u
lt
in
g
c
o
mp
r
e
s
s
i
on
im
a
g
e
f
o
r
b
ot
h
R
OI
a
nd
no
n
-
R
O
I
is
r
e
c
ons
t
r
uc
te
d.
F
i
na
l
ly
,
t
he
i
de
a
l
c
om
p
r
e
s
s
io
n
i
ma
ge
t
ha
t
m
e
e
ts
t
he
ma
in
obj
e
c
ti
ve
s
o
f
t
he
s
tu
dy
be
in
g
c
a
r
r
ied
ou
t
.
F
igur
e
3.
T
he
ge
ne
r
a
l
block
diagr
a
m
Of
R
OI
-
b
a
s
e
d
hybr
id
medic
a
l
im
a
ge
c
ompr
e
s
s
ion
T
a
ble
1
s
hows
the
c
ompar
is
on
of
the
tec
hniques
that
ha
ve
be
e
n
us
ing
by
the
pr
e
vious
s
tudi
e
s
.
T
he
r
e
a
r
e
va
r
ious
s
e
gmenta
ti
on
tec
hniques
to
s
pli
t
the
R
OI
a
nd
non
-
R
OI
of
the
medic
a
l
im
a
ge
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
A
r
e
v
iew
on
r
e
gion
of
int
e
r
e
s
t
-
bas
e
d
hy
br
id
me
dical
image
c
ompr
e
s
s
ion
algor
it
hms
(
Suhail
a
A
b
A
z
iz
)
1655
S
uc
h
a
s
M
AT
L
AB
manua
ll
y
f
r
e
e
ha
nd
tool
s
,
mor
ph
ologi
c
a
l
s
e
gmenta
ti
on
,
f
lood
f
il
l
a
lgor
i
thm
,
s
pli
t
a
n
d
mer
ge
,
a
nd
R
OI
mas
ks
.
M
os
t
r
e
s
e
a
r
c
he
r
s
c
hoos
e
to
us
e
M
AT
L
A
B
ma
nua
ll
y
f
r
e
e
ha
nd
tool
s
tec
hniques
f
or
s
e
gmenta
ti
on
f
or
the
medic
a
l
im
a
ge
be
c
a
us
e
the
main
objec
ti
ve
is
to
f
oc
us
on
the
c
ompr
e
s
s
ion
a
lgor
it
h
m
r
a
ther
than
s
e
gmenta
ti
on
tec
hniques
.
I
n
c
olum
n
3
a
nd
c
olum
n
4,
it
s
hows
that
a
ll
r
e
s
e
a
r
c
he
r
us
e
the
los
s
les
s
c
ompr
e
s
s
ion
a
lgor
it
hm
to
c
omp
r
e
s
s
the
R
OI
s
e
gment,
a
nd
the
non
-
R
OI
s
e
gment
is
c
omp
r
e
s
s
e
d
us
ing
the
los
s
y
c
ompr
e
s
s
ion
a
lgor
it
hm.
T
he
r
e
a
r
e
many
f
o
r
ms
to
e
va
luate
the
s
uc
c
e
s
s
of
the
c
ompr
e
s
s
ion
a
lgor
it
hm,
s
uc
h
a
s
M
S
E
,
P
S
NR
,
C
R
,
B
it
R
a
te,
a
nd
ti
me
pe
r
f
or
manc
e
[
27
]
.
W
e
f
ound
that
the
s
e
lec
ti
on
of
the
e
va
luation
tec
hniq
ue
mus
t
be
in
li
ne
with
the
objec
ti
ve
of
the
s
tudi
e
s
to
ge
t
the
op
ti
mum
r
e
s
ult
of
the
s
ur
ve
y
is
e
s
s
e
n
ti
a
l
[
28]
.
T
a
ble
2
s
hows
the
e
va
luation
met
r
ics
us
e
d
by
the
p
r
e
vious
s
tudi
e
s
.
T
a
ble
1.
C
ompar
is
on
of
tec
hnique
us
e
d
by
p
r
e
viou
s
s
tudi
e
s
R
e
f
S
e
gme
nt
a
ti
on
R
O
I
(
lo
s
s
le
s
s
c
ompr
e
s
s
io
n)
N
on
-
R
O
I
(
lo
s
s
y
c
ompr
e
s
s
io
n)
[
20]
M
A
T
L
A
B
ma
nua
ll
y
f
r
e
e
ha
nd t
ool
H
A
A
R
w
a
ve
le
t
(
H
W
T
)
a
nd t
he
D
a
ub
e
c
hi
e
s
(
D
b)
l
os
s
le
s
s
c
ompr
e
s
s
io
n T
e
c
hni
que
s
E
mbe
dde
d Z
e
r
ot
r
e
e
W
a
ve
le
t
[
21]
M
or
phol
ogi
c
a
l
s
e
gme
nt
a
ti
on
1.
B
ur
r
ow
-
W
he
e
le
r
T
r
a
ns
f
or
m
-
M
ove
t
o f
r
ont
T
r
a
ns
f
or
m
2.
H
uf
f
ma
n C
odi
ng
Q
ua
dr
a
tr
e
e
D
e
c
ompo
s
it
io
n
[
16]
F
lo
od F
il
l
A
lg
or
it
h
m
L
os
s
le
s
s
B
e
tt
e
r
P
or
ta
bl
e
G
r
a
phi
c
s
L
os
s
y B
e
tt
e
r
P
or
ta
bl
e
G
r
a
phi
c
s
[
22]
1.
S
pl
it
a
nd
M
e
r
ge
2.
R
O
I
M
a
s
k
1.
R
un L
e
ngt
h C
odi
ng
2.
H
uf
f
ma
n C
odi
ng
3.
A
r
it
hme
ti
c
C
odi
ng
V
e
c
to
r
Q
ua
nt
iz
a
ti
on
[
23]
M
A
T
L
A
B
ma
nua
ll
y
f
r
e
e
ha
nd t
ool
R
e
ve
r
s
ib
le
D
is
c
r
e
te
C
os
in
e
T
r
a
ns
f
or
ma
ti
on
S
e
t
P
a
r
ti
ti
oni
ng I
n
H
ie
r
a
r
c
hi
c
a
l
T
r
e
e
[
24]
M
A
T
L
A
B
ma
nua
ll
y
f
r
e
e
ha
nd t
ool
C
ont
e
xt
T
r
e
e
W
e
ig
ht
in
g
F
r
a
c
ta
l
L
os
s
y C
ompr
e
s
s
io
n
T
a
ble
2.
S
umm
a
r
is
e
d
o
f
the
e
va
luation
metr
ics
by
pr
e
vious
s
tudi
e
s
R
e
f
e
r
e
nc
e
s
E
va
lu
a
ti
on M
e
tr
ic
s
CR
M
S
E
P
N
S
R
S
pa
c
e
T
im
e
C
oC
[
20]
√
√
[
21]
√
√
√
√
[
16]
√
√
√
[
22]
√
√
√
√
[
23]
√
√
[
24]
√
√
√
T
he
C
R
is
us
e
d
to
e
va
luate
the
r
a
ti
o
of
c
ompr
e
s
s
ion
that
the
p
r
opos
e
d
tec
hnique.
T
he
higher
the
c
ompr
e
s
s
ion
r
a
ti
o,
the
mo
r
e
e
f
f
e
c
ti
ve
the
c
omp
r
e
s
s
ion
a
lgor
it
hm
will
be
in
the
i
mage
.
M
S
E
a
nd
P
S
NR
a
r
e
us
e
d
to
e
va
luate
the
qua
li
ty
of
the
im
a
ge
.
M
S
E
is
r
e
f
e
r
r
ing
to
the
nois
e
in
the
i
mage
,
while
the
P
S
NR
indi
c
a
tes
the
qua
li
ty
of
c
omp
r
e
s
s
ion
im
a
ge
c
ompar
e
d
t
o
the
or
igi
na
l
i
mage
.
T
he
va
lue
of
M
S
E
a
n
d
P
S
NR
is
inver
s
e
ly
pr
opor
ti
ona
l.
T
he
ou
ts
tanding
c
ompr
e
s
s
ion
tec
hnique
will
s
how
low
of
M
S
E
a
nd
high
o
f
P
S
NR
.
M
os
t
r
e
s
e
a
r
c
he
r
s
pr
e
f
e
r
to
us
e
P
S
NR
a
s
the
pr
im
a
r
y
e
va
luation
tec
hniques
a
nd
s
uppor
t
by
the
c
om
pr
e
s
s
ion
r
a
ti
o
a
nd
M
S
E
[
17]
.
F
igur
e
4
a
nd
F
igur
e
5
s
hows
the
tr
e
nd
r
e
s
ult
o
f
the
c
ompr
e
s
s
ion
im
a
ge
us
ing
C
R
a
nd
P
S
NR
f
or
the
pa
r
ti
c
ular
R
OI
-
ba
s
e
d
hybr
id
medic
a
l
im
a
ge
c
ompr
e
s
s
ion
by
pr
e
vious
s
tudi
e
s
.
T
he
gr
a
ph
s
hows
that
the
r
e
s
e
a
r
c
he
r
[
20,
22,
24]
a
r
e
us
ing
both
C
R
a
nd
P
S
NR
a
s
the
e
va
luation
matr
ix
to
mea
s
ur
e
the
a
lgor
it
hm.
T
he
be
s
t
r
e
s
ult
is
by
the
r
e
s
e
a
r
c
h
[
24]
,
whic
h
ge
ne
r
a
te
the
highes
t
C
R
a
nd
P
S
NR
c
omp
r
e
s
s
im
a
ge
.
T
his
r
e
s
e
a
r
c
h
uti
li
s
e
d
f
r
a
c
tal
los
s
y
c
ompr
e
s
s
ion
f
o
r
non
-
R
OI
a
nd
c
ontext
tr
e
e
we
ight
ing
los
s
les
s
f
or
the
R
OI
s
e
gment.
T
he
r
e
s
e
a
r
c
he
r
[
16]
not
c
ons
ider
e
d
the
qua
li
ty
of
the
outpu
t
im
a
ge
;
ther
e
f
o
r
e
,
he
did
not
c
ons
ider
the
P
S
NR
va
lue.
T
he
r
e
s
e
a
r
c
he
r
[
21
,
23
]
a
im
f
or
t
he
qua
li
ty
of
the
output
im
a
ge
.
T
hus
,
the
c
omp
r
e
s
s
ion
r
a
ti
o
is
not
mea
s
ur
e
d
by
them.
He
r
e
we
c
a
n
s
e
e
that
the
s
e
lec
ti
on
of
e
va
luation
methods
f
or
c
ompr
e
s
s
io
n
im
a
ge
s
de
pe
nding
on
the
objec
ti
ve
s
a
nd
goa
ls
o
f
the
s
tudy
.
T
he
r
e
s
e
a
r
c
he
r
r
e
a
li
s
e
s
the
c
ompr
e
s
s
ion
a
lgor
it
hm
by
e
xe
c
uti
ng
a
nd
e
va
luate
us
ing
M
AT
L
AB
.
M
AT
L
AB
wide
ly
us
e
d
by
the
im
a
ge
pr
oc
e
s
s
in
g
a
nd
c
omput
e
r
vis
ion
c
omm
unit
y
[
29
]
.
M
AT
L
AB
a
nd
the
I
mage
P
r
oc
e
s
s
ing
T
oolbox
p
r
ovides
f
unc
ti
ons
a
nd
int
e
r
a
c
ti
ve
tool
s
f
o
r
im
p
r
ove
d
a
nd
a
na
lyzin
g
digi
tal
im
a
ge
s
a
nd
de
ve
lopi
n
g
im
a
ge
pr
oc
e
s
s
ing
a
lgor
it
h
ms
.
T
he
late
s
t
ve
r
s
ion
is
M
AT
L
AB
R
2019
b
[3
0
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
3
,
J
une
2020:
1650
-
1657
1656
M
e
dica
l
s
c
i
e
nc
e
e
volut
ion
is
a
ga
p
b
e
twe
e
n
medic
a
l
s
c
ienc
e
a
nd
a
v
a
il
a
ble
tec
hnologi
e
s
to
s
uppor
t
it
with
a
pur
pos
e
.
I
t
is
i
mper
a
ti
ve
to
r
e
s
tor
e
the
e
f
f
icie
nc
y
of
r
e
s
olut
ion
a
nd
qua
li
ty
o
f
pe
r
c
e
pti
on
whe
n
pe
r
f
or
mi
ng
c
ompr
e
s
s
ion
in
medic
a
l
im
a
ge
s
.
T
he
mas
s
ive
da
ta
with
mi
nim
a
l
c
ha
nne
l
c
a
pa
c
it
y
f
or
tr
a
ns
mi
s
s
ion
pur
pos
e
s
lea
ds
to
a
los
s
of
vi
tal
in
f
or
mation
.
T
he
r
e
f
o
r
e
,
c
o
mpr
e
s
s
ion
of
the
medic
a
l
im
a
ge
pe
r
f
o
r
ms
a
s
a
n
e
s
s
e
nti
a
l
r
e
s
e
a
r
c
h
topi
c
a
bout
the
de
gr
e
e
of
c
omp
r
e
s
s
ion
a
nd
maintena
nc
e
of
the
r
e
leva
nt
in
f
or
mation
[3
1
]
.
F
igur
e
4.
T
he
C
R
of
the
c
ompr
e
s
s
ion
im
a
ge
tr
e
nds
by
the
p
r
e
vious
s
tudi
e
s
F
igur
e
5.
T
he
P
S
NR
of
the
c
omp
r
e
s
s
ion
im
a
ge
tr
e
nds
by
the
p
r
e
vious
s
tudi
e
s
5.
CONC
L
USI
ON
T
he
medic
a
l
im
a
ge
is
a
n
e
s
s
e
nti
a
l
digi
tal
r
e
s
our
c
e
in
the
hos
pit
a
l
inf
o
r
mation
s
ys
tem.
T
he
r
e
a
r
e
dif
f
e
r
e
nt
types
of
modalit
ies
f
o
r
medic
a
l
im
a
ge
s
,
s
uc
h
a
s
C
T
,
X
-
r
a
y,
M
R
I
,
P
E
T
,
a
nd
S
P
E
C
.
Ge
ne
r
a
ll
y
known
that
the
medic
a
l
im
a
ge
c
ons
umes
a
high
volum
e
o
f
s
ize
.
T
he
medic
a
l
pr
a
c
ti
t
ioner
de
mand
to
s
ha
r
e
t
he
im
a
ge
in
or
de
r
to
diagnos
e
a
nd
a
na
lys
e
the
dis
e
a
s
e
with
a
n
a
c
c
ur
a
te,
s
e
c
ur
e
a
nd
f
a
s
t
tr
a
ns
it
ion
.
T
he
r
e
f
o
r
e
,
t
he
mos
t
c
ha
ll
e
nging
is
s
ue
with
s
ha
r
ing
medic
a
l
im
a
ge
is
on
da
ta
s
tor
a
ge
a
nd
ne
twor
k
pe
r
f
or
manc
e
.
T
hi
s
matter
,
whic
h
mot
ivate
d
the
c
ompr
e
s
s
ion
of
the
medic
a
l
im
a
ge
to
e
ns
ur
e
a
c
c
ur
a
cy
,
s
e
c
ur
ity
a
nd
f
a
s
t
medic
a
l
im
a
ge
s
ha
r
ing
.
T
his
pa
pe
r
d
is
c
us
s
e
s
the
r
e
view
on
a
r
e
gion
of
i
nter
e
s
t
-
ba
s
e
d
hybr
id
medic
a
l
im
a
ge
c
ompr
e
s
s
ion
a
lgor
it
hms
.
C
ompr
e
s
s
ion
is
us
e
f
ul
a
s
a
r
e
s
ult
of
it
he
lps
to
c
ut
ba
c
k
r
e
s
our
c
e
c
on
s
umpt
ion
s
uc
h
a
s
di
s
k
s
pa
c
e
a
nd
a
ls
o
the
ba
ndwidth
of
ne
twor
k
tr
a
ns
mi
s
s
ion.
I
n
the
r
e
gion
of
int
e
r
e
s
t
-
ba
s
e
d
hybr
id
medic
a
l
im
a
ge
c
ompr
e
s
s
ion
a
lgor
it
hms
,
R
OI
,
a
nd
non
-
R
OI
e
a
c
h
a
r
e
e
s
s
e
nti
a
l
to
r
e
c
ons
tr
uc
t
the
id
e
a
l
c
omp
r
e
s
s
ion
im
a
ge
.
T
his
pa
pe
r
s
umm
a
r
is
e
s
the
c
ompar
is
on
o
f
the
a
lgor
i
thm
a
nd
pe
r
f
or
manc
e
.
T
he
r
e
c
ons
tr
uc
t
medic
a
l
im
a
ge
us
ing
thi
s
tec
hnique
he
lps
to
r
e
duc
e
the
s
ize
of
the
im
a
ge
while
not
de
gr
a
ding
the
qua
li
ty
of
the
medic
a
l
da
t
a
.
T
his
r
e
view
pa
pe
r
o
f
f
e
r
s
a
tr
a
ns
pa
r
e
nt
idea
c
onc
e
r
ning
e
s
s
e
nti
a
l
c
ompr
e
s
s
ion
a
lgor
it
hm.
T
he
c
hoice
of
the
a
na
lys
is
tec
hnique
is
a
ddit
ionally
e
s
s
e
nti
a
l
to
c
onf
ir
m
that
the
a
lgor
i
thm
a
c
hieve
s
opti
mum
pe
r
f
o
r
manc
e
.
T
he
pr
ojec
t
objec
ti
ve
e
nga
ge
s
the
pr
a
c
ti
c
a
l
de
mand
of
c
ompr
e
s
s
i
on
of
medic
a
l
i
mage
pr
oc
e
dur
e
s
a
nd
t
he
r
e
f
or
e
,
the
e
va
luation
tec
hniques
.
F
or
thi
s
pur
pos
e
,
we
tend
to
s
hould
make
s
ur
e
that
the
e
va
luation
tec
hniqu
e
s
got
to
be
in
li
ne
with
the
tar
ge
t
of
the
s
tudi
e
s
.
W
e
tend
to
c
onc
lude
that
the
thr
e
e
f
a
c
tor
s
that
in
f
luenc
e
the
c
hoice
of
c
ompr
e
s
s
ion
a
lgor
it
hm
a
r
e
:
im
a
ge
qua
li
ty,
c
ompr
e
s
s
ion
r
a
ti
o,
a
nd
c
omp
r
e
s
s
ion
s
pe
e
d.
AC
KNOWL
E
DGE
M
E
NT
S
T
he
a
uthor
s
f
ull
y
a
c
knowle
dge
the
M
ini
s
tr
y
of
E
du
c
a
ti
on
(
M
OE
)
a
nd
Unive
r
s
it
y
T
e
knologi
M
a
lays
ia
f
or
the
r
e
s
e
a
r
c
h
gr
a
nt
(
Q.
K130000
.
3556.
0
5G32)
th
a
t
he
lp
in
f
unding
the
r
e
s
e
a
r
c
h
wor
ks
.
S
pe
c
ial
thanks
to
a
ll
the
r
e
view
e
r
s
f
or
their
va
luable
c
omm
e
nts
a
nd
c
on
s
tr
uc
ti
ve
s
ugge
s
ti
ons
.
RE
F
E
RE
NC
E
S
[1
]
V
arma
D
.,
"
Man
ag
i
n
g
D
IC
O
M
Imag
e
s
:
T
i
p
s
a
n
d
T
r
i
c
k
s
f
o
r
t
h
e
Rad
i
o
l
o
g
i
s
t
,"
In
d
i
a
n
J
R
a
d
i
o
l
I
m
a
g
i
n
g
,
v
o
l
.
2
2
,
n
o
.
1
,
p
p
.
4
-
1
3
,
2
0
1
2
.
[2
]
G
eo
rg
e
C.
K
ag
ad
i
s
,
St
e
v
e
G
.
L
an
g
er
,
"
In
fo
rmat
i
c
s
i
n
Med
i
ca
l
Imag
i
n
g
,"
CR
C
P
r
es
s
,
2
0
1
4
.
[3
]
A
mi
n
e
N
aï
t
‐
A
l
i
,
Ch
ri
s
t
i
n
e
Cav
aro
‐Mé
n
ar
d
,
"
Co
mp
res
s
i
o
n
o
f
Bi
o
me
d
i
ca
l
Imag
es
an
d
Si
g
n
a
l
s
,"
L
o
n
d
o
n
,
U
K
:
IST
E
;
2
0
0
8
.
[O
n
l
i
n
e].
A
v
ai
l
ab
l
e:
h
t
t
p
:
/
/
d
o
i
.
w
i
l
e
y
.
co
m/
1
0
.
1
0
0
2
/
9
7
8
0
4
7
0
6
1
1
1
5
9
[4
]
A
l
i
A
.
H
.
,
A
b
b
a
s
A
.
N
.
,
G
eo
rg
e
L
.
E
.
,
Mo
k
h
t
ar
M
.
R.
,
"
I
mag
e
an
d
au
d
i
o
fract
a
l
co
mp
re
s
s
i
o
n
:
Co
mp
re
h
en
s
i
v
e
rev
i
ew
,
en
h
a
n
cemen
t
s
a
n
d
re
s
earch
d
i
rec
t
i
o
n
s
,"
In
d
o
n
es
i
a
n
J
o
u
r
n
a
l
o
f
E
l
ect
r
i
c
a
l
E
n
g
i
n
eer
i
n
g
a
n
d
C
o
m
p
u
t
er
S
ci
e
n
ce
,
v
o
l
.
15
,
n
o
.
3
,
p
p
.
1
5
6
4
-
15
70
,
2
0
1
9
.
[5
]
A
zman
N
.
A
.
N
.
,
A
l
i
S
.
,
Ras
h
i
d
R
.
A
.
,
Sap
aru
d
i
n
F
.
A
.
,
Sari
j
ari
M
.
A.
,
"
A
h
y
b
r
i
d
p
red
i
ct
i
v
e
t
ec
h
n
i
q
u
e
fo
r
l
o
s
s
l
es
s
i
mag
e
co
m
p
res
s
i
o
n
,
"
B
u
l
l
et
i
n
o
f
E
l
ect
r
i
c
a
l
E
n
g
i
n
ee
r
i
n
g
a
n
d
I
n
f
o
r
m
a
t
i
c
s
,
v
o
l
.
8
,
n
o
.
4
,
p
p
.
1
2
8
9
-
12
96
,
2
0
1
9
.
[6
]
C.
Mc
W
ay
D
.
,
"
T
o
d
ay
’s
H
ea
l
t
h
In
fo
rmat
i
o
n
Man
ag
e
men
t
:
A
n
In
t
e
g
rat
e
d
A
p
p
ro
ac
h
,
Seco
n
d
E
d
i
t
i
o
n
,"
D
el
m
a
r
Cen
g
a
g
e
Lea
r
n
i
n
g
,
2
0
1
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
A
r
e
v
iew
on
r
e
gion
of
int
e
r
e
s
t
-
bas
e
d
hy
br
id
me
dical
image
c
ompr
e
s
s
ion
algor
it
hms
(
Suhail
a
A
b
A
z
iz
)
1657
[7
]
N
a`am
J
.
,
H
arl
an
J
.
,
Pu
t
ra
I
.
,
H
ard
i
a
n
t
o
R
.
,
Prat
i
w
i
M.
,
"
A
n
au
t
o
ma
t
i
c
RO
I
o
f
t
h
e
fu
n
d
u
s
p
h
o
t
o
g
rap
h
y
.
,"
In
t
e
r
n
a
t
i
o
n
a
l
Jo
u
r
n
a
l
o
f
E
l
ect
r
i
c
a
l
a
n
d
C
o
m
p
u
t
er
E
n
g
i
n
ee
r
i
n
g
,
v
o
l
.
8
,
n
o
.
6
,
p
p
.
4
5
4
5
-
45
53
,
2
0
1
8
.
[8
]
D
o
u
k
a
s
C
.
,
Mag
l
o
g
i
a
n
n
i
s
I.
,
"
Reg
i
o
n
o
f
i
n
t
eres
t
co
d
i
n
g
t
ec
h
n
i
q
u
es
f
o
r
med
i
cal
i
mag
e
c
o
mp
re
s
s
i
o
n
,"
IE
E
E
E
n
g
i
n
eer
i
n
g
i
n
M
ed
i
ci
n
e
a
n
d
B
i
o
l
o
g
y
M
a
g
a
z
i
n
e
,
v
o
l
.
26
,
n
o
.
5
,
p
p
.
29
-
35
,
2
0
0
7
.
[9
]
E
ri
n
L
ei
g
h
P.
,
"
Co
mp
res
s
i
o
n
o
f
Med
i
ca
l
Imag
e
s
U
s
i
n
g
L
o
cal
N
ei
g
h
b
o
r
D
i
ffere
n
ce
,"
U
n
i
v
ers
i
t
y
o
f
D
a
y
t
o
n
,
2
0
1
7
.
[1
0
]
Rah
u
l
Sh
arma
C
.
K
.
an
d
D
.
M
.
S.
,
"
H
y
b
ri
d
Med
i
ca
l
Imag
e
Co
mp
res
s
i
o
n
:
Su
rv
e
y
,"
In
t
er
n
a
t
i
o
n
a
l
Jo
u
r
n
a
l
o
f
A
d
v
a
n
ced
R
es
e
a
r
c
h
i
n
C
o
m
p
u
t
er
E
n
g
i
n
ee
r
i
n
g
&
Tech
n
o
l
o
g
y
(IJA
R
CE
T)
,
v
o
l
.
5
,
n
o
.
4
,
p
p
.
1
0
3
6
-
1
0
3
8
,
2
0
1
6
.
[1
1
]
U
k
r
i
t
M
.
F
.
,
Su
res
h
G
.
R.
,
"
E
ffect
i
v
e
L
o
s
s
l
es
s
Co
m
p
res
s
i
o
n
fo
r
M
ed
i
cal
Imag
e
Seq
u
e
n
ces
U
s
i
n
g
Co
m
p
o
s
i
t
e
A
l
g
o
r
i
t
h
m
,"
P
r
o
ceed
i
n
g
s
o
f
I
E
E
E
I
n
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
er
en
ce
o
n
C
i
r
c
u
i
t
,
P
o
wer
a
n
d
C
o
m
p
u
t
i
n
g
Tec
h
n
o
l
o
g
i
e
s
(
ICCP
CT
2
0
1
3
),
p
p
.
1
1
2
2
-
1
1
2
6
,
2
0
1
3
.
[1
2
]
T
.
G
.
Sh
i
rs
at
,
V.
K
.
Bai
ra
g
i
,
"
L
o
s
s
l
es
s
Med
i
cal
Imag
e
Co
mp
res
s
i
o
n
b
y
IW
T
an
d
Pred
i
c
t
i
v
e
Co
d
i
n
g
,"
in
2
0
1
3
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
er
e
n
ce
o
n
E
n
e
r
g
y
E
f
f
i
c
i
en
t
Tech
n
o
l
o
g
i
es
f
o
r
S
u
s
t
a
i
n
a
b
i
l
i
t
y
,
p
p
.
1
2
7
9
-
12
83
,
2
0
1
3
.
[1
3
]
T
h
a
n
k
i
R
.
M
.
,
K
o
t
h
ari
A
.
,
"
H
y
b
ri
d
an
d
A
d
v
an
ce
d
Co
m
p
res
s
i
o
n
T
ec
h
n
i
q
u
es
fo
r
Me
d
i
ca
l
Imag
es
,"
S
p
r
i
n
g
e
r
N
a
t
u
r
e
S
wi
t
z
er
l
a
n
d
,
2
0
1
9
.
[1
4
]
A
l
i
m
O
.
A
.
,
H
amd
y
N
.
,
E
l
-
d
i
n
W
.
G.
,
"
D
et
ermi
n
at
i
o
n
o
f
t
h
e
Reg
i
o
n
o
f
In
t
ere
s
t
i
n
t
h
e
Co
mp
re
s
s
i
o
n
o
f
Bi
o
me
d
i
cal
Imag
es
Co
m
p
res
s
i
o
n
,"
in
24
th
Na
t
l
R
a
d
i
o
S
ci
Co
n
f
(N
R
S
C
2
0
0
7
)
,
p
p
.
1
-
6
,
2
0
0
7
.
[1
5
]
Fi
ro
o
zb
a
k
h
t
M
.
,
D
eh
mes
h
k
i
J
.
,
Mart
i
n
i
M
.
,
E
b
rah
i
md
o
o
s
t
Y
.
,
A
mi
n
H
.
,
D
eh
k
o
r
d
i
M
.
,
et
al
.
,
"
Co
mp
res
s
i
o
n
o
f
D
i
g
i
t
a
l
Med
i
ca
l
Imag
es
Bas
ed
o
n
Mu
l
t
i
p
l
e
Reg
i
o
n
s
o
f
In
t
eres
t
,"
in
2
0
1
0
F
o
u
r
t
h
IE
E
E
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
er
e
n
ce
o
n
D
i
g
i
t
a
l
S
o
c
i
et
y
,
p
p
.
2
6
0
-
26
3
,
2
0
1
0
.
[O
n
l
i
n
e].
A
v
a
i
l
a
b
l
e
:
h
t
t
p
:
/
/
i
e
eex
p
l
o
re.
i
eee.
o
rg
/
d
o
cu
me
n
t
/
5
4
3
2
7
8
6
/
[1
6
]
Y
ee
D
.
,
So
l
t
an
i
n
e
j
ad
S
.
,
H
azari
k
a
D
.
,
Mb
u
y
i
G
.
,
Barn
w
al
R
.
,
Bas
u
A
.
,
"
Med
i
cal
Imag
e
Co
mp
res
s
i
o
n
Bas
e
d
o
n
Reg
i
o
n
o
f
In
t
ere
s
t
U
s
i
n
g
Bet
t
er
Po
r
t
ab
l
e
G
rap
h
i
cs
(BPG
)
,"
in
2
0
1
7
IE
E
E
I
n
t
C
o
n
f
S
ys
t
M
a
n
,
Cy
b
ern
SMC
2
0
1
7
,
p
p
.
2
1
6
-
2
21
,
J
an
u
ary
2
0
1
7
.
[1
7
]
D
i
n
u
A
.
J
.
,
G
an
es
an
R
.
,
K
eb
ed
e
A
.
A
.
,
V
eeras
amy
B.
,
"
Perfo
rman
ce
A
n
a
l
y
s
i
s
an
d
Co
mp
ar
i
s
o
n
o
f
Med
i
cal
Im
ag
e
Co
mp
re
s
s
i
o
n
T
ech
n
i
q
u
e
s
,"
2
0
1
6
In
t
Co
n
f
Co
n
t
r
o
l
In
s
t
r
u
m
Co
m
m
u
n
Co
m
p
u
t
Tech
n
o
l
ICCICCT
,
p
p
.
7
3
8
–
7
45
,
2
0
1
7
.
[1
8
]
Bu
i
V
.
,
Ch
a
n
g
L
.
,
L
i
D
.
,
H
s
u
L
.
,
Ch
en
M
.
Y
.
,
"
D
e
s
cri
p
t
i
o
n
A
D
.
Co
m
p
ari
s
o
n
o
f
L
o
s
s
l
e
s
s
V
i
d
eo
an
d
Im
ag
e
Co
mp
re
s
s
i
o
n
C
o
d
ec
s
fo
r
Me
d
i
ca
l
Co
m
p
u
t
ed
T
o
m
o
g
ra
p
h
y
D
at
a
s
e
t
s
,"
in
2
0
1
6
I
E
E
E
In
t
er
n
a
t
i
o
n
a
l
C
o
n
f
er
e
n
ce
o
n
B
i
g
D
a
t
a
(B
i
g
D
a
t
a
)
,
p
p
.
3
9
6
0
-
3
9
6
2
,
2
0
1
6
.
[1
9
]
A
z
m
a
n
N
.,
A
q
i
l
a
h
b
i
n
t
i
N
.
,
"
H
y
b
r
i
d
Pr
e
d
i
c
t
i
o
n
f
o
r
L
o
s
s
l
e
s
s
Im
a
g
e
C
o
m
p
re
s
s
i
o
n
,"
U
n
i
v
e
r
s
i
t
i
T
e
k
n
o
l
o
g
i
M
a
l
a
y
s
i
a
,
2
0
1
8
.
[2
0
]
Red
d
y
B.
V
en
k
a
t
es
w
ara,
P.
Bh
as
k
ara
Red
d
y
,
P.
Sat
ees
h
K
u
mar
A
.
S
.
R.
,
"
L
o
s
s
l
es
s
Co
m
p
res
s
i
o
n
o
f
Med
i
c
al
Ima
g
es
fo
r
Bet
t
er
D
i
a
g
n
o
s
i
s
,"
IE
E
E
6
th
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
en
c
e
o
n
A
d
va
n
ced
Co
m
p
u
t
i
n
g
2
0
1
6
,
p
p
.
4
0
4
-
40
8
,
2
0
1
6
.
[2
1
]
D
ev
a
d
o
s
s
C
.
P
.
,
San
k
arag
o
ma
t
h
i
B.
,
"
N
ear
L
o
s
s
l
es
s
Med
i
ca
l
Im
ag
e
Co
mp
re
s
s
i
o
n
U
s
i
n
g
Bl
o
ck
BW
T
-
MT
F
an
d
H
y
b
ri
d
Fract
al
C
o
mp
re
s
s
i
o
n
T
ech
n
i
q
u
e
s
,"
Cl
u
s
t
e
r
Co
m
p
u
t
i
n
g
,
v
o
l
.
2
2
,
p
p.
1
2
9
2
9
-
1
2
9
3
7
,
2
0
1
8
;
[2
2
]
So
p
h
i
a
P
.
E.
,
"
Imp
l
emen
t
at
i
o
n
o
f
Reg
i
o
n
Bas
e
d
Med
i
ca
l
Imag
e
Co
m
p
res
s
i
o
n
fo
r
T
el
eme
d
i
c
i
n
e
A
p
p
l
i
cat
i
o
n
,"
2
0
1
4
IE
E
E
In
t
C
o
n
f
Co
m
p
u
t
In
t
el
l
Co
m
p
u
t
R
e
s
,
p
p
.
1
-
4
,
2
0
1
4
.
[2
3
]
E
l
h
an
n
ach
i
S
.
A
.
,
Ben
amra
n
e
N
.
,
A
b
d
el
ma
l
i
k
T
.
A.
,
"
A
d
a
p
t
i
v
e
Me
d
i
ca
l
Imag
e
Co
m
p
res
s
i
o
n
Ba
s
ed
o
n
L
o
s
s
y
an
d
L
o
s
s
l
es
s
E
mb
e
d
d
e
d
Z
ero
t
ree
Met
h
o
d
s
,"
J
In
f
P
r
o
ces
s
S
y
s
t
,
v
o
l
.
13
,
n
o
.
1
,
p
p
.
40
-
56
,
2
0
1
7
.
[2
4
]
K
au
r
M
.
,
W
as
s
o
n
V
.
,
"
RO
I
Bas
ed
Med
i
cal
Imag
e
Co
m
p
res
s
i
o
n
fo
r
T
el
eme
d
i
c
i
n
e
A
p
p
l
i
ca
t
i
o
n
,"
P
r
o
ced
i
a
Co
m
p
u
t
S
ci
,
v
o
l
.
70
,
p
p
.
5
7
9
-
5
85
,
2
0
1
5
.
[2
5
]
E
my
Set
y
an
i
n
g
s
i
h
A
.
H.
,
"
Su
rv
ey
o
f
H
y
b
r
i
d
Imag
e
Co
mp
res
s
i
o
n
T
ec
h
n
i
q
u
es
,"
In
t
J
E
l
ec
t
r
Co
m
p
u
t
E
n
g
,
v
o
l
.
7
,
n
o
.
4
,
p
p
.
2
2
0
6
-
22
14
,
2
0
1
8
.
[2
6
]
D
eep
a
k
.
S.
T
h
o
mas
,
M.
Mo
o
rt
h
i
,
R.
Mu
t
h
al
a
g
u
,
"
Med
i
c
al
Imag
e
Co
mp
res
s
i
o
n
Bas
e
d
o
n
A
u
t
o
mat
e
d
RO
I
Sel
ect
i
o
n
fo
r
T
e
l
emed
i
ci
n
e
A
p
p
l
i
cat
i
o
n
,"
In
t
e
r
n
a
t
i
o
n
a
l
Jo
u
r
n
a
l
o
f
E
n
g
i
n
eer
i
n
g
A
n
d
Co
m
p
u
t
e
r
S
ci
e
n
ce
,
v
o
l
.
3
,
n
o
.
1
,
p
p
.
3
6
3
8
-
36
42
,
2
0
1
4
.
[2
7
]
U
t
h
ay
a
k
u
mar
J
.
,
V
en
g
at
t
araman
T
.
,
D
h
a
v
ach
e
l
v
a
n
P
.
,
"
A
Su
rv
ey
o
n
D
a
t
a
Co
m
p
res
s
i
o
n
T
ech
n
i
q
u
e
s
:
Fr
o
m
t
h
e
Pers
p
ect
i
v
e
o
f
D
a
t
a
Q
u
a
l
i
t
y
,
Co
d
i
n
g
Sch
eme
s
,
D
at
a
T
y
p
e
an
d
A
p
p
l
i
ca
t
i
o
n
s
,"
Jo
u
r
n
a
l
o
f
Ki
n
g
S
a
u
d
U
n
i
ve
r
s
i
t
y
-
Co
m
p
u
t
er
a
n
d
I
n
f
o
r
m
a
t
i
o
n
S
c
i
en
ce
s
,
2
0
1
8
.
[2
8
]
E
l
h
o
u
d
a
A
.
C
.
,
A
b
d
es
s
el
am
B
.
,
Is
mah
an
e
B
.
,
Mu
s
t
a
p
h
a
K
.
,
"
Sel
ec
t
i
o
n
o
f
c
o
mp
re
s
s
i
o
n
t
e
s
t
i
mag
e
s
u
s
i
n
g
v
ari
a
n
ce
-
b
a
s
ed
s
t
at
i
s
t
i
ca
l
met
h
o
d
,"
In
d
o
n
es
J
E
l
ect
r
E
n
g
Co
m
p
u
t
S
ci
,
v
o
l
.
16
,
n
o
.
1
,
p
p
.
2
4
3
-
2
58
,
2
0
1
9
.
[2
9
]
G
o
n
zal
ez
R
.
,
W
o
o
d
s
R
.
,
E
d
d
i
n
s
S.
,
"
D
i
g
i
t
al
Ima
g
e
Pro
ces
s
i
n
g
u
s
i
n
g
M
A
T
L
A
B,
"
Seco
n
d
E
d
i
t
i
o
n
,
G
a
t
es
m
a
r
k
Pu
b
l
i
s
h
i
n
g
,
2
0
0
9
.
[3
0
]
Mat
h
w
o
r
k
s
,
"
MA
T
L
A
B
fo
r
A
rt
i
fi
c
i
a
l
In
t
el
l
i
g
en
ce
D
es
i
g
n
A
I
mo
d
el
s
a
n
d
A
I
-
d
r
i
v
e
n
s
y
s
t
ems
,"
[
O
n
l
i
n
e].
A
v
a
i
l
ab
l
e
h
t
t
p
s
:
/
/
w
w
w
.
ma
t
h
w
o
r
k
s
.
c
o
m/
[3
1
]
G
e
o
f
f
D
o
u
g
h
e
r
t
y
,
"
D
i
g
i
t
a
l
I
ma
g
e
P
r
o
c
e
s
s
i
n
g
f
o
r
M
e
d
i
ca
l
A
p
p
l
i
c
a
t
i
o
n
s
,"
C
a
m
b
r
i
d
g
e
U
n
i
v
e
r
s
i
t
y
P
r
e
s
s
,
N
e
w
Y
o
r
k
;
2
0
0
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.