I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
,
p
p
.
428
~
438
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
14
.i
1
.
p
p
4
2
8
-
4
3
8
428
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Dete
c
tion o
f
lo
ca
t
io
n
-
specific
int
ra
-
cran
ia
l brain tu
m
o
rs
Sh
o
la
Ush
a
ra
ni
1
,
Ra
m
a
P
a
rv
a
t
hy
L
a
k
s
hm
a
na
n
1
,
G
a
y
a
t
hri R
a
j
a
k
u
m
a
ra
n
1
,
Arit
ra
B
a
s
u
1
,
Anj
a
na
Dev
i N
a
nd
a
m
2
,
Siv
a
k
um
a
r
Depuru
3
1
S
c
h
o
o
l
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
V
e
l
l
o
r
e
I
n
st
i
t
u
t
e
o
f
Te
c
h
n
o
l
o
g
y
,
C
h
e
n
n
a
i
,
I
n
d
i
a
2
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
K
o
n
e
r
u
L
a
k
sh
m
a
i
a
h
E
d
u
c
a
t
i
o
n
F
o
u
n
d
a
t
i
o
n
,
V
a
d
d
e
sw
a
r
a
m
,
I
n
d
i
a
3
S
c
h
o
o
l
o
f
C
o
m
p
u
t
i
n
g
,
M
o
h
a
n
B
a
b
u
U
n
i
v
e
r
si
t
y
,
T
i
r
u
p
a
t
i
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
an
25
,
2
0
2
4
R
ev
is
ed
Sep
6
,
2
0
2
4
Acc
ep
ted
Sep
30
,
2
0
2
4
M
u
tatio
n
s
o
r
a
b
n
o
rm
a
li
ti
e
s
i
n
g
e
n
e
s
c
a
n
o
c
c
a
sio
n
a
ll
y
c
a
u
se
c
e
ll
s
to
g
ro
w
u
n
c
o
n
tro
ll
e
d
,
re
su
lt
i
n
g
i
n
a
tu
m
o
r,
wh
ich
is
v
e
ry
d
a
n
g
e
ro
u
s.
Th
e
se
a
re
th
e
m
o
st
p
re
v
a
len
t
c
a
n
c
e
r
c
a
u
se
s.
Th
e
y
a
re
c
a
u
se
d
b
y
sig
n
ifi
c
a
n
t
d
a
m
a
g
e
to
g
e
n
e
s
in
a
sp
e
c
ifi
c
c
e
ll
d
u
rin
g
a
p
e
r
so
n
'
s
e
x
isten
c
e
.
Bra
in
t
u
m
o
rs
a
re
i
n
c
re
a
sin
g
ra
p
id
l
y
,
m
a
j
o
rly
b
ra
i
n
tu
m
o
r
c
a
se
s
in
t
h
e
US
a
re
p
r
o
jec
ted
to
rise
fr
o
m
2
7
,
0
0
0
in
2
0
2
0
t
o
3
1
,
0
0
0
i
n
2
0
2
3
a
t
a
n
a
n
n
u
a
l
g
r
o
wth
ra
te
o
f
1
.
5
%
,
a
ll
t
h
e
c
a
se
s
a
re
risin
g
b
e
c
a
u
se
o
f
th
e
d
e
tec
ti
o
n
o
f
th
e
tu
m
o
rs
i
n
t
h
e
late
p
h
a
se
.
T
h
u
s
,
it
n
e
e
d
s
th
e
h
o
u
r
to
c
re
a
te so
m
e
th
in
g
wh
ich
c
a
n
s
o
lv
e
th
is a
n
o
m
a
ly
a
n
d
h
e
l
p
u
s
d
e
tec
t
th
e
tu
m
o
r
ra
p
id
ly
a
n
d
e
fficie
n
tl
y
.
Wh
il
e
m
a
jo
r
re
se
a
rc
h
p
a
p
e
rs
o
n
b
ra
in
tu
m
o
r
d
e
tec
ti
o
n
m
a
in
l
y
fo
c
u
s
o
n
th
e
d
e
t
e
c
ti
o
n
a
n
d
c
las
sifica
ti
o
n
o
f
th
e
t
u
m
o
rs,
th
e
p
re
se
n
ted
re
se
a
rc
h
a
ims
to
f
irst
d
e
tec
t
th
e
tu
m
o
r
u
si
n
g
p
re
-
re
c
o
g
n
i
z
e
d
p
h
o
to
s
u
sin
g
m
a
c
h
i
n
e
lea
rn
in
g
o
b
jec
t
d
e
tec
ti
o
n
m
o
d
e
ls
.
T
h
e
n
a
fter
su
c
c
e
ss
fu
l
d
e
tec
ti
o
n
o
f
th
e
tu
m
o
r
,
t
h
e
st
u
d
y
tea
m
p
la
n
s
t
o
d
e
term
in
e
i
ts
p
re
c
ise
c
o
o
rd
i
n
a
tes
a
n
d
d
isp
lay
t
h
e
tu
m
o
r
a
n
d
i
ts l
o
c
a
ti
o
n
i
n
t
h
e
p
ict
u
re
.
K
ey
w
o
r
d
s
:
B
r
ain
tu
m
o
r
C
o
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
E
f
f
icien
tNet
Gen
es
Ma
ch
in
e
lear
n
in
g
m
o
d
els
Mo
b
ileNet
v
2
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
:
Gay
ath
r
i Rajak
u
m
ar
an
Sch
o
o
l o
f
C
o
m
p
u
ter
Scien
ce
a
n
d
E
n
g
in
ee
r
in
g
,
Vello
r
e
I
n
s
titu
te
o
f
T
ec
h
n
o
lo
g
y
C
h
en
n
ai,
I
n
d
ia
E
m
ail:
g
ay
ath
r
i.r
@
v
it.a
c.
in
1.
I
NT
RO
D
UCT
I
O
N
A
tu
m
o
r
is
a
ti
s
s
u
e
f
o
r
m
ed
b
y
an
ac
cu
m
u
latio
n
o
f
ab
e
r
r
an
t
ce
lls
.
T
h
ese
ab
n
o
r
m
al
ce
lls
co
n
s
u
m
e
th
e
h
ea
lth
y
b
o
d
ily
ce
lls
,
o
b
liter
ate
th
em
,
an
d
co
n
tin
u
e
to
s
well.
B
r
ain
tu
m
o
r
is
o
n
e
o
f
th
ese
tu
m
o
r
s
.
I
t
h
as
an
im
p
ac
t
o
n
th
e
b
r
ain
,
n
e
u
r
o
lo
g
ical
s
y
s
tem
,
g
lan
d
s
,
an
d
b
r
ain
-
s
u
r
r
o
u
n
d
in
g
m
e
m
b
r
an
es.
I
m
ag
i
n
g
a
n
d
p
ath
o
lo
g
y
ca
n
b
o
th
b
e
u
s
ed
to
d
iag
n
o
s
e
tu
m
o
r
s
.
M
ag
n
etic
r
eso
n
an
ce
im
ag
in
g
(
M
R
I
)
,
wh
ich
p
r
o
d
u
ce
s
cr
o
s
s
-
s
ec
t
io
n
al
im
ag
es
o
f
th
e
b
r
ain
,
is
u
s
ed
to
im
a
g
e
b
r
ain
t
u
m
o
r
s
.
On
e
o
f
th
e
m
o
s
t
s
er
io
u
s
co
n
d
itio
n
s
in
v
o
lv
in
g
th
e
b
r
ai
n
is
a
b
r
ain
tu
m
o
r
,
in
wh
ich
an
u
n
co
n
tr
o
lle
d
g
r
o
wth
o
f
a
b
n
o
r
m
al
ce
lls
o
cc
u
r
s
in
a
g
r
o
u
p
o
f
ce
lls
.
T
h
e
d
ev
elo
p
m
en
t o
f
v
a
r
io
u
s
d
ee
p
lear
n
in
g
alg
o
r
ith
m
s
h
as
led
to
a
n
o
tab
le
e
x
p
an
s
io
n
o
f
t
h
e
f
ield
o
f
im
a
g
e
p
r
o
ce
s
s
in
g
f
o
r
u
s
e
in
b
io
m
e
d
ical
ap
p
licatio
n
s
.
T
h
e
f
o
cu
s
o
f
th
e
r
esear
ch
p
ap
er
is
to
d
etec
t
b
r
ai
n
tu
m
o
r
s
alo
n
g
with
th
e
b
r
ain
tu
m
o
r
lo
ca
tio
n
co
o
r
d
in
ate
in
t
h
e
g
iv
en
MRI
s
ca
n
o
f
th
e
b
r
ain
.
B
r
ain
tu
m
o
r
ca
s
e
s
ar
e
r
is
in
g
r
ap
id
ly
b
ec
au
s
e
o
f
th
e
late
d
etec
tio
n
o
f
b
r
ain
tu
m
o
r
s
.
B
r
ain
tu
m
o
r
ca
s
e
ar
e
in
cr
ea
s
in
g
r
ap
id
ly
,
th
e
ca
s
es
in
th
e
U
n
ited
S
tate
i
s
ex
p
ec
ted
to
r
is
e
f
r
o
m
2
7
,
0
0
0
i
n
2
0
2
0
to
3
1
,
0
0
0
i
n
2
0
2
3
at
an
an
n
u
al
g
r
o
wth
r
ate
o
f
1
.
5
%,
all
th
e
ca
s
es a
r
e
r
i
s
in
g
b
ec
au
s
e
o
f
th
e
late
d
etec
tio
n
o
f
b
r
ain
tu
m
o
r
.
S
o
th
is
p
ap
er
aim
s
ea
r
ly
d
etec
tio
n
o
f
t
h
e
tu
m
o
r
b
y
e
x
p
licitly
g
iv
i
n
g
th
e
lin
k
ca
tio
n
co
o
r
d
i
n
ates o
f
th
e
tu
m
o
r
alo
n
g
with
a
v
is
u
al
r
ep
r
esen
tatio
n
o
f
a
r
ec
tan
g
u
la
r
b
o
x
en
c
o
m
p
ass
in
g
th
e
tu
m
o
r
f
o
r
e
f
f
icien
t d
ete
ctio
n
.
B
r
ain
im
ag
e
p
r
o
ce
s
s
in
g
r
esear
ch
is
cu
r
r
en
tly
f
o
c
u
s
ed
o
n
th
e
i
d
en
tific
atio
n
o
f
b
r
ain
tu
m
o
r
s
.
T
h
is
s
tu
d
y
s
u
g
g
ests
a
s
tr
ateg
y
f
o
r
u
s
in
g
m
ag
n
etic
r
eso
n
a
n
ce
im
ag
es
to
s
eg
m
en
t
an
d
ca
te
g
o
r
ize
b
r
ain
tu
m
o
r
s
(
MRI)
[
1
]
.
Fo
r
tu
m
o
r
s
eg
m
en
tatio
n
,
a
d
e
ep
n
eu
r
al
n
etwo
r
k
s
(
DNN)
-
b
a
s
ed
ar
ch
itectu
r
e
is
u
s
ed
[
2
]
–
[
4
]
ar
ti
c
les
v
alid
atio
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Dete
ctio
n
o
f lo
ca
tio
n
-
s
p
ec
ific in
tr
a
-
cra
n
ia
l b
r
a
in
tu
mo
r
s
(
S
h
o
la
Ush
a
r
a
n
i
)
429
tech
n
iq
u
es u
s
ed
s
u
ch
as
d
ice
s
im
ilar
ity
co
ef
f
icien
t
an
d
th
e
J
a
cc
ar
d
s
im
ilar
ity
i
n
d
ex
.
T
h
e
cla
s
s
if
icatio
n
p
r
o
ce
s
s
is
cr
u
cial
in
th
e
d
iag
n
o
s
is
o
f
b
r
ain
ca
n
ce
r
s
wh
ich
aim
s
to
id
en
tify
b
r
ain
ca
n
ce
r
s
u
s
in
g
v
a
r
io
u
s
ca
teg
o
r
izatio
n
[
5
]
tech
n
iq
u
es
ar
e
ea
r
lier
r
esear
ch
ac
tiv
ities
[
6
]
.
Ho
wev
er
,
cu
r
r
en
t
class
if
icatio
n
m
eth
o
d
s
h
a
v
e
s
u
b
s
tan
tial
f
alse
alar
m
r
ates
(
FAR
s
)
.
T
h
e
w
eig
h
ted
co
r
r
elatio
n
f
ea
t
u
r
e
s
elec
tio
n
b
ased
iter
ativ
e
b
a
y
esian
m
u
ltiv
ar
iate
d
ee
p
n
eu
r
al
lear
n
i
n
g
(
W
C
FS
-
I
B
M
DNL
)
[
7
]
–
[
9
]
tec
h
n
iq
u
e
is
s
u
g
g
ested
in
th
is
s
tu
d
y
t
o
f
asten
th
e
class
if
icatio
n
d
etec
tio
n
o
f
ea
r
ly
-
s
tag
e
b
r
ain
tu
m
o
r
s
f
o
r
f
aster
d
iag
n
o
s
is
m
an
y
au
to
m
ate
d
s
eg
m
en
tatio
n
an
d
class
if
icatio
n
alg
o
r
ith
m
s
th
at
ar
e
s
u
cc
ess
f
u
l in
em
p
lo
y
i
n
g
m
u
ltimo
d
al
MR
I
s
to
ex
tr
ac
t th
e
k
ey
ch
ar
a
cter
i
s
tics
o
f
b
r
ain
tu
m
o
r
d
etec
tio
n
[
1
0
]
,
[
1
1
]
.
So
m
e
m
ajo
r
al
g
o
r
ith
m
s
u
s
ed
ar
e
d
ee
p
lear
n
in
g
m
eth
o
d
s
,
K
-
m
ea
n
s
clu
s
ter
in
g
,
f
u
zz
y
C
-
m
ea
n
s
,
K
-
n
ea
r
est
n
eig
h
b
o
u
r
s
,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
es,
a
n
d
d
ec
is
io
n
tr
ee
s
[
1
2
]
–
[
1
4
]
.
T
h
e
s
y
s
te
m
'
s
p
r
o
g
n
o
s
ticated
d
elica
cy
f
o
r
th
e
test
d
ata
was
9
9
.
1
2
[
1
5
]
.
I
t
s
u
b
s
tan
tially
em
p
lo
y
s
th
e
m
ar
k
s
o
f
p
er
ce
p
tiv
it
y
,
p
ar
ticu
lar
ity
,
an
d
p
r
ec
is
io
n
to
m
ea
s
u
r
e
n
etwo
r
k
p
er
f
o
r
m
a
n
ce
in
ad
d
itio
n
t
o
th
e
d
elica
cy
c
r
iter
io
n
,
wh
ile
it
o
n
ly
d
etec
ts
ex
cr
escen
ce
s
p
r
esen
t,
n
o
v
is
i
b
ilit
y
o
f
ex
c
r
escen
ce
s
is
s
h
o
wn
[
1
6
]
.
T
h
is
s
tu
d
y
f
o
c
u
s
es
o
n
th
e
ass
ess
ab
le
ch
ar
ac
ter
is
tics
o
f
b
r
ain
tu
m
o
r
s
,
lik
e
s
h
ap
e,
s
ig
n
al
in
ten
s
ity
an
d
tex
tu
r
e,
t
o
p
r
e
d
ict
h
ig
h
er
ac
c
u
r
ac
y
with
a
lo
wer
er
r
o
r
r
ate
an
d
th
e
ca
p
ac
ity
f
o
r
f
u
t
u
r
e
wo
r
k
in
th
e
f
ield
[
1
7
]
.
T
h
e
p
ap
er
m
ain
ly
co
v
er
s
c
o
n
v
o
lu
ti
o
n
al
n
eu
r
a
l
n
etwo
r
k
(
C
NN
)
,
wate
r
s
h
ed
al
g
o
r
ith
m
,
a
n
d
r
ec
tifie
d
lin
ea
r
u
n
it (
R
eL
U)
.
T
h
e
m
ain
a
d
v
an
ta
g
es o
f
th
is
p
ap
er
a
r
e
h
ig
h
ac
cu
r
ac
y
,
ad
v
an
ce
d
n
o
v
el
b
r
ain
tu
m
o
r
id
en
tific
atio
n
m
eth
o
d
,
wh
ile
it
o
n
ly
d
etec
ts
tu
m
o
r
s
p
r
esen
t,
n
o
v
is
ib
ilit
y
o
f
t
u
m
o
r
s
is
s
h
o
wn
[
1
8
]
.
T
h
is
ex
p
lo
r
atio
n
p
a
p
er
p
r
o
p
o
s
es
a
wate
r
f
all
o
f
C
NNs
to
m
em
b
e
r
b
r
ain
ex
cr
escen
ce
s
with
h
ier
ar
ch
ic
al
s
u
b
-
r
eg
io
n
s
f
r
o
m
m
u
lti
-
m
o
d
al
g
lam
o
r
o
u
s
r
eso
n
an
ce
i
m
ag
es
(
MRI)
,
an
d
in
tr
o
d
u
ce
a
2
.
5
D
n
etwo
r
k
t
h
at'
s
a
tr
ad
e
-
o
f
f
b
etwe
en
m
em
o
r
y
co
n
s
u
m
p
tio
n
,
m
o
d
el
co
m
p
le
x
ity
an
d
o
p
e
n
f
iel
d
[
1
9
]
.
Alg
o
r
ith
m
s
co
v
er
ed
in
al
ter
n
ate
p
ap
er
ar
e
Mo
n
te
C
ar
lo
s
im
u
latio
n
,
s
tr
u
ctu
r
e
-
w
is
e
q
u
er
y
,
an
d
v
o
x
el
-
wis
e
q
u
er
y
.
T
h
e
m
ain
ad
v
a
n
tag
es
a
r
e
th
e
u
s
es
o
f
Mo
n
te
C
ar
lo
s
im
u
latio
n
to
p
r
o
g
n
o
s
ticate
th
e
p
r
o
b
a
b
ilit
y
o
f
b
r
ai
n
ex
cr
escen
ce
s
eg
m
en
tatio
n
p
o
s
s
ib
ilit
ies
in
ar
b
itra
r
y
s
am
p
les,
h
ig
h
d
elica
cy
o
f
ar
b
itra
r
y
s
am
p
les,
wh
ile
it
o
n
ly
d
etec
ts
ex
cr
escen
ce
s
p
r
esen
t,
n
o
v
is
ib
ilit
y
o
f
ex
cr
escen
ce
s
is
s
h
o
wn
[
2
0
]
.
An
o
th
er
r
esear
c
h
p
ap
er
p
r
o
p
o
s
ed
an
alg
o
r
ith
m
to
s
eg
m
e
n
t
b
r
ai
n
tu
m
o
r
s
f
r
o
m
2
D
MRI
b
y
a
C
N
N
wh
ich
is
f
o
llo
wed
b
y
tr
ad
it
io
n
al
class
if
ier
s
an
d
d
ee
p
lear
n
in
g
m
eth
o
d
s
(
C
NN
an
d
SVM
class
if
ier
m
ain
ly
)
[
2
1
]
.
T
h
is
p
ap
er
'
s
C
NN
m
eth
o
d
h
elp
s
t
o
d
etec
t
t
h
e
tu
m
o
r
f
ast
h
elp
s
in
m
ed
ical
in
d
u
s
tr
y
.
I
n
o
n
e
o
f
th
e
ex
p
lo
r
atio
n
s
,
d
ee
p
f
ea
tu
r
es
ar
e
u
p
r
o
o
ted
f
r
o
m
th
e
in
ce
p
tio
n
v
3
m
o
d
el,
in
wh
ich
s
co
r
e
v
ec
to
r
i
s
ac
q
u
ir
ed
f
r
o
m
So
f
tMa
x
an
d
s
u
p
p
lied
to
th
e
am
o
u
n
t
v
ar
iatio
n
al
class
if
ier
(
QV
R
)
f
o
r
d
em
a
r
ca
tio
n
b
etwe
en
g
lio
m
a,
m
en
in
g
i
o
m
a,
n
o
ex
cr
esce
n
ce
,
an
d
p
itu
itar
y
tu
m
o
r
.
Alg
o
r
ith
m
s
in
clu
d
ed
ar
e
s
u
b
s
tan
tially
f
u
zz
y
c
-
m
ea
n
s
,
QVR
cla
s
s
if
ier
,
an
d
n
eu
r
al
n
etwo
r
k
[
2
2
]
.
T
h
e
m
ain
ad
v
an
tag
es
o
f
th
is
p
ap
er
ar
e
th
e
d
is
tin
ct
co
m
p
a
r
is
o
n
with
Kag
g
le
an
d
B
r
aT
s
s
tan
d
ar
d
m
o
d
els,
class
if
y
in
g
b
r
ain
ex
c
r
escen
ce
in
th
e
ea
r
ly
s
tag
e,
wh
ile
it o
n
ly
d
etec
ts
ex
c
r
escen
ce
s
p
r
esen
t,
n
o
v
is
ib
ilit
y
o
f
ex
c
r
escen
ce
s
ar
e
o
b
s
er
v
e
d
[
2
3
]
.
I
n
a
n
o
th
er
s
tu
d
y
ar
tific
ial
n
eu
r
al
n
etwo
r
k
(
ANN
)
an
d
C
NN
(
m
ajo
r
ly
u
s
ed
alg
o
r
ith
m
s
i
n
th
i
s
s
tu
d
y
)
is
u
s
ed
in
th
e
b
r
ac
k
et
o
f
n
o
r
m
al
an
d
e
x
cr
escen
ce
b
r
ain
.
ANN
wo
r
k
s
lik
e
a
m
o
r
tal
b
r
ain
n
er
v
o
u
s
s
y
s
tem
,
o
n
th
is
b
ase
a
d
ig
ital
co
m
p
u
ter
is
co
n
n
ec
ted
with
lar
g
e
q
u
a
n
tu
m
o
f
in
ter
co
n
n
ec
ted
elem
en
ts
an
d
n
etwo
r
k
in
g
wh
ich
m
ak
es
th
e
n
eu
r
al
n
etwo
r
k
to
tr
ain
with
th
e
u
s
e
o
f
s
im
p
le
p
r
o
c
ess
in
g
u
n
its
ap
p
lied
o
n
th
e
tr
ai
n
in
g
s
et
an
d
s
to
r
es
th
e
ex
is
ten
tial
k
n
o
wled
g
e
[
2
4
]
.
Ma
jo
r
ad
v
an
tag
es
in
clu
d
e
th
e
h
ig
h
ac
c
u
r
ac
y
ac
h
iev
e
d
o
win
g
to
th
e
u
s
e
o
f
C
NN
DNN
tech
n
iq
u
es wh
ile
o
n
ly
d
etec
ts
ex
cr
escen
ce
s
p
r
esen
t,
n
o
v
is
ib
ilit
y
o
f
ex
c
r
escen
ce
s
is
s
h
o
wn
[
2
5
]
.
2.
RE
S
E
ARCH
O
B
J
E
CT
I
V
E
B
r
ain
tu
m
o
r
ca
s
es
ar
e
r
is
in
g
r
ap
id
ly
b
ec
au
s
e
o
f
t
h
e
late
d
ete
ctio
n
o
f
b
r
ai
n
tu
m
o
r
s
.
T
h
e
m
a
in
f
o
cu
s
o
f
th
e
wo
r
k
is
to
d
etec
t
b
r
ain
tu
m
o
r
s
alo
n
g
with
th
e
b
r
ain
tu
m
o
r
lo
ca
tio
n
co
o
r
d
i
n
ate
in
th
e
b
r
ai
n
MRI
s
ca
n
.
I
t
aim
s
to
d
etec
t
th
e
tu
m
o
r
ea
r
ly
b
y
ex
p
licitly
g
iv
in
g
th
e
lin
k
ca
ti
o
n
co
o
r
d
in
ates
o
f
th
e
tu
m
o
r
alo
n
g
with
a
v
is
u
al
r
ep
r
esen
tatio
n
o
f
a
r
ec
tan
g
u
lar
b
o
x
e
n
co
m
p
ass
in
g
th
e
tu
m
o
r
f
o
r
ef
f
icie
n
t d
etec
tio
n
.
T
h
e
p
r
o
p
o
s
ed
wo
r
k
is
d
is
cu
s
s
ed
in
th
e
n
e
x
t
s
ec
tio
n
.
T
h
is
will
elab
o
r
ate
th
e
m
et
h
o
d
o
lo
g
y
an
d
ar
ch
itectu
r
e
m
o
d
el
u
s
ed
.
T
h
e
n
it
f
o
llo
ws
with
th
r
ee
alg
o
r
it
h
m
s
u
s
ed
in
th
e
p
r
o
p
o
s
ed
m
o
d
el
with
f
lo
w
ch
ar
t
d
iag
r
am
s
.
Fin
ally
,
th
e
n
ex
t
s
ec
tio
n
is
ab
o
u
t
r
esu
lts
an
d
t
h
eir
d
is
cu
s
s
io
n
s
f
o
llo
wed
b
y
co
n
c
lu
s
io
n
s
with
f
u
t
u
r
e
wo
r
k
s
.
3.
P
RO
P
O
SE
D
M
O
D
E
L
T
h
e
cu
r
r
en
t
w
o
r
k
is
ca
teg
o
r
iz
ed
in
to
two
s
ec
tio
n
s
.
T
h
e
f
ir
s
t
s
eg
m
en
t
s
ee
k
s
to
id
e
n
tify
b
r
a
in
im
ag
es,
an
d
th
e
s
ec
o
n
d
s
ec
tio
n
aim
s
to
en
clo
s
e
th
e
tu
m
o
r
s
s
ca
n
in
a
r
ec
tan
g
u
lar
b
o
x
with
th
e
tu
m
o
r
’
s
g
eo
g
r
ap
h
ical
co
o
r
d
in
ates.
T
h
e
two
n
e
u
r
al
n
etwo
r
k
m
o
d
els
u
s
ed
th
r
o
u
g
h
o
u
t
th
e
e
n
tire
wo
r
k
,
o
n
e
is
m
o
b
ile
N
et
v
2
an
d
o
th
er
o
n
e
is
ef
f
icien
t
n
et
lite
0
.
W
h
ile
th
e
f
ir
s
t
m
o
d
el
i.e
.
th
e
m
o
b
ile
n
et
e
f
f
icien
t
V2
m
o
d
el
is
a
m
o
b
ile
n
et
class
T
en
s
o
r
Flo
w
m
o
d
el
u
s
ed
f
o
r
o
b
ject
d
etec
tio
n
,
th
e
s
ec
o
n
d
m
o
d
el
i.e
.
th
e
ef
f
icien
t
N
et
is
a
d
if
f
er
en
t
T
en
s
o
r
Flo
w
m
o
d
el
th
at
is
u
s
ed
f
o
r
b
esp
o
k
e
o
b
ject
d
etec
tio
n
.
Fig
u
r
e
1
d
en
o
tes th
e
ar
ch
itectu
r
e
o
f
th
e
o
v
er
all
p
r
o
ce
s
s
wh
er
e
th
e
f
ir
s
t
p
ar
t
is
t
h
e
f
ee
d
e
r
m
o
d
e
l
f
o
r
th
e
im
ag
es
a
n
d
d
etec
ts
wh
eth
er
th
e
tu
m
o
r
is
p
r
esen
t
o
r
n
o
t.
T
h
en
th
e
im
ag
es
o
f
th
e
tu
m
o
r
ar
e
f
ed
in
t
o
th
e
s
ec
o
n
d
p
a
r
t f
o
r
tu
m
o
r
lo
ca
tio
n
d
etec
tio
n
.
Fo
r
th
e
f
ir
s
t sectio
n
,
1
0
0
0
p
h
o
to
s
with
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
428
-
4
3
8
430
s
p
lit
o
f
6
0
:2
0
:2
0
o
f
test
:tra
in
:v
alid
ate
h
av
e
b
ee
n
u
s
ed
,
with
i
m
ag
es
o
f
b
r
ai
n
tu
m
o
r
s
class
if
i
ed
tu
m
o
r
s
class
if
ied
as
Yes
an
d
im
ag
es
o
f
n
o
n
-
t
u
m
o
r
s
class
if
ied
as
No
.
Fo
llo
win
g
tr
ain
in
g
an
d
v
alid
atio
n
with
th
e
im
ag
e,
th
e
m
o
d
el
g
en
er
ates
its
o
u
tp
u
t
o
n
test
im
a
g
es
with
an
ac
cu
r
ac
y
r
ate
o
f
9
8
%.
Fo
llo
win
g
tu
m
o
r
d
et
ec
tio
n
m
o
d
el,
wh
ich
d
eter
m
in
es
th
e
tu
m
o
r
’
s
co
o
r
d
i
n
ate
d
etec
tio
n
,
th
e
d
ata
is
f
ed
in
to
th
e
ef
f
icien
t
n
et
lite
0
m
o
d
el.
L
ab
el
im
g
h
as
b
ee
n
u
s
ed
to
a
n
n
o
tate
p
h
o
t
o
s
o
f
tu
m
o
r
s
with
a
b
o
x
ar
o
u
n
d
th
e
m
as
test
a
n
d
tr
ai
n
d
atasets
f
o
r
th
e
s
ec
o
n
d
m
o
d
el,
co
u
p
led
with
an
x
m
l
f
ile
with
th
e
tu
m
o
r
’
s
co
o
r
d
in
ates.
T
h
e
im
ag
e
with
th
e
tu
m
o
r
e
f
f
icien
t
N
et
m
o
d
el,
is
th
en
ch
ec
k
ed
ag
ai
n
s
t th
e
o
u
tco
m
e
is
th
en
d
is
p
lay
ed
to
g
eth
er
with
th
e
co
o
r
d
in
ates a
n
d
an
ac
c
u
r
at
e
9
6
p
er
ce
n
t v
is
u
al
d
ep
ictio
n
o
f
th
e
im
ag
e
o
f
t
h
e
b
r
ain
tu
m
o
r
.
Fig
u
r
e
1
.
Ar
c
h
itectu
r
e
p
ip
elin
e
d
iag
r
am
f
o
r
th
e
p
r
o
p
o
s
ed
m
o
d
el
4.
M
E
T
H
O
D
T
h
e
en
tire
s
tu
d
y
was
im
p
lem
en
ted
in
Go
o
g
le
C
o
lab
(
GUI
)
,
with
th
e
f
ir
s
t
p
ar
t'
s
d
ataset
tak
en
f
r
o
m
Kag
g
le
an
d
th
e
s
ec
o
n
d
p
ar
t'
s
s
am
p
le
o
f
1
0
0
tu
m
o
r
-
co
n
tain
i
n
g
im
ag
es
b
ein
g
lab
eled
in
L
a
b
el
-
im
g
f
o
r
t
h
e
test
tr
ain
s
et.
T
h
e
m
o
d
els
ar
e
f
r
o
m
th
e
T
e
n
s
o
r
Flo
w
API
,
a
n
d
in
d
iv
id
u
al
twea
k
in
g
h
as
b
ee
n
ap
p
lied
to
ea
c
h
m
o
d
el
to
im
p
r
o
v
e
ac
cu
r
ac
y
.
Mo
b
ile
NetV2
is
an
ex
ce
llen
t
f
ea
tu
r
e
ex
tr
ac
to
r
f
o
r
o
b
ject
r
e
co
g
n
itio
n
an
d
s
eg
m
e
n
tatio
n
.
Fo
r
in
s
tan
ce
,
th
e
n
ew
m
o
d
el
i
s
ap
p
r
o
x
im
ately
3
5
%
f
aster
f
o
r
d
etec
tio
n
wh
en
u
s
ed
with
th
e
r
ec
en
tly
r
elea
s
ed
SS
D
L
ite
wh
ile
m
ain
tain
in
g
th
e
s
am
e
ac
cu
r
ac
y
as
Mo
b
ileNetV1
.
T
h
is
ar
ch
itectu
r
e
m
ain
ly
in
v
o
lv
es
a
3
2
-
f
ilte
r
in
itial
f
u
lly
co
n
v
o
lu
tio
n
lay
e
r
as
well
as
1
9
ad
d
itio
n
al
b
o
ttlen
ec
k
lay
er
s
.
As
s
h
o
wn
in
Fig
u
r
e
2
,
wh
ich
d
ep
icts
Mo
b
ile
n
et
V2
a
r
ch
itectu
r
e
as b
lo
ck
d
iag
r
am
.
T
h
e
r
e
ar
e
two
d
if
f
er
en
t
k
i
n
d
s
o
f
b
lo
c
k
s
in
Mo
b
ileNetV2
.
B
lo
ck
o
n
e
h
as
a
s
tr
id
e
o
f
o
n
e.
An
o
t
h
er
is
a
b
lo
ck
f
o
r
s
h
r
in
k
in
g
with
a
s
tr
id
e
o
f
two
.
Fo
r
b
o
th
v
ar
ie
ties
o
f
b
lo
c
k
s
,
th
er
e
ar
e
th
r
ee
lev
els.
T
h
is
tim
e,
1
1
co
n
v
o
lu
tio
n
s
wit
h
R
eL
U6
m
a
k
e
u
p
th
e
f
ir
s
t
lay
er
.
T
h
e
d
ep
t
h
-
wis
e
co
n
v
o
lu
tio
n
is
th
e
s
ec
o
n
d
lay
er
.
T
h
e
th
ir
d
l
ay
er
is
an
1
1
co
n
v
o
lu
tio
n
o
n
ce
m
o
r
e,
b
u
t th
is
tim
e
th
er
e
is
n
o
n
o
n
-
lin
ea
r
ity
.
Acc
o
r
d
in
g
to
th
is
ar
c
h
itectu
r
e
,
n
eu
r
al
n
etwo
r
k
s
o
n
ly
h
av
e
th
e
ca
p
ac
ity
o
f
a
lin
ea
r
class
if
ier
o
n
t
h
e
non
-
ze
r
o
v
o
lu
m
e
p
o
r
tio
n
o
f
th
e
r
esu
ltin
g
d
o
m
ai
n
if
R
eL
U
is
ap
p
lied
o
n
ce
m
o
r
e
.
As
s
h
o
w
n
i
n
Fig
u
r
e
3
(
a)
d
ep
icts
th
e
ar
ch
itectu
r
e
f
o
r
ca
lcu
latio
n
in
Mo
b
ile
N
et
ten
s
o
r
lay
er
s
.
W
h
er
e
t:
ex
p
an
s
io
n
f
ac
to
r
,
c
:
n
u
m
b
er
o
f
o
u
tp
u
t
ch
an
n
els,
n
:
r
e
p
ea
tin
g
n
u
m
b
e
r
,
s
:
s
tr
id
e.
3
×3
k
er
n
els
ar
e
u
s
e
d
f
o
r
s
p
atial
co
n
v
o
l
u
tio
n
.
T
y
p
ically
,
th
e
p
r
im
ar
y
n
etwo
r
k
(
wid
th
m
u
ltip
lier
1
,
2
2
4
2
2
4
)
u
s
es
3
.
4
m
illi
o
n
p
a
r
am
eter
s
an
d
r
eq
u
i
r
es
3
0
0
m
illi
o
n
m
u
ltip
ly
-
a
d
d
o
p
er
atio
n
s
.
I
n
M
o
b
ileNetV1
,
a
wid
th
m
u
ltip
lier
is
in
tr
o
d
u
ce
d
.
Fo
r
in
p
u
t
r
eso
lu
tio
n
s
r
a
n
g
in
g
f
r
o
m
9
6
to
2
2
4
an
d
wid
th
m
u
ltip
lier
s
r
an
g
in
g
f
r
o
m
0
.
3
5
to
1
.
4
,
th
e
p
er
f
o
r
m
an
c
e
tr
ad
e
-
o
f
f
s
ar
e
f
u
r
th
er
in
v
esti
g
ated
.
W
h
ile
m
o
d
el
s
izes
r
an
g
e
f
r
o
m
1
.
7
M
to
6
.
9
M
p
ar
a
m
ete
r
s
,
n
etwo
r
k
co
m
p
u
tatio
n
ca
n
co
s
t
u
p
t
o
5
8
5
M
MA
d
d
s
.
Usi
n
g
1
6
GPUs
an
d
a
b
atch
s
ize
o
f
9
6
,
t
h
e
n
etwo
r
k
is
tr
ain
ed
.
Fig
u
r
e
3
(
b
)
d
ep
icts
ef
f
icien
t
n
et
ar
ch
itectu
r
e
d
iag
r
a
m
s
wh
ich
is
u
s
ed
f
o
r
o
b
ject
-
d
etec
tin
g
m
o
d
els
h
av
e
3
s
ec
tio
n
s
n
am
ely
b
ac
k
b
o
n
e,
f
ea
tu
r
e
n
etwo
r
k
,
an
d
class
/b
o
x
n
etwo
r
k
.
W
h
ile
th
e
b
ac
k
b
o
n
e
is
th
e
in
p
u
t
s
ec
tio
n
th
at
ex
tr
ac
ts
th
e
in
ten
d
e
d
f
ea
tu
r
es
o
f
an
im
ag
e,
th
e
f
ea
tu
r
e
n
etwo
r
k
r
e
p
r
esen
ts
a
f
u
s
ed
co
m
b
in
atio
n
o
f
th
e
b
ac
k
b
o
n
e
ex
tr
ac
ted
f
ea
tu
r
es,
i.
e.
,
a
co
llectio
n
o
f
f
ea
tu
r
e
ch
a
r
a
cter
is
tic
o
f
th
e
im
ag
e.
T
h
e
b
o
x
n
etwo
r
k
u
s
es
f
u
s
ed
f
ea
tu
r
es
to
p
r
ed
ict
th
e
class
ass
ig
n
ed
to
th
e
r
esp
ec
tiv
e
f
ea
tu
r
e
an
d
its
lo
ca
ti
o
n
.
W
h
ile
tr
ad
itio
n
al
b
ac
k
b
o
n
es
lik
e
R
es
N
et
5
0
an
d
R
esNeXT
ar
e
s
u
itab
le
b
ac
k
b
o
n
es
f
o
r
a
f
ea
t
u
r
e
ex
tr
ac
tio
n
p
r
o
ce
s
s
,
r
ec
en
t
s
tu
d
ies
in
d
icate
th
at
ju
s
t
b
y
r
ep
lacin
g
th
e
tr
a
d
itio
n
al
b
ac
k
b
o
n
es
with
E
f
f
icien
t
N
et
ca
n
im
p
r
o
v
e
th
e
m
o
d
el
ac
c
u
r
ac
y
b
y
3
%
wh
ile
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Dete
ctio
n
o
f lo
ca
tio
n
-
s
p
ec
ific in
tr
a
-
cra
n
ia
l b
r
a
in
tu
mo
r
s
(
S
h
o
la
Ush
a
r
a
n
i
)
431
r
ed
u
cin
g
c
o
m
p
u
tatio
n
p
o
wer
b
y
2
0
%.
W
h
ile
tr
ad
itio
n
al
f
ea
tu
r
e
n
etwo
r
k
lik
e
to
p
-
d
o
wn
f
ea
tu
r
e
p
y
r
am
id
n
etwo
r
k
(
FPN
)
r
estricts
th
e
in
f
o
r
m
atio
n
f
lo
w
to
o
n
e
way
,
b
o
tto
m
-
u
p
f
lo
w
lik
e
n
eu
r
al
ar
ch
itectu
r
e
s
ea
r
ch
(
NAS)
_
FP
N
s
tr
u
ctu
r
e
p
r
o
v
ed
t
o
b
e
a
lo
t
co
m
p
lex
.
So
th
e
in
ten
d
ed
m
o
d
el
p
r
o
p
o
s
ed
a
b
i
d
ir
ec
tio
n
al
f
ea
tu
r
e
n
etwo
r
k
,
b
id
ir
ec
tio
n
al
f
ea
tu
r
e
p
y
r
am
id
n
etwo
r
k
(
B
iFP
N
)
th
at
e
n
ab
l
es
b
o
th
ap
p
r
o
ac
h
es
d
is
cu
s
s
ed
ab
o
v
e
as
well
as
p
o
r
tr
ay
e
d
in
Fig
u
r
e
3
(
b
)
,
wh
i
ch
in
cr
ea
s
e
ac
c
u
r
ac
y
b
y
4
%
an
d
r
e
d
u
ce
s
c
o
m
p
u
tatio
n
al
p
o
wer
b
y
5
0
%
.
T
h
e
ef
f
icien
t n
et
lie
0
a
r
ch
itectu
r
e
h
as b
ee
n
im
p
lem
e
n
ted
b
y
u
s
in
g
a
m
ix
tu
r
e
o
f
n
ew
b
ac
k
b
o
n
e
an
d
B
iFP
N.
Fig
u
r
e
2
.
B
lo
ck
d
iag
r
am
fo
r
Mo
b
ileNetV2
ar
ch
itectu
r
e
(
a)
(
b
)
F
i
g
u
r
e
3
.
A
r
c
h
i
t
ec
t
u
r
e
m
o
d
e
l
f
o
r
(
a
)
M
o
b
i
l
e
N
e
t
t
e
n
s
o
r
s
i
n
M
o
b
i
l
e
N
e
t
[
2
4
]
a
n
d
(
b
)
E
f
f
i
c
i
e
n
tNe
t
l
i
te
d
i
a
g
r
a
m
[
9
]
3
.
1
.
P
r
o
po
s
ed
a
lg
o
rit
hm
f
o
r
det
ec
t
io
n o
f
bra
in t
um
o
r
Fig
u
r
e
4
(
a)
d
en
o
tes
th
e
f
l
o
w
ch
ar
t
p
ip
elin
e
wh
ic
h
d
etec
ts
wh
eth
er
th
e
tu
m
o
r
is
p
r
esen
t
o
r
n
o
t,
its
r
elate
d
s
tep
s
f
o
r
id
en
tific
atio
n
o
f
tu
m
o
r
is
g
iv
en
in
th
e
A
lg
o
r
i
th
m
1
.
I
t
b
asically
co
llects
th
e
s
am
p
le
MRI
im
ag
e
class
if
ied
d
ata
s
et.
I
t c
h
ec
k
s
its
tr
ain
in
g
d
ata
s
et
an
d
test
in
g
d
ata
s
et
f
o
r
v
alid
atin
g
its
ac
cu
r
ac
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
428
-
4
3
8
432
Alg
o
r
ith
m
1
: Step
s
f
o
r
d
etec
ti
o
n
o
f
b
r
ain
t
u
m
o
r
I
n
p
u
t: Sam
p
le
d
ata
s
et
h
av
in
g
y
es a
n
d
n
o
tu
m
o
r
in
im
a
g
e
Ou
tp
u
t: Pr
ed
icted
tu
m
o
r
im
ag
es o
n
test
d
ata
s
et
1.
Star
t Co
llectio
n
Ph
ase
2.
R
ea
d
Sam
p
le
test
im
ag
es h
av
in
g
y
es tu
m
o
r
an
d
n
o
tu
m
o
r
in
MRI
s
ca
n
as tr
ain
,
test
an
d
v
al
id
atio
n
d
ata
s
et
3.
T
r
ain
an
d
v
alid
ate
d
ata
with
m
o
b
ile
n
et_
v
2
teso
r
f
lo
w
m
o
d
el
4.
if
ac
cu
r
ac
y
>.
9
5
th
e
n
5.
C
all
m
o
d
el.
p
r
ed
ict
with
test
im
ag
es to
test
th
e
m
o
d
el
with
s
am
p
le
d
ata
s
et
6.
else
7.
wh
ile
ac
cu
r
ac
y
<
.
9
5
/*
H
y
p
er
T
u
n
e
m
o
d
el
f
ea
tu
r
es *
/
8.
ac
cu
r
ac
y
=T
r
u
e
Po
s
itiv
e
(
I
m
ag
es h
a
v
in
g
tu
m
o
r
p
r
ed
ict
ed
as tu
m
o
r
)
/(
T
r
u
e
Po
s
itiv
e+
T
r
u
e
Neg
ativ
e(
I
m
ag
es h
a
v
in
g
n
o
tu
m
o
r
p
r
ed
icted
n
o
tu
m
o
r
)
)
9.
en
d
10.
en
d
11.
en
d
3.
2
.
P
r
o
po
s
ed
wo
rk
ing
f
lo
w
f
o
r
det
ec
t
io
n o
f
bra
in t
um
o
r
T
h
e
an
n
o
tatio
n
o
f
lo
ca
tio
n
co
o
r
d
in
ates
f
o
r
b
r
ain
tu
m
o
r
s
is
in
d
is
p
en
s
ab
le
f
o
r
d
etec
tin
g
lo
ca
tio
n
-
s
p
ec
if
ic
in
tr
a
-
cr
an
ial
tu
m
o
r
s
.
I
t
en
h
a
n
c
es
p
r
ec
is
io
n
in
tr
ea
tm
e
n
t
p
la
n
n
in
g
,
f
ac
ilit
ates
s
u
r
g
ical
n
av
ig
atio
n
an
d
tar
g
etin
g
,
in
teg
r
ates
with
ad
v
an
ce
d
i
m
ag
in
g
tech
n
o
lo
g
ies,
f
o
s
ter
s
m
u
ltid
is
cip
lin
ar
y
co
llab
o
r
atio
n
,
an
d
s
u
p
p
o
r
ts
lo
n
g
itu
d
in
al
m
o
n
ito
r
in
g
an
d
f
o
llo
w
-
u
p
,
u
ltima
tely
im
p
r
o
v
in
g
p
atien
t
o
u
tco
m
es
an
d
q
u
ality
o
f
ca
r
e.
Fig
u
r
e
4
(
b
)
d
en
o
tes
th
e
f
lo
w
c
h
ar
t
p
ip
elin
e
to
an
n
o
tate
th
e
b
r
ai
n
tu
m
o
r
f
r
o
m
th
e
im
ag
e.
L
a
b
el
I
m
g
is
u
s
ed
f
o
r
a
n
n
o
tatin
g
th
e
tu
m
o
r
s
p
r
esen
t
in
th
e
im
a
g
e
an
d
th
e
x
m
l
co
o
r
d
in
ates
o
f
th
e
t
u
m
o
r
g
et
s
to
r
ed
alo
n
g
wi
th
th
e
im
ag
e
f
o
r
th
e
n
ex
t
m
o
d
el
tr
ain
in
g
p
r
o
ce
s
s
.
T
h
e
XM
L
co
o
r
d
in
ates
ar
e
cr
u
cia
l
s
in
ce
th
ey
ac
t
as
a
f
ee
d
er
s
et
i
n
th
e
n
e
x
t
alg
o
r
ith
m
alo
n
g
with
th
e
d
etec
ted
tu
m
o
r
im
ag
es.
I
ts
r
elate
d
s
tep
s
ar
e
g
i
v
en
as a
n
A
lg
o
r
ith
m
2
.
(
a)
(
b
)
Fig
u
r
e
4
.
Flo
wch
ar
t
f
o
r
(
a)
d
et
ec
tio
n
o
f
b
r
ain
tu
m
o
r
a
n
d
(
b
)
a
n
n
o
tatin
g
o
f
d
etec
tio
n
b
r
ai
n
tu
m
o
r
Alg
o
r
ith
m
2
: Step
s
f
o
r
an
n
o
tat
in
g
th
e
lo
ca
tio
n
s
o
f
b
r
ain
tu
m
o
r
I
n
p
u
t: Sam
p
le
d
ata
s
et
h
av
in
g
y
es a
n
d
n
o
tu
m
o
r
in
im
a
g
e
Ou
tp
u
t: L
o
ca
tio
n
c
o
o
r
d
in
ate
o
f
tu
m
o
r
in
im
ag
e
i
n
x
m
l f
o
r
m
a
t
1.
Star
t c
o
llectio
n
p
h
ase
/*
Fro
m
p
r
ev
io
u
s
alg
o
r
ith
m
o
u
tp
u
t*
/
2.
R
ea
d
s
am
p
le
test
im
ag
es f
r
o
m
p
r
ev
io
u
s
m
o
d
el
r
esu
lts
3.
do
4.
if
tu
m
o
r
p
r
esen
t t
h
en
5.
a
n
n
o
tate
tu
m
o
r
with
a
r
ec
tan
g
u
lar
b
o
x
u
s
in
g
lab
elim
g
6.
L
o
c(
tu
m
o
r
)
=
(
x
lef
t,y
lef
t
)
,
(
x
lef
t,y
r
ig
h
t)
,
(
x
r
ig
h
t,
y
lef
t)
,
(
x
r
i
g
h
t,y
r
ig
h
t)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Dete
ctio
n
o
f lo
ca
tio
n
-
s
p
ec
ific in
tr
a
-
cra
n
ia
l b
r
a
in
tu
mo
r
s
(
S
h
o
la
Ush
a
r
a
n
i
)
433
7.
s
to
r
e
L
o
c(
tu
m
o
r
)
in
x
m
l
f
o
r
m
at
8.
en
d
9.
till
im
ag
es a
r
e
p
r
esen
t
10.
en
d
11.
Pre
p
ar
e
th
e
an
n
o
tated
im
a
g
es a
n
d
x
m
l
f
ile
co
n
tain
in
g
lo
ca
tio
n
co
o
r
d
in
ates o
f
tu
m
o
r
as test,
tr
ain
an
d
v
alid
atio
n
s
et
f
o
r
n
ex
t a
lg
o
r
ith
m
12.
E
n
d
3.
3
.
P
r
o
po
s
ed
a
lg
o
rit
hm
f
o
r
f
ind
ing
t
he
lo
ca
t
io
n c
o
o
rdina
t
es o
f
bra
in t
um
o
r
Acc
u
r
ate
lo
ca
tio
n
co
o
r
d
i
n
ates
p
r
o
v
id
e
p
r
ec
is
e
in
f
o
r
m
atio
n
ab
o
u
t
th
e
tu
m
o
r
'
s
p
o
s
itio
n
with
in
th
e
b
r
ain
.
T
h
is
p
r
ec
is
io
n
is
cr
u
cial
f
o
r
tr
ea
tm
en
t
p
lan
n
in
g
,
en
ab
lin
g
h
ea
lth
ca
r
e
p
r
o
v
id
er
s
to
d
et
er
m
in
e
th
e
o
p
tim
al
ap
p
r
o
ac
h
f
o
r
s
u
r
g
er
y
,
r
ad
iatio
n
th
er
ap
y
,
o
r
o
t
h
e
r
in
ter
v
en
tio
n
s
.
I
t
allo
ws
f
o
r
th
e
d
ev
elo
p
m
en
t
o
f
cu
s
to
m
ized
tr
ea
tm
en
t
s
tr
ateg
ies
tailo
r
ed
t
o
th
e
tu
m
o
r
'
s
s
p
ec
if
ic
lo
ca
tio
n
,
m
in
im
izin
g
th
e
r
is
k
o
f
d
am
ag
e
to
cr
itical
b
r
ain
s
tr
u
ctu
r
es
an
d
im
p
r
o
v
in
g
tr
ea
tm
en
t
o
u
tco
m
es.
Fig
u
r
e
5
d
e
n
o
tes
th
e
a
r
ch
itectu
r
e
p
ip
e
lin
e
th
at
p
r
o
ce
s
s
es
th
e
an
n
o
tated
tu
m
o
r
im
a
g
es
alo
n
g
with
XM
L
co
o
r
d
in
ates
in
to
a
n
ef
f
icien
t
n
et
lite
0
T
en
s
o
r
Flo
w
m
o
d
el.
T
h
is
s
tep
is
th
e
m
o
s
t
cr
u
cial
o
n
e
f
o
r
th
e
r
esear
ch
p
u
r
p
o
s
e
s
in
ce
it
i
n
tr
o
d
u
ce
s
a
n
i
n
n
o
v
ativ
e
way
o
f
lo
ca
tin
g
th
e
b
r
ain
tu
m
o
r
s
in
th
e
im
a
g
e
with
in
cr
e
ased
ac
cu
r
ac
y
.
T
h
e
in
p
u
ts
ar
e
f
u
r
th
er
s
p
lit
in
to
test
tr
ain
an
d
v
alid
ate
s
et
an
d
th
e
m
o
d
el
is
h
y
p
er
p
a
r
am
eter
tu
n
e
d
till
th
e
tar
g
et
ac
cu
r
ac
y
is
r
ea
ch
ed
af
ter
wh
ich
it
g
iv
es
th
e
d
esire
d
o
u
tp
u
t
im
ag
es
h
av
in
g
t
u
m
o
r
l
o
ca
tio
n
s
as c
o
o
r
d
in
ates in
th
e
p
ictu
r
e
its
elf
alo
n
g
with
a
r
ec
tan
g
u
la
r
an
n
o
tatio
n
.
Fig
u
r
e
5
.
Flo
wch
ar
t
f
o
r
a
n
n
o
ta
tin
g
an
d
s
to
r
i
n
g
im
ag
es
Alg
o
r
ith
m
3
: Step
s
f
o
r
d
etec
ti
o
n
o
f
lo
ca
tio
n
c
o
o
r
d
in
ates o
f
b
r
ain
tu
m
o
r
I
n
p
u
t: Sam
p
le
b
r
ain
m
r
i im
ag
e
h
av
in
g
tu
m
o
r
Ou
tp
u
t: Pr
ed
icted
lo
ca
tio
n
co
o
r
d
in
ates o
f
tu
m
o
r
im
a
g
e
1.
Star
t Co
llectio
n
Ph
ase
/*
S
to
r
ed
x
m
l a
n
d
th
e
im
a
g
es f
r
o
m
t
h
e
last
alg
o
r
ith
m
*
/
2.
R
ea
d
Sam
p
le
test
im
ag
es h
av
in
g
an
n
o
tated
im
ag
es a
n
d
lo
ca
t
io
n
co
o
r
d
in
ates o
f
t
u
m
o
r
in
x
m
l
as
tr
ain
,
test
an
d
v
alid
atio
n
d
ata
s
et
3.
T
r
ain
an
d
v
alid
ate
d
ata
with
e
f
f
icien
t n
et
lite 0
ten
s
o
r
f
l
o
w
m
o
d
el
4.
if
ac
cu
r
ac
y
>
.
9
5
t
h
en
5.
C
all
m
o
d
el.
p
r
ed
ict
with
test
im
ag
es to
test
th
e
m
o
d
el
with
s
am
p
le
d
ata
s
et
to
p
r
e
d
ict
tu
m
o
r
lo
ca
tio
n
in
v
is
u
al
r
ep
r
esen
tatio
n
6.
else
7.
wh
ile
ac
cu
r
ac
y
<
.
9
5
/*
Hy
p
er
T
u
n
e
m
o
d
el
f
ea
tu
r
es *
/
8.
ac
cu
r
ac
y
=T
r
u
e
Po
s
itiv
e(
I
m
ag
es h
av
in
g
tu
m
o
r
p
r
ed
icted
as tu
m
o
r
)
/(
T
r
u
e
Po
s
itiv
e+
T
r
u
e
Neg
ativ
e(
I
m
ag
es h
av
in
g
n
o
tu
m
o
r
p
r
ed
icted
n
o
tu
m
o
r
)
9.
en
d
10.
en
d
11.
en
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
428
-
4
3
8
434
3.
4
.
E
qu
a
t
io
ns
u
s
ed
E
q
u
atio
n
s
f
o
r
co
m
p
u
tatio
n
al
co
s
ts
o
f
co
n
v
en
tio
n
al
c
o
n
v
o
l
u
tio
n
an
d
d
e
p
th
wis
e
s
ep
ar
a
b
le
co
n
v
o
l
u
tio
n
as u
s
ed
in
Alg
o
r
i
th
m
s
1
an
d
3
an
d
h
as u
s
ed
tr
ai
n
in
g
an
d
test
s
et
r
esp
ec
tiv
ely
.
=
ℎ
∗
∗
∗
∗
2
(
1
)
=
ℎ
∗
∗
∗
(
+
2
)
(
2
)
W
h
er
e
:
co
s
t
o
f
d
e
p
th
wis
e
s
ep
ar
ab
le
co
n
v
o
lu
tio
n
,
:
co
s
t
o
f
c
o
n
v
e
n
tio
n
al
c
o
n
v
o
lu
tio
n
,
i:
in
d
ex
o
f
in
p
u
t
lay
er
,
j:
in
d
ex
o
f
o
u
t
p
u
t
lay
er
,
ℎ
:
in
p
u
t
f
ea
t
u
r
e
m
ap
s
h
eig
h
t,
:
in
p
u
t
f
ea
tu
r
e
m
ap
s
wid
th
,
: in
p
u
t f
ea
tu
r
e
m
ap
s
n
u
m
b
er
,
: o
u
tp
u
t
f
ea
tu
r
e
m
a
p
s
n
u
m
b
er
,
an
d
K:
f
ilter
s
ize.
T
r
u
e
p
o
s
itiv
e
r
ates
eq
u
atio
n
a
s
u
s
ed
in
m
et
r
ic
s
co
r
es
o
f
m
o
d
el
ev
alu
atio
n
wh
ich
tak
es
in
t
o
tr
ain
in
g
,
test
,
an
d
v
alid
atio
n
d
ataset
in
to
ac
co
u
n
t:
=
[
1
,
1
⋯
1
,
4
⋮
⋱
⋮
4
,
1
⋯
4
,
4
]
(
3
)
tpr
i
=
m
i
,
i
∑
m
i
,
j
4
j
=
1
(
4
)
W
h
er
e
,
:
co
u
n
ts
elem
en
ts
lab
e
led
with
class
I
,
b
u
t
p
r
ed
icted
as
class
j,
M
m
atr
ix
:
co
n
f
u
s
io
n
m
at
r
ix
with
d
iag
o
n
al
elem
e
n
ts
as tr
u
e
p
o
s
i
tiv
es a
n
d
r
est as m
is
class
if
icat
io
n
s
,
an
d
tpr
i
: tr
u
e
p
o
s
itiv
e
r
ate
f
o
r
class
i
.
E
r
r
o
r
r
ate
f
o
r
ep
o
c
h
a
n
d
class
as
u
s
ed
in
m
etr
ic
s
co
r
es o
f
m
o
d
el
ev
alu
atio
n
wh
ich
tak
es
in
t
o
tr
ain
in
g
,
test
,
an
d
v
alid
atio
n
d
ataset
in
to
ac
co
u
n
t:
,
=
1
−
,
(
5
)
W
h
er
e
,
: e
r
r
o
r
r
ate
f
o
r
ep
o
c
h
t
an
d
class
I
,
,
: tr
u
e
p
o
s
itiv
e
r
ate
f
o
r
ep
o
ch
t a
n
d
class
i
Def
au
lt
b
o
u
n
d
a
r
y
b
o
x
wid
th
an
d
h
eig
h
t
o
f
ef
f
icien
t
n
et
lite
0
as
u
s
ed
in
Alg
o
r
ith
m
3
m
o
d
el
b
u
ild
p
h
ase
f
o
r
tr
ain
in
g
,
test
,
an
d
v
alid
atio
n
d
ataset.
w1
=
s
c
a
l
e
∗
√
a
s
pe
c
t
r
a
tio
(
6
)
h1
=
s
c
al
e
√
as
p
ect
r
at
i
o
(
7
)
E
f
f
icien
t n
et
ad
d
s
an
ex
tr
a
d
ef
au
lt scale
b
o
x
:
s
c
a
l
e
=
√
s
c
a
l
e
∗
s
c
a
l
e
at
n
e
x
t
l
e
ve
l
(
asp
ec
t r
atio
=1
)
(
8
)
W
h
er
e
w1
:
wid
th
o
f
b
o
u
n
d
ar
y
b
o
x
,
h
1
:
h
eig
h
t
o
f
b
o
u
n
d
ar
y
b
o
x
.
Sk
ip
p
in
g
c
o
n
n
ec
ti
o
n
in
r
esn
et
as
u
s
ed
i
n
Alg
o
r
ith
m
1
m
o
d
el
a
r
ch
itectu
r
e:
L
in
ea
r
lay
er
1
:
+
1
=
+
1
∗
+
+
1
R
eL
u
o
p
er
atio
n
o
n
lay
e
r
1
:
+
1
=
(
+
1
)
L
in
ea
r
lay
er
1
:
+
2
=
+
2
∗
+
+
2
R
eL
u
o
p
er
atio
n
o
n
lay
e
r
1
:
+
2
=
(
+
2
+
)
=
(
+
2
∗
+
+
2
+
)
(
+
2
an
d
+
2
=0
wh
er
e,
s
in
ce
L
2
r
eg
u
lar
izatio
n
is
u
s
ed
)
.
=
(
)
=
(
Sk
ip
p
in
g
+
1
lay
er
)
(
1
2
)
wh
er
e
a:
in
co
m
in
g
r
esid
u
al
n
e
two
r
k
,
l :
lev
el
o
f
lay
er
,
W
,
b
:
weig
h
t d
ec
ay
,
a
n
d
g
(
)
:r
elu
f
u
n
ctio
n
5.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
W
h
ile
ea
r
lier
s
tu
d
ies
h
av
e
ex
p
lo
r
ed
th
e
d
etec
tio
n
o
f
tu
m
o
r
,
th
e
p
r
o
p
o
s
ed
m
o
d
el
ten
d
s
to
d
etec
t
th
e
tu
m
o
r
lo
ca
tio
n
c
o
o
r
d
i
n
ates
as
p
er
th
e
g
iv
e
n
m
eth
o
d
o
l
o
g
y
d
is
cu
s
s
ed
in
p
r
ev
io
u
s
s
ec
tio
n
s
an
d
d
e
r
iv
e
th
e
f
in
al
r
esu
lts
as
d
is
cu
s
s
ed
in
th
is
s
ec
t
io
n
.
Fig
u
r
e
6
(
a)
(
b
r
ain
h
av
in
g
tu
m
o
r
)
is
th
e
o
u
tp
u
t
o
f
th
e
m
o
b
ile
n
et
m
o
d
el
with
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Dete
ctio
n
o
f lo
ca
tio
n
-
s
p
ec
ific in
tr
a
-
cra
n
ia
l b
r
a
in
tu
mo
r
s
(
S
h
o
la
Ush
a
r
a
n
i
)
435
9
8
p
er
ce
n
t a
cc
u
r
ac
y
.
T
h
e
f
ig
u
r
e
was a
cc
ep
tab
le
as
a
te
s
t d
ata
s
et
in
p
u
t in
to
th
e
m
ac
h
in
e
lear
n
in
g
m
o
d
el
wh
ich
p
r
o
d
u
ce
d
th
e
class
if
icatio
n
r
esu
lt
o
f
y
es
o
r
n
o
b
ased
o
n
t
h
e
p
r
esen
ce
o
f
tu
m
o
r
in
t
h
e
p
ictu
r
e,
s
h
o
win
g
h
o
w
t
h
e
f
o
llo
win
g
r
esu
lt
y
ield
e
d
th
e
f
o
llo
w
in
g
r
esu
lts
.
Fig
u
r
e
6
(
b
)
i
s
th
e
o
u
tp
u
t
o
f
th
e
r
ec
ta
n
g
u
lar
b
o
x
e
n
co
m
p
ass
in
g
th
e
tu
m
o
r
w
h
ich
p
a
r
ticu
lar
ly
was
m
o
s
tly
y
ield
ed
b
y
ef
f
icie
n
t
n
et
lite
0
m
o
d
el
with
9
6
p
e
r
ce
n
t
ac
cu
r
ac
y
.
T
h
e
o
u
tp
u
t
also
in
co
r
p
o
r
ate
d
th
e
g
iv
en
x
m
l
co
o
r
d
in
ates
o
f
th
e
tu
m
o
r
lo
ca
tio
n
i
n
t
h
e
p
ictu
r
e.
Fi
g
u
r
e
s
7
(
a)
a
n
d
7
(
b
)
d
escr
ib
e
th
e
m
o
d
el
ac
cu
r
ac
y
o
f
test
an
d
v
alid
ate
d
ata
s
et
r
e
s
p
ec
tiv
ely
with
tim
e
h
o
w
t
h
e
m
o
d
el
test
ed
o
n
th
e
test
d
ata
s
et
an
d
v
alid
ate
d
ata
s
et
an
d
h
o
w
its
ac
cu
r
ac
y
in
cr
ea
s
ed
in
a
m
ajo
r
way
.
Fig
u
r
e
s
8
(
a)
an
d
8
(
b
)
d
escr
ib
e
th
e
ep
o
ch
lo
s
s
d
ec
r
ea
s
e
with
tim
e
f
o
r
t
h
e
m
o
d
el
with
tim
e
f
o
r
th
e
test
an
d
v
alid
atio
n
d
ata
s
et
r
esp
ec
tiv
ely
.
(
a)
(
b
)
Fig
u
r
e
6
.
Dete
ctio
n
o
f
tu
m
o
r
(
a)
o
u
tp
u
t
b
y
M
o
b
ileNet
V2
an
d
(
b
)
o
u
tp
u
t b
y
E
f
f
icien
tNet
m
o
d
el
(
a)
(
b
)
Fig
u
r
e
7
.
E
v
alatio
n
o
f
ac
cu
r
ac
y
b
ased
o
n
(
a)
tim
e
a
n
d
(
b
)
ite
r
atio
n
s
(
a)
(
b
)
Fig
u
r
e
8
.
E
v
alu
atio
n
o
f
lo
s
s
b
a
s
ed
o
n
(
a)
iter
atio
n
an
d
(
b
)
tim
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
428
-
4
3
8
436
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
was
th
en
co
m
p
ar
e
d
with
C
NN,
ML
PNN,
R
C
NN,
an
d
PF
m
o
d
els
f
r
o
m
th
e
liter
atu
r
e
s
u
r
v
e
y
m
e
n
tio
n
ed
i
n
an
ea
r
lier
s
ec
tio
n
.
A
c
o
m
p
ar
ativ
e
s
tu
d
y
f
o
r
th
e
ab
o
v
e
a
lg
o
r
ith
m
s
with
t
h
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
o
f
th
e
p
a
p
er
was
d
o
n
e
as
m
en
tio
n
ed
in
Fig
u
r
e
9
an
d
T
ab
le
1
.
Fig
u
r
e
9
d
en
o
tes
th
e
m
o
d
el
m
etr
ics
in
b
ar
ch
a
r
t
co
m
p
ar
is
o
n
f
o
r
m
at
an
d
T
ab
le
1
i
n
tab
u
lar
f
o
r
m
at
with
ac
c
u
r
ac
y
an
d
av
er
ag
e
p
r
ec
is
io
n
as
m
etr
ics
to
co
m
p
ar
e.
W
h
ich
it
i
s
ev
id
en
t
th
at
E
f
f
icien
tNet
lite
0
o
b
tain
ed
t
h
e
h
ig
h
est
ac
cu
r
ac
y
th
an
o
th
e
r
m
o
d
els,
wh
ile
it
f
ailed
t
o
ac
h
i
ev
e
th
e
h
ig
h
est
av
er
ag
e
p
r
ec
is
io
n
an
d
m
ea
n
av
e
r
ag
e
p
r
ec
is
io
n
s
in
ce
i
t’
s
a
cu
s
to
m
-
tailo
r
e
d
o
b
ject
d
etec
tio
n
m
o
d
el.
Hen
ce
th
e
g
iv
en
r
esear
c
h
wo
r
k
f
i
n
ally
co
n
cl
u
d
es
o
n
th
e
s
elec
tio
n
o
f
E
f
f
icien
t
N
et
lite
0
as
th
e
s
u
itab
le
m
o
d
el
f
o
r
co
o
r
d
in
ate
lo
ca
tio
n
s
d
etec
tio
n
b
as
ed
o
n
t
h
e
f
in
d
in
g
s
ab
o
v
e.
I
t
a
ls
o
co
n
clu
d
es
SS
D
R
esNet
b
ein
g
a
ju
s
t selec
tio
n
f
o
r
th
e
tu
m
o
r
d
etec
tio
n
p
r
o
ce
s
s
b
ased
o
n
t
h
e
m
etr
ics d
is
cu
s
s
ed
ab
o
v
e.
Fig
u
r
e
9
.
C
o
m
p
a
r
is
o
n
c
h
a
r
t b
e
twee
n
d
if
f
er
en
t CNN m
o
d
els
T
ab
le
1
.
C
o
m
p
a
r
is
o
n
t
ab
le
M
o
d
e
l
A
v
e
r
a
g
e
p
r
e
c
i
s
i
o
n
A
c
c
u
r
a
c
y
M
e
a
n
a
v
e
r
a
g
e
p
r
e
c
i
s
i
o
n
Ef
f
i
c
i
e
n
t
_
n
e
t
_
l
i
t
e
0
87
96
83
M
LPN
N
90
95
87
R
C
N
N
N
92
98
85
C
N
N
88
98
78
PF
80
95
82
Ho
wev
er
,
th
e
em
er
g
e
n
ce
o
f
m
o
r
e
ef
f
icien
t
n
eu
r
al
n
etwo
r
k
m
o
d
els
in
th
e
n
ea
r
f
u
tu
r
e
co
u
l
d
lead
to
an
im
p
r
o
v
e
d
cu
s
to
m
o
b
ject
d
ete
ctio
n
alg
o
r
ith
m
.
W
h
ile
th
e
p
r
o
p
o
s
ed
m
o
d
el
is
r
ec
o
m
m
en
d
e
d
f
o
r
ac
h
ie
v
in
g
th
e
r
esear
ch
p
ap
er
'
s
o
b
jectiv
es
.
I
ts
p
o
ten
tial
b
en
ef
its
to
th
e
m
ed
ical
in
d
u
s
tr
y
m
ay
ex
p
an
d
f
u
r
th
er
with
th
e
in
teg
r
atio
n
o
f
m
o
r
e
ef
f
ec
tiv
e
m
o
d
els in
th
e
f
u
tu
r
e.
6.
CO
NCLU
SI
O
N
T
h
e
o
b
jectiv
e
o
f
t
h
is
s
tu
d
y
is
to
id
en
tify
th
e
co
o
r
d
in
ates
o
f
in
tr
a
-
cr
a
n
ial
b
r
ain
tu
m
o
r
s
u
s
in
g
two
m
ac
h
in
e
lear
n
in
g
m
o
d
els
.
T
h
e
s
e
m
o
d
els
an
aly
ze
X
-
r
a
y
d
ata
t
o
p
in
p
o
in
t
tu
m
o
r
r
e
g
io
n
s
.
I
n
itially
,
th
e
m
o
b
ile
n
et
-
v
m
o
d
el
s
u
cc
ess
f
u
lly
d
etec
ts
tu
m
o
r
p
r
esen
ce
,
f
o
ll
o
wed
b
y
th
e
ap
p
licatio
n
o
f
th
e
ef
f
icien
t
N
et
lite
0
m
o
d
el,
wh
ich
id
en
tifie
s
an
n
o
tatio
n
b
o
x
es,
p
in
p
o
in
tin
g
tu
m
o
r
lo
ca
tio
n
c
o
o
r
d
in
ates
th
r
o
u
g
h
im
ag
e
lab
eli
n
g
.
Ultim
ately
,
th
is
r
esear
ch
p
r
esen
ts
a
b
r
ain
tu
m
o
r
d
etec
tio
n
m
o
d
el
p
o
is
ed
to
b
en
ef
it
th
e
m
ed
ical
co
m
m
u
n
ity
b
y
en
ab
lin
g
ea
r
l
y
tu
m
o
r
d
etec
tio
n
.
M
o
v
in
g
f
o
r
war
d
,
th
e
aim
is
to
en
h
an
ce
th
e
p
r
o
p
o
s
ed
m
o
d
el
b
y
in
t
eg
r
atin
g
b
r
ain
s
ize
d
im
en
s
io
n
s
to
ex
p
ed
ite
tu
m
o
r
d
etec
tio
n
an
d
in
cr
ea
s
e
ac
cu
r
ac
y
in
ea
r
ly
-
s
tag
e
d
iag
n
o
s
is
.
RE
F
E
R
E
NC
E
S
[
1
]
J.
A
mi
n
,
M
.
S
h
a
r
i
f
,
M
.
Y
a
s
mi
n
,
a
n
d
S
.
L.
F
e
r
n
a
n
d
e
s,
“
B
i
g
d
a
t
a
a
n
a
l
y
si
s
f
o
r
b
r
a
i
n
t
u
m
o
r
d
e
t
e
c
t
i
o
n
:
D
e
e
p
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,”
F
u
t
u
re
G
e
n
e
r
a
t
i
o
n
C
o
m
p
u
t
e
r
S
y
s
t
e
m
s
,
v
o
l
.
8
7
,
p
p
.
2
9
0
–
2
9
7
,
2
0
1
8
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
f
u
t
u
r
e
.
2
0
1
8
.
0
4
.
0
6
5
.
0
20
40
60
80
100
120
Co
m
p
ari
so
n
Ch
art
A
c
c
u
racy
A
ve
rag
e
P
re
c
i
s
i
o
n
Me
an
A
P
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
Dete
ctio
n
o
f lo
ca
tio
n
-
s
p
ec
ific in
tr
a
-
cra
n
ia
l b
r
a
in
tu
mo
r
s
(
S
h
o
la
Ush
a
r
a
n
i
)
437
[
2
]
P
.
M
.
S
h
a
k
e
e
l
,
T.
E
.
E.
T
o
b
e
l
y
,
H
.
A
l
-
F
e
e
l
,
G
.
M
a
n
o
g
a
r
a
n
,
a
n
d
S
.
B
a
s
k
a
r
,
“
N
e
u
r
a
l
n
e
t
w
o
r
k
b
a
se
d
b
r
a
i
n
t
u
m
o
r
d
e
t
e
c
t
i
o
n
u
si
n
g
w
i
r
e
l
e
s
s i
n
f
r
a
r
e
d
i
ma
g
i
n
g
se
n
s
o
r
,
”
I
E
EE
A
c
c
e
ss
,
v
o
l
.
7
,
p
p
.
5
5
7
7
–
5
5
8
8
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
1
8
.
2
8
8
3
9
5
7
.
[
3
]
H
.
F
a
b
e
l
o
e
t
a
l
.
,
“
I
n
-
v
i
v
o
h
y
p
e
r
s
p
e
c
t
r
a
l
h
u
ma
n
b
r
a
i
n
i
m
a
g
e
d
a
t
a
b
a
s
e
f
o
r
b
r
a
i
n
c
a
n
c
e
r
d
e
t
e
c
t
i
o
n
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
7
,
p
p
.
3
9
0
9
8
–
3
9
1
1
6
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
1
9
.
2
9
0
4
7
8
8
.
[
4
]
J.
A
m
i
n
,
M
.
S
h
a
r
i
f
,
M
.
R
a
z
a
,
T
.
S
a
b
a
,
a
n
d
M
.
A
.
A
n
j
u
m,
“
B
r
a
i
n
t
u
mo
r
d
e
t
e
c
t
i
o
n
u
si
n
g
s
t
a
t
i
s
t
i
c
a
l
a
n
d
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
me
t
h
o
d
,
”
C
o
m
p
u
t
e
r
M
e
t
h
o
d
s
a
n
d
Pr
o
g
r
a
m
s
i
n
Bi
o
m
e
d
i
c
i
n
e
,
v
o
l
.
1
7
7
,
p
p
.
6
9
–
7
9
,
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
m
p
b
.
2
0
1
9
.
0
5
.
0
1
5
.
[
5
]
N
.
K
e
sa
v
a
n
d
M
.
G
.
J
i
b
u
k
u
mar,
“
Ef
f
i
c
i
e
n
t
a
n
d
l
o
w
c
o
m
p
l
e
x
a
r
c
h
i
t
e
c
t
u
r
e
f
o
r
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
o
f
b
r
a
i
n
t
u
mo
r
u
s
i
n
g
R
C
N
N
w
i
t
h
t
w
o
c
h
a
n
n
e
l
C
N
N
,
”
J
o
u
r
n
a
l
o
f
K
i
n
g
S
a
u
d
U
n
i
v
e
rsi
t
y
-
C
o
m
p
u
t
e
r
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
c
i
e
n
c
e
s
,
v
o
l
.
3
4
,
n
o
.
8
,
p
p
.
6
2
2
9
–
6
2
4
2
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
k
su
c
i
.
2
0
2
1
.
0
5
.
0
0
8
.
[
6
]
M
.
G
h
a
f
f
a
r
i
,
A
.
S
o
w
my
a
,
a
n
d
R
.
O
l
i
v
e
r
,
“
A
u
t
o
ma
t
e
d
b
r
a
i
n
t
u
m
o
r
se
g
m
e
n
t
a
t
i
o
n
u
s
i
n
g
mu
l
t
i
m
o
d
a
l
b
r
a
i
n
s
c
a
n
s:
a
s
u
r
v
e
y
b
a
se
d
o
n
mo
d
e
l
s
s
u
b
mi
t
t
e
d
t
o
t
h
e
B
r
a
TS
2
0
1
2
–
2
0
1
8
c
h
a
l
l
e
n
g
e
s
,
”
I
EEE
Re
v
i
e
w
s
i
n
B
i
o
m
e
d
i
c
a
l
En
g
i
n
e
e
r
i
n
g
,
v
o
l
.
1
3
,
p
p
.
1
5
6
–
1
6
8
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
R
B
M
E.
2
0
1
9
.
2
9
4
6
8
6
8
.
[
7
]
S
.
I
r
sh
e
i
d
a
t
a
n
d
R
.
D
u
w
a
i
r
i
,
“
B
r
a
i
n
t
u
mo
r
d
e
t
e
c
t
i
o
n
u
si
n
g
a
r
t
i
f
i
c
i
a
l
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
2
0
2
0
1
1
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
I
n
f
o
rm
a
t
i
o
n
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
S
y
s
t
e
m
s
(
I
C
I
C
S
)
,
2
0
2
0
,
p
p
.
1
9
7
–
2
0
3
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
I
C
S
4
9
4
6
9
.
2
0
2
0
.
2
3
9
5
2
2
.
[
8
]
A
.
K
u
mar
,
M
.
R
a
ma
c
h
a
n
d
r
a
n
,
A
.
H
.
G
a
n
d
o
m
i
,
R
.
P
a
t
a
n
,
S
.
L
u
k
a
s
i
k
,
a
n
d
R
.
K
.
S
o
u
n
d
a
r
a
p
a
n
d
i
a
n
,
“
A
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
b
a
se
d
c
l
a
ss
i
f
i
e
r
f
o
r
b
r
a
i
n
t
u
m
o
r
d
i
a
g
n
o
s
i
s,
”
Ap
p
l
i
e
d
S
o
f
t
C
o
m
p
u
t
i
n
g
,
v
o
l
.
8
2
,
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
s
o
c
.
2
0
1
9
.
1
0
5
5
2
8
.
[
9
]
P
y
t
o
r
c
h
T
e
a
m
,
“
M
o
b
i
l
e
N
e
t
v
2
:
Ef
f
i
c
i
e
n
t
n
e
t
w
o
r
k
s
o
p
t
i
m
i
z
e
d
f
o
r
s
p
e
e
d
a
n
d
m
e
mo
r
y
,
w
i
t
h
r
e
si
d
u
a
l
b
l
o
c
k
s,
”
L
i
n
u
x
F
o
u
n
d
a
t
i
o
n
.
2
0
2
4
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s:
/
/
p
y
t
o
r
c
h
.
o
r
g
/
h
u
b
/
p
y
t
o
r
c
h
_
v
i
si
o
n
_
m
o
b
i
l
e
n
e
t
_
v
2
/
[
1
0
]
T.
A
.
S
o
o
mr
o
e
t
a
l
.
,
“
I
mag
e
s
e
g
me
n
t
a
t
i
o
n
f
o
r
M
R
b
r
a
i
n
t
u
m
o
r
d
e
t
e
c
t
i
o
n
u
si
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
:
a
r
e
v
i
e
w
,
”
I
EEE
Re
v
i
e
w
s
i
n
Bi
o
m
e
d
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
,
v
o
l
.
1
6
,
p
p
.
7
0
–
9
0
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
R
B
M
E.
2
0
2
2
.
3
1
8
5
2
9
2
.
[
1
1
]
S
.
G
u
l
l
a
n
d
S
.
A
k
b
a
r
,
“
A
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
i
n
b
r
a
i
n
t
u
m
o
r
d
e
t
e
c
t
i
o
n
t
h
r
o
u
g
h
M
R
I
sca
n
s,
”
i
n
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
I
n
t
e
r
n
e
t
o
f
T
h
i
n
g
s
,
1
st
e
d
.
,
F
l
o
r
i
d
a
,
U
S
:
C
R
C
P
r
e
ss,
2
0
2
1
.
[
1
2
]
A
.
C
h
a
t
t
o
p
a
d
h
y
a
y
a
n
d
M
.
M
a
i
t
r
a
,
“
M
R
I
-
b
a
se
d
b
r
a
i
n
t
u
mo
u
r
i
ma
g
e
d
e
t
e
c
t
i
o
n
u
si
n
g
C
N
N
b
a
s
e
d
d
e
e
p
l
e
a
r
n
i
n
g
me
t
h
o
d
,
”
S
m
a
r
t
Ag
ri
c
u
l
t
u
r
a
l
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
,
n
o
.
4
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
r
i
.
2
0
2
2
.
1
0
0
0
6
0
.
[
1
3
]
M
.
S
.
I
.
K
h
a
n
e
t
a
l
.
,
“
A
c
c
u
r
a
t
e
b
r
a
i
n
t
u
m
o
r
d
e
t
e
c
t
i
o
n
u
si
n
g
d
e
e
p
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
,
”
C
o
m
p
u
t
a
t
i
o
n
a
l
a
n
d
S
t
r
u
c
t
u
r
a
l
Bi
o
t
e
c
h
n
o
l
o
g
y
J
o
u
r
n
a
l
,
v
o
l
.
2
0
,
p
p
.
4
7
3
3
–
4
7
4
5
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
s
b
j
.
2
0
2
2
.
0
8
.
0
3
9
.
[
1
4
]
T.
H
o
ssa
i
n
,
F
.
S
.
S
h
i
sh
i
r
,
M
.
A
s
h
r
a
f
,
M
.
D
.
A
.
A
l
N
a
si
m,
a
n
d
F
.
M
.
S
h
a
h
,
“
B
r
a
i
n
t
u
m
o
r
d
e
t
e
c
t
i
o
n
u
s
i
n
g
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
,
”
i
n
2
0
1
9
1
st
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
A
d
v
a
n
c
e
s
i
n
S
c
i
e
n
c
e
,
E
n
g
i
n
e
e
ri
n
g
a
n
d
R
o
b
o
t
i
c
s
T
e
c
h
n
o
l
o
g
y
(
I
C
AS
ERT)
,
2
0
1
9
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
A
S
ER
T
.
2
0
1
9
.
8
9
3
4
5
6
1
.
[
1
5
]
J.
A
mi
n
,
M
.
A
.
A
n
j
u
m,
M
.
S
h
a
r
i
f
,
S
.
Ja
b
e
e
n
,
S
.
K
a
d
r
y
,
a
n
d
P
.
M
.
G
e
r
,
“
A
n
e
w
m
o
d
e
l
f
o
r
b
r
a
i
n
t
u
m
o
r
d
e
t
e
c
t
i
o
n
u
s
i
n
g
e
n
s
e
m
b
l
e
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
a
n
d
q
u
a
n
t
u
m
v
a
r
i
a
t
i
o
n
a
l
c
l
a
ssi
f
i
e
r
,
”
C
o
m
p
u
t
a
t
i
o
n
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
N
e
u
r
o
sc
i
e
n
c
e
,
v
o
l
.
2
0
2
2
,
p
p
.
1
–
1
3
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
2
2
/
3
2
3
6
3
0
5
.
[
1
6
]
M
.
S
i
a
r
a
n
d
M
.
T
e
sh
n
e
h
l
a
b
,
“
B
r
a
i
n
t
u
m
o
r
d
e
t
e
c
t
i
o
n
u
si
n
g
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
a
n
d
mac
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m,
”
i
n
2
0
1
9
9
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
a
n
d
K
n
o
w
l
e
d
g
e
En
g
i
n
e
e
ri
n
g
(
I
C
C
K
E)
,
2
0
1
9
,
p
p
.
3
6
3
–
3
6
8
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
K
E4
8
5
6
9
.
2
0
1
9
.
8
9
6
4
8
4
6
.
[
1
7
]
P
.
G
.
B
r
i
n
d
h
a
,
M
.
K
a
v
i
n
r
a
j
,
P
.
M
a
n
i
v
a
s
a
k
a
m,
a
n
d
P
.
P
r
a
sa
n
t
h
,
“
B
r
a
i
n
t
u
mo
r
d
e
t
e
c
t
i
o
n
f
r
o
m
M
R
I
i
ma
g
e
s
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,
”
I
O
P
C
o
n
f
e
r
e
n
c
e
S
e
r
i
e
s:
Ma
t
e
r
i
a
l
s
S
c
i
e
n
c
e
a
n
d
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
1
0
5
5
,
n
o
.
1
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
8
8
/
1
7
5
7
-
8
9
9
X
/
1
0
5
5
/
1
/
0
1
2
1
1
5
.
[
1
8
]
S
.
K
u
m
a
r
,
R
.
D
h
i
r
,
a
n
d
N
.
C
h
a
u
r
a
si
a
,
“
B
r
a
i
n
t
u
m
o
r
d
e
t
e
c
t
i
o
n
a
n
a
l
y
s
i
s
u
si
n
g
C
N
N
:
A
r
e
v
i
e
w
,
”
i
n
2
0
2
1
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
S
m
a
rt
S
y
st
e
m
s (I
C
AI
S
)
,
2
0
2
1
,
p
p
.
1
0
6
1
–
1
0
6
7
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
A
I
S
5
0
9
3
0
.
2
0
2
1
.
9
3
9
5
9
2
0
.
[
1
9
]
A
.
G
a
n
e
s
h
,
S
.
D
e
p
u
r
u
,
B
.
R
.
A
.
,
a
n
d
G
.
S
u
j
a
t
h
a
,
“
S
t
r
e
a
ml
i
n
i
n
g
c
a
n
c
e
r
d
i
a
g
n
o
s
i
s
a
n
d
p
r
o
g
n
o
s
i
s
sy
s
t
e
m
u
si
n
g
h
y
b
r
i
d
C
N
N
-
N
P
R
:
d
e
e
p
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
e
s
,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
I
n
t
e
l
l
i
g
e
n
t
S
y
st
e
m
s
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s i
n
E
n
g
i
n
e
e
ri
n
g
,
v
o
l
.
1
2
,
n
o
.
3
s,
p
p
.
1
9
0
–
2
0
1
,
2
0
2
3
.
[
2
0
]
G
.
W
a
n
g
,
W
.
Li
,
S
.
O
u
r
s
e
l
i
n
,
a
n
d
T.
V
e
r
c
a
u
t
e
r
e
n
,
“
A
u
t
o
ma
t
i
c
b
r
a
i
n
t
u
m
o
r
seg
m
e
n
t
a
t
i
o
n
b
a
se
d
o
n
c
a
sca
d
e
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 wi
t
h
u
n
c
e
r
t
a
i
n
t
y
e
st
i
ma
t
i
o
n
,
”
Fr
o
n
t
i
e
rs
i
n
C
o
m
p
u
t
a
t
i
o
n
a
l
N
e
u
ro
s
c
i
e
n
c
e
,
v
o
l
.
1
3
,
2
0
1
9
,
d
o
i
:
1
0
.
3
3
8
9
/
f
n
c
o
m.
2
0
1
9
.
0
0
0
5
6
.
[
2
1
]
R
.
R
a
n
j
b
a
r
z
a
d
e
h
,
A
.
B
.
K
a
sg
a
r
i
,
S
.
J
.
G
h
o
u
s
h
c
h
i
,
S
.
A
n
a
r
i
,
M
.
N
a
s
e
r
i
,
a
n
d
M
.
B
e
n
d
e
c
h
a
c
h
e
,
“
B
r
a
i
n
t
u
m
o
r
se
g
m
e
n
t
a
t
i
o
n
b
a
s
e
d
o
n
d
e
e
p
l
e
a
r
n
i
n
g
a
n
d
a
n
a
t
t
e
n
t
i
o
n
m
e
c
h
a
n
i
sm
u
s
i
n
g
M
R
I
mu
l
t
i
-
m
o
d
a
l
i
t
i
e
s
b
r
a
i
n
i
ma
g
e
s
,
”
S
c
i
e
n
t
i
f
i
c
Re
p
o
rt
s
,
v
o
l
.
1
1
,
n
o
.
1
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
9
8
-
0
2
1
-
9
0
4
2
8
-
8.
[
2
2
]
M
.
La
t
h
e
r
a
n
d
P
.
S
i
n
g
h
,
“
I
n
v
e
s
t
i
g
a
t
i
n
g
b
r
a
i
n
t
u
m
o
r
se
g
me
n
t
a
t
i
o
n
a
n
d
d
e
t
e
c
t
i
o
n
t
e
c
h
n
i
q
u
e
s
,
”
Pr
o
c
e
d
i
a
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
1
6
7
,
p
p
.
1
2
1
–
1
3
0
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
r
o
c
s.2
0
2
0
.
0
3
.
1
8
9
.
[
2
3
]
M
.
T
a
n
,
R
.
P
a
n
g
,
a
n
d
Q
.
V
.
Le
,
“
Ef
f
i
c
i
e
n
t
D
e
t
:
sc
a
l
a
b
l
e
a
n
d
e
f
f
i
c
i
e
n
t
o
b
j
e
c
t
d
e
t
e
c
t
i
o
n
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
I
E
EE/
C
VF
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
Vi
s
i
o
n
a
n
d
P
a
t
t
e
r
n
R
e
c
o
g
n
i
t
i
o
n
(
C
VP
R)
,
2
0
2
0
,
p
p
.
1
0
7
8
1
-
1
0
7
9
0
[
2
4
]
M
.
S
a
n
d
l
e
r
,
A
.
H
o
w
a
r
d
,
M
.
Z
h
u
,
A
.
Z
h
m
o
g
i
n
o
v
,
a
n
d
L
.
-
C
.
C
h
e
n
,
“
M
o
b
i
l
e
n
e
t
v
2
:
I
n
v
e
r
t
e
d
r
e
s
i
d
u
a
l
s
a
n
d
l
i
n
e
a
r
b
o
t
t
l
e
n
e
c
k
s
,
”
i
n
2
0
1
8
I
E
E
E
/
C
V
F
C
o
n
f
e
r
e
n
c
e
o
n
C
o
m
p
u
t
e
r
V
i
s
i
o
n
a
n
d
P
a
t
t
e
r
n
R
e
c
o
g
n
i
t
i
o
n
,
2
0
1
8
,
p
p
.
4
5
1
0
–
4
5
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
.
2
0
1
8
.
0
0
4
7
4
.
[
2
5
]
M
.
T
a
n
,
R
.
P
a
n
g
,
a
n
d
Q
.
V
L
e
,
“
Ef
f
i
c
i
e
n
t
D
e
t
:
sc
a
l
a
b
l
e
a
n
d
e
f
f
i
c
i
e
n
t
o
b
j
e
c
t
d
e
t
e
c
t
i
o
n
,
”
i
n
2
0
2
0
I
EEE
/
C
VF
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
Vi
si
o
n
a
n
d
Pa
t
t
e
rn
Re
c
o
g
n
i
t
i
o
n
(
C
V
P
R)
,
2
0
2
0
,
p
p
.
1
0
7
7
8
–
1
0
7
8
7
,
d
o
i
:
1
0
.
1
1
0
9
/
C
V
P
R
4
2
6
0
0
.
2
0
2
0
.
0
1
0
7
9
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Dr
.
S
h
o
la
Us
h
a
r
a
n
i
is
c
u
rre
n
tl
y
a
ffil
iate
d
wit
h
Ve
ll
o
re
In
st
it
u
te
o
f
Tec
h
n
o
l
o
g
y
(VIT),
Ch
e
n
n
a
i
a
s
As
so
c
iate
P
r
o
fe
ss
o
r
-
G
ra
d
e
II
in
th
e
S
c
h
o
o
l
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
i
n
e
e
rin
g
(
S
COPE
).
S
h
e
re
c
e
iv
e
d
h
e
r
P
h
.
D
.
in
Ve
ll
o
re
In
sti
t
u
te
o
f
Tec
h
n
o
lo
g
y
(VIT),
Ch
e
n
n
a
i
i
n
2
0
2
0
u
n
d
e
r
h
e
a
lt
h
c
a
re
c
o
m
p
u
ti
n
g
.
He
r
sp
e
c
ializa
ti
o
n
d
o
m
a
in
s
in
c
lu
d
e
c
lo
u
d
se
c
u
rit
y
,
in
fo
rm
a
ti
o
n
,
a
n
d
c
y
b
e
r
se
c
u
rit
y
,
Io
T,
a
n
d
m
a
c
h
in
e
lea
rn
i
n
g
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
sh
o
lau
sh
a
.
ra
n
i@v
i
t.
a
c
.
in
.
Evaluation Warning : The document was created with Spire.PDF for Python.