I
n
t
e
r
n
at
ion
al
Jou
r
n
al
of
E
lec
t
r
ical
an
d
Com
p
u
t
e
r
E
n
gin
e
e
r
in
g
(
I
JE
CE
)
Vol.
1
4
,
No.
5
,
Oc
tober
20
2
4
,
pp
.
5200
~
5208
I
S
S
N:
2088
-
8708
,
DO
I
:
10
.
11591/i
jec
e
.
v
1
4
i
5
.
pp
5
200
-
5208
5200
Jou
r
n
al
h
omepage
:
ht
tp:
//
ij
e
c
e
.
iaes
c
or
e
.
c
om
B
lo
od
gl
u
c
ose
p
r
e
d
ic
t
io
n
u
si
n
g n
on
-
i
n
vasi
ve
op
t
ic
al
syste
m
b
ase
d
o
n
p
h
ot
o
p
le
t
h
ysm
og
r
a
p
h
y
M
oh
am
m
e
d
Anes
B
e
r
e
k
s
i
Re
gu
ig
1
,2
,
Nas
s
im
a
L
ab
d
e
ll
i
1
,
2
,
3
1
D
e
pa
r
tm
e
nt
of
B
io
me
di
c
a
l
E
ngi
ne
e
r
in
g, F
a
c
ul
ty
of
T
e
c
hnol
og
y
, A
bou
-
B
e
kr
B
e
lk
a
id
U
ni
ve
r
s
it
y, T
le
m
c
e
n, A
lg
e
r
ia
2
R
e
s
e
a
r
c
h L
a
bor
a
to
r
y of
B
io
me
di
c
a
l
E
ngi
ne
e
r
in
g, A
bou
-
B
e
kr
B
e
lk
a
id
U
ni
ve
r
s
it
y, T
le
m
c
e
n, A
lg
e
r
ia
3
H
ig
he
r
S
c
hool
of
A
ppl
ie
d S
c
ie
nc
e
s
,
T
le
mc
e
n
,
A
lg
e
r
ia
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
M
a
r
15,
2024
R
e
vis
e
d
J
un
20,
2024
Ac
c
e
pted
J
ul
2,
2024
Sev
eral
p
eo
p
l
e
mu
s
t
fre
q
u
e
n
t
l
y
e
v
al
u
at
e
t
h
e
i
r
b
l
o
o
d
g
l
u
c
o
s
e
s
i
n
ce
i
t
i
s
a
n
i
mp
o
rt
a
n
t
i
n
d
i
cat
o
r
o
f
h
eal
t
h
p
r
o
b
l
ems
mai
n
l
y
d
i
a
b
et
e
s
.
D
i
fferen
t
med
i
ca
l
s
y
s
t
ems
are
co
mmerc
i
al
i
zed
t
o
meas
u
re
b
l
o
o
d
g
l
u
c
o
s
e
l
ev
e
l
s
;
s
o
me
are
i
n
v
as
i
v
e
o
t
h
ers
are
n
o
n
i
n
v
as
i
v
e.
T
h
e
ma
i
n
p
u
r
p
o
s
e
o
f
t
h
i
s
art
i
cl
e
i
s
t
o
d
ev
e
l
o
p
a
n
o
n
-
i
n
v
as
i
v
e
d
e
v
i
ce
fo
r
meas
u
r
i
n
g
b
l
o
o
d
g
l
u
co
s
e
l
e
v
el
s
b
as
e
d
o
n
t
h
e
d
e
t
ect
i
o
n
an
d
an
al
y
s
i
s
o
f
t
h
e
p
h
o
t
o
p
l
et
h
y
s
mo
g
ram
s
i
g
n
a
l
.
T
h
e
d
e
v
el
o
p
e
d
s
y
s
t
ems
i
n
c
l
u
d
e
an
o
p
t
i
ca
l
s
e
n
s
o
r
t
o
d
e
t
ect
t
h
e
p
h
o
t
o
p
l
et
h
y
s
m
o
g
r
ap
h
y
(PPG
)
s
i
g
n
a
l
,
d
i
g
i
t
a
l
i
z
i
n
g
an
d
acq
u
i
r
i
n
g
b
o
ar
d
s
t
o
a
co
mp
u
t
er
an
d
a
s
o
ft
w
are
p
ro
g
ram
t
o
p
ro
ce
s
s
a
n
d
a
n
al
y
ze
t
h
e
d
i
g
i
t
al
i
zed
PPG
s
i
g
n
a
l
reg
ard
i
n
g
s
o
me
feat
u
re
s
ex
t
ract
e
d
fro
m
i
t
s
w
av
ef
o
rm.
T
h
e
s
e
fea
t
u
r
es
are
t
h
e
s
y
s
t
o
l
i
c
amp
l
i
t
u
d
e
an
d
t
h
e
/
amp
l
i
t
u
d
e
rat
i
o
i
n
t
h
e
s
eco
n
d
d
eri
v
a
t
i
v
e
PPG
(SD
PPG
)
w
av
ef
o
rm.
A
n
i
n
v
a
s
i
v
e
g
l
u
c
o
met
er
i
s
al
s
o
u
s
ed
al
o
n
g
w
i
t
h
t
h
e
S
a
an
d
b
/
a
ra
t
i
o
d
e
t
ermi
n
ed
fr
o
m
t
h
e
d
ev
e
l
o
p
ed
s
y
s
t
em
t
o
g
en
erat
e
a
ca
l
i
b
rat
i
o
n
mo
d
e
l
w
h
i
ch
i
s
u
s
ed
t
o
d
e
d
u
ce
b
l
o
o
d
g
l
u
co
s
e
l
ev
e
l
(BG
L
)
v
al
u
es
.
T
h
e
r
es
u
l
t
s
h
o
w
e
d
t
h
a
t
t
h
e
cal
i
b
ra
t
i
o
n
mo
d
e
l
u
s
i
n
g
t
h
e
b
/
a
rat
i
o
i
s
mo
re
accu
ra
t
e
fo
r
non
-
i
n
v
a
s
i
v
e
b
l
o
o
d
l
ev
e
l
meas
u
reme
n
t
t
h
en
t
h
at
o
f
Sa
w
i
t
h
a
d
i
ffere
n
ce
i
n
g
l
u
co
s
e
es
t
i
mat
i
o
n
aro
u
n
d
2
m
g
/
d
l
an
d
w
i
t
h
t
h
e
co
rre
l
a
t
i
o
n
co
eff
i
ci
e
n
t
(R
2
)
o
f
t
h
e
g
l
u
co
s
e
l
ev
e
l
p
re
d
i
c
t
i
o
n
b
e
t
w
e
en
0
.
8
9
0
4
an
d
0
.
9
7
7
5
.
K
e
y
w
o
r
d
s
:
Dia
be
tes
Non
-
invas
ive
opti
c
a
l
s
e
n
s
or
blood
glucos
e
moni
tor
ing
P
hotopl
e
thys
mogr
a
m
s
ignal
P
hys
iol
ogica
l
s
ignal
S
ignal
pr
oc
e
s
s
ing
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
:
M
oha
mm
e
d
Ane
s
B
e
r
e
ks
i
R
e
guig
De
pa
r
tm
e
nt
of
B
iom
e
dica
l
E
nginee
r
ing,
F
a
c
ult
y
of
T
e
c
hnology
,
Abou
-
B
e
kr
B
e
lkaid
Unive
r
s
it
y
T
lemc
e
n,
Alge
r
ia
E
mail:
mohammed_a
ne
s
.
be
r
e
ks
i_r
e
guig@um
mt
o.
dz
1.
I
NT
RODU
C
T
I
ON
T
he
human
body
us
e
s
glucos
e
a
s
a
s
igni
f
ica
nt
s
our
c
e
of
e
ne
r
gy.
T
o
c
a
r
r
y
ou
t
da
il
y
tas
ks
,
the
body
ne
e
ds
blood
glucos
e
leve
ls
in
the
nor
mal
r
a
nge
(
80
to
150
mg/
dl)
[
1]
.
How
e
ve
r
,
a
higher
or
lowe
r
glucos
e
leve
l
c
ould
r
e
s
ult
in
a
numbe
r
o
f
int
e
r
na
l
c
ompl
ica
ti
ons
.
I
n
a
ddit
ion
,
the
vit
a
l
hor
mone
ins
uli
n
is
pr
od
uc
e
d
by
the
body
a
s
a
r
e
s
ult
of
e
a
ti
ng
.
Ge
ne
r
a
ll
y
high
bloo
d
glucos
e
c
onc
e
ntr
a
ti
on
is
pos
s
ibl
e
if
the
c
r
e
a
ti
on
of
α
c
e
ll
s
is
gr
e
a
ter
a
s
c
ompar
e
d
to
that
of
the
β
c
e
ll
s
.
B
e
c
a
us
e
of
thi
s
c
ondit
ion,
e
nough
ins
uli
n
is
not
s
e
c
r
e
ted
in
the
body
f
or
g
lucos
e
c
ons
umpt
ion.
T
his
c
ondit
ion
make
s
mention
a
diabe
tes
melli
tus
[
2]
.
Dia
be
t
e
s
is
a
n
incur
a
ble
il
lnes
s
in
whic
h
the
body
is
una
ble
to
m
a
int
a
in
blood
g
lucos
e
leve
ls
withi
n
the
nor
mal
r
a
nge
(
5
to
8
mi
lm
ol/
l)
[
3]
.
T
he
I
nter
na
ti
ona
l
Dia
be
tes
F
e
de
r
a
ti
on
(
I
DF
)
e
s
ti
mate
s
that
387
mi
ll
ion
pe
ople
will
ha
ve
diabe
tes
by
2035
[
4]
.
T
he
diabe
tes
c
onduc
t
to
dif
f
e
r
e
nt
he
a
lt
h
pr
ob
lems
,
e
.
g.
c
a
r
diovas
c
ular
dis
e
a
s
e
s
,
a
ther
os
c
ler
os
is
pr
oblems
,
c
hr
onic
r
e
na
l
dis
e
a
s
e
(
C
KD
)
,
ne
r
vous
s
ys
tem
dis
e
a
s
e
s
,
f
oot
a
mput
a
ti
o
n
due
to
ulce
r
a
ti
on,
a
nd
e
a
r
ly
de
a
th
[
5
]
.
S
ince
mos
t
c
omm
e
r
c
ially
a
va
il
a
ble
glucos
e
tes
ts
a
r
e
now
unc
omf
or
t
a
ble
a
nd
c
a
r
r
y
a
r
is
k
o
f
inf
e
c
ti
on
,
invas
ive
mea
s
ur
e
ment
tec
hniques
ha
ve
be
e
n
us
e
d
the
mos
t
f
r
e
que
ntl
y
wor
ld
wide
[
6]
.
Ne
ve
r
thele
s
s
,
numer
ous
c
omm
e
r
c
ial
de
vice
s
f
or
c
onti
nuous
blood
glucos
e
moni
tor
ing
uti
li
z
e
e
lec
tr
o
c
he
mi
c
a
l
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
E
lec
&
C
omp
E
ng
I
S
S
N:
2088
-
8708
B
lood
glucos
e
pr
e
diction
us
ing
non
-
invas
ive
opti
c
al
s
y
s
tem
bas
e
d
on
…
(
M
ohamm
e
d
A
ne
s
B
e
r
e
k
s
i
R
e
guig
)
5201
s
e
ns
or
s
[
7]
.
W
hich
pr
e
s
e
nt
a
high
r
e
s
pons
e
f
or
g
lucos
e
de
tec
ti
on
in
blood
[
8]
.
F
or
pr
icking
the
bl
ood,
the
lanc
e
ts
a
r
e
us
e
d
in
the
pr
im
a
r
y
s
tage
f
or
blood
glucos
e
moni
tor
ing
f
or
many
c
omm
e
r
c
ial
de
vice
s
a
va
il
a
ble
in
the
mar
ke
t
[
9]
.
As
a
r
e
s
ult
o
f
taking
blood
s
a
mpl
e
s
f
r
om
the
f
inger
ti
p
mor
e
than
twice
a
da
y
f
or
f
r
e
que
nt
moni
tor
ing,
the
pr
a
c
ti
c
e
o
f
r
e
pe
a
ted
mea
s
ur
e
ment
is
quit
e
pa
inf
ul
[
10
]
.
F
o
r
de
c
a
de
s
,
di
f
f
e
r
e
nt
non
-
invas
ive
methods
a
r
e
e
f
f
e
c
ti
ve
s
ubs
ti
tut
e
s
f
or
thes
e
invas
i
ve
methods
be
c
a
us
e
pa
ti
e
nt
s
will
f
ind
it
mor
e
c
onve
nient,
plea
s
a
nt,
a
nd
a
f
f
or
da
ble.
T
he
r
e
a
r
e
numer
ous
no
ninvas
ive
methods
f
or
mea
s
ur
ing
blood
glucos
e
t
ha
t
ha
ve
be
e
n
pr
opos
e
d
[
11]
.
C
hu
e
t
al
.
[
12]
de
ve
loped
a
c
omput
a
ti
ona
l
m
ode
l
ba
s
e
d
on
e
ne
r
gy
c
ons
e
r
va
ti
on
pr
inciples
,
int
e
gr
a
ti
ng
phys
iol
ogica
l
da
ta
li
ke
he
a
r
t
r
a
te,
oxyg
e
n
s
a
tur
a
ti
on,
a
nd
blood
f
low
ve
locity
.
T
he
y
a
ls
o
d
e
vis
e
d
a
ha
r
dwa
r
e
s
ys
tem
f
o
r
da
ta
c
oll
e
c
ti
on,
including
a
dua
l
-
wa
ve
length
photopl
e
thys
mogr
a
phy
(
P
P
G)
a
c
quis
it
ion
s
e
tup.
Ha
bbu
e
t
al.
[
13]
int
r
oduc
e
d
a
ne
ur
a
l
ne
tw
or
k
model
us
ing
s
ingl
e
puls
e
a
na
lys
is
a
nd
f
r
e
que
nc
y
-
ti
me
f
e
a
tur
e
s
f
r
om
P
P
G
s
ignals
.
S
a
tt
e
r
e
t
al.
[
14]
uti
li
z
e
d
wr
is
t
P
P
G
s
ignals
to
e
s
ti
mate
blood
glucos
e
leve
l
(
B
GL
)
,
f
oc
us
ing
on
wa
ve
f
or
m
-
ba
s
e
d
f
e
a
tur
e
s
li
ke
puls
a
ti
l
e
pa
r
t
(
AC
)
a
nd
s
tea
dy
pa
r
t
(
DC
)
r
a
ti
o
a
nd
f
e
a
tur
e
s
de
r
ived
f
r
om
e
mpi
r
ica
l
mode
de
c
ompos
it
ion
(
E
M
D)
.
Addi
ti
ona
ll
y,
a
s
tudy
in
[
15]
p
r
opos
e
d
a
non
-
invas
ive
method
to
e
s
ti
mate
HbA
1c
us
ing
P
P
G
s
ignals
with
two
wa
ve
lengths
.
T
he
s
e
a
ppr
oa
c
he
s
de
mons
tr
a
te
the
ongoi
ng
e
f
f
o
r
t
to
de
ve
lop
non
-
invas
ive
tec
hniques
f
or
glucos
e
mo
nit
or
ing.
I
n
thi
s
pa
pe
r
we
de
ve
lop
a
ne
w
s
ys
tem
mea
s
ur
ing
a
g
lucos
e
leve
l
in
non
-
invas
ive
wa
y.
T
he
PPG
tec
hnique
is
us
e
d
f
or
thi
s
pur
pos
e
.
T
his
tec
hnol
ogy
is
s
im
ple,
low
-
c
os
t,
a
nd
ha
s
a
lr
e
a
dy
be
e
n
e
f
f
e
c
ti
ve
ly
a
ppli
e
d
to
the
de
tec
ti
on
of
oxyge
n
s
a
tur
a
ti
on
a
nd
puls
a
ti
on
r
a
te
on
va
r
ious
we
a
r
a
ble
de
vice
s
[
12]
.
A
PPG
s
ys
tem
uti
li
z
e
s
li
ght
to
e
va
luate
c
ha
nge
s
in
blood
c
ir
c
ulation
volum
e
by
mea
s
ur
ing
the
a
bs
or
pti
on
o
f
li
ght
in
the
ti
s
s
ue
.
I
t
is
obs
e
r
ve
d
that
s
ome
wa
ve
lengths
of
li
ght’
s
a
bs
or
pti
on
a
nd
r
e
f
lec
tanc
e
a
r
e
s
e
ns
it
ive
to
the
body’
s
he
modynamic
c
ha
r
a
c
ter
is
ti
c
s
,
whic
h
a
r
e
c
los
e
ly
ti
e
d
to
the
c
ondit
ion
of
the
c
a
r
diovas
c
ular
s
ys
tem,
while
it
is
we
ll
known
that
blood
glucos
e
l
e
ve
ls
,
whic
h
may
be
im
media
tely
mea
s
ur
e
d
a
s
puls
e
mor
phologi
c
a
l
p
r
of
il
e
s
,
ha
ve
a
long
-
ter
m
im
pa
c
t
on
the
c
a
r
d
iovas
c
ular
s
ys
tem
[
16]
.
F
or
that
r
e
a
s
on,
e
s
tablis
hing
a
li
nk
be
twe
e
n
P
P
G
puls
e
s
ha
pe
a
nd
blood
glucos
e
leve
ls
may
of
f
e
r
a
us
e
f
ul
mea
ns
of
noninvas
ively
pr
e
dicting
blood
glucos
e
leve
ls
.
T
he
main
goa
l
of
ou
r
wor
ks
is
to
r
e
plac
e
the
c
on
ve
nti
ona
l
invas
ive
-
glucos
e
mea
s
ur
e
ment
with
our
pr
opos
e
d
non
-
invas
iv
e
glucos
e
moni
tor
ing
de
vice
.
Due
to
the
de
vice
’
s
inexpe
ns
ive
pr
ice
,
many
indi
viduals
c
a
n
uti
li
z
e
it
.
Additi
ona
ll
y,
it
is
qui
te
s
im
ple
to
u
s
e
a
nd
ought
to
e
a
s
il
y
f
it
the
major
i
ty
of
pa
ti
e
nts
’
f
inger
s
.
T
he
r
e
s
t
o
f
th
is
pa
pe
r
is
f
ir
s
t
c
onc
e
r
ne
d
with
a
br
ie
f
de
s
c
r
ipt
ion
o
f
the
glucos
e
mea
s
ur
e
ment
pr
inciple
.
T
his
is
f
oll
owe
d
by
the
pr
e
s
e
ntation
of
the
ha
r
dwa
r
e
s
ys
tem
whic
h
de
ve
loped
to
de
tec
t
the
P
P
G
s
ignals
on
dif
f
e
r
e
nt
s
ubjec
ts
.
T
he
n
the
de
ve
loped
s
of
twa
r
e
is
de
tailed
.
F
inally,
in
the
las
t
s
e
c
ti
on,
a
n
a
na
lys
is
of
the
P
P
G
a
nd
s
e
c
ond
de
r
ivative
P
P
G
(
S
DPP
G)
pa
r
a
mete
r
va
r
iat
ions
a
c
c
or
ding
to
glucos
e
mea
s
ur
e
ment
a
r
e
pr
e
s
e
nted
a
nd
dis
c
us
s
e
d
be
f
or
e
a
c
onc
lus
ion
is
given.
2.
M
AT
E
R
I
AL
AN
D
M
E
T
HO
DS
T
he
s
ugge
s
ted
de
s
ign
a
nd
the
method
of
e
xe
c
uti
on
a
r
e
de
s
c
r
ibed
in
de
tail
in
thi
s
s
e
c
ti
on.
T
he
e
nti
r
e
pr
oc
e
dur
e
is
s
e
pa
r
a
ted
int
o
thr
e
e
s
teps
:
ha
r
d
wa
r
e
de
ve
lopm
e
nt;
invas
ive
da
ta
c
oll
e
c
ti
on;
s
of
twa
r
e
de
ve
lopm
e
nt.
E
a
c
h
s
tep
is
c
r
uc
ial
f
o
r
the
s
uc
c
e
s
s
f
u
l
c
ompl
e
ti
on
o
f
the
p
r
ojec
t.
2.
1.
Hardwar
e
d
e
ve
lop
m
e
n
t
T
he
P
P
G
s
ignal
is
de
tec
ted
by
us
ing
the
e
a
s
y
puls
e
s
hield
a
s
s
hown
in
F
igur
e
1.
I
t
is
a
puls
e
s
e
ns
or
de
ve
loped
by
AR
DU
I
NO
whic
h
wa
s
de
s
igned
to
i
ll
us
tr
a
te
the
pr
inciple
of
pho
topl
e
thys
mogr
a
phy
(
P
P
G)
a
s
a
non
-
invas
ive
opti
c
a
l
tec
hnique
f
or
de
tec
ti
ng
c
a
r
d
io
-
va
s
c
ular
puls
e
wa
ve
f
r
om
a
f
inger
ti
p.
T
he
e
a
s
y
puls
e
ope
r
a
tes
us
ing
tr
a
ns
mi
s
s
ion
mode.
W
it
hin
the
s
e
ns
or
c
a
s
ing,
a
n
in
f
r
a
r
e
d
(
I
R
)
L
E
D
a
nd
a
photodete
c
tor
a
r
e
pos
it
ioned
on
oppos
it
e
s
ides
,
f
a
c
ing
e
a
c
h
other
.
T
he
c
hoice
of
a
n
I
R
L
E
D
is
ba
s
e
d
on
the
f
a
c
t
that
glucos
e
,
be
ing
a
type
of
monos
a
c
c
ha
r
ide
with
the
c
he
mi
c
a
l
f
or
mul
a
C
6H12O
6,
e
xhibi
ts
va
r
ious
ne
a
r
-
inf
r
a
r
e
d
(
NI
R
)
a
bs
or
pti
on
pe
a
ks
a
t
s
pe
c
if
ic
wa
ve
lengths
(
940,
9
70,
1408
,
1688
,
1925
,
2261
,
a
nd
2326
nm
)
.
How
e
ve
r
,
a
t
940nm,
the
int
e
r
f
e
r
e
nc
e
of
op
ti
c
a
l
s
ignals
by
other
blood
c
omponents
s
uc
h
a
s
wa
ter
,
plate
lets
,
a
nd
r
e
d
blood
c
e
ll
s
is
mi
nim
a
l.
T
his
a
ll
ows
f
o
r
opti
mal
de
pth
o
f
pe
ne
tr
a
ti
on
a
nd
a
c
c
ur
a
te
pr
e
diction
of
glucos
e
leve
ls
[
17]
.
2.
2.
I
n
vas
ive
d
at
a
c
oll
e
c
t
ion
T
o
de
ve
lop
the
c
a
li
br
a
ti
on
model
of
our
de
ve
lop
e
d
s
ys
tems
,
we
us
e
d
Dia
gno
-
C
he
c
k
S
e
ns
gluco
s
e
mete
r
D3D
L
661938
a
s
s
hown
in
F
igur
e
2
[
18]
.
T
his
wa
s
us
e
d,
us
ing
c
li
nica
ll
y
r
e
c
omm
e
nde
d
a
ppr
oa
c
h,
to
mea
s
ur
e
s
im
ult
a
ne
ous
ly
with
the
de
ve
loped
s
ys
tem
the
blood
glucos
e
leve
l
in
mi
li
mol
e
pe
r
li
ter
.
F
igur
e
3
il
lus
tr
a
tes
the
block
diagr
a
m
of
ou
r
p
r
opos
e
d
s
ys
tem
to
mea
s
ur
e
blood
glucos
e
leve
l.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2088
-
8708
I
nt
J
E
lec
&
C
omp
E
ng
,
Vol
.
1
4
,
No.
5
,
Oc
tober
2
02
4
:
5200
-
5208
5202
F
igur
e
1.
E
a
s
y
puls
e
s
hield
F
igur
e
2.
Dia
gno
-
C
he
c
k
S
e
ns
glucos
e
mete
r
F
igur
e
3.
T
he
block
diagr
a
m
of
our
pr
opos
e
d
s
ys
tem
to
mea
s
ur
e
blood
glucos
e
leve
l
2.
3.
S
o
f
t
war
e
d
e
ve
lop
m
e
n
t
2.
3.
1.
P
P
G
f
e
at
u
r
e
s
e
xt
r
ac
t
ion
I
n
or
de
r
to
de
tec
t
the
s
ys
tol
ic
a
mpl
it
ude
o
f
the
P
P
G
s
ignal,
the
pe
a
k
f
inder
f
unc
ti
on
in
M
AT
L
AB
[
19]
is
a
da
pted
to
identif
ied
pe
a
ks
a
nd
va
ll
e
y
f
or
e
a
c
h
puls
e
in
P
P
G
s
ignal
a
s
s
hown
in
F
ig
ur
e
4(
a
)
,
F
igur
e
4
(
b)
us
ing
a
us
e
r
-
de
f
ined
magnitude
th
r
e
s
hold,
thi
s
f
unc
ti
on
quickly
loca
tes
loca
l
pe
a
ks
o
r
va
ll
e
ys
(
loca
l
e
xtr
e
ma)
in
a
nois
y
ve
c
tor
.
I
t
then
de
ter
mi
n
e
s
if
e
a
c
h
pe
a
k
is
s
igni
f
ica
ntl
y
gr
e
a
ter
(
or
s
maller
)
than
the
da
ta
a
r
ound
it
.
I
n
ou
r
s
ys
tem,
the
thr
e
s
hold
va
lue
i
s
f
ixed
e
mpi
r
ica
ll
y.
W
he
r
e
a
s
pe
a
ks
s
ys
tol
ic
wa
s
de
noted
a
s
“
”
a
nd
va
ll
e
y
point
wa
s
de
noted
a
s
“
”
.
T
he
s
ys
tol
ic
a
mpl
it
ude
is
obtaine
d
us
ing
(
2)
,
=
–
(
2)
F
oll
owing
the
a
utom
a
ted
a
ppli
c
a
ti
on
o
f
our
M
AT
L
AB
-
de
v
e
loped
a
lgor
it
hm,
s
ys
tol
ic
a
mpl
it
ude
s
we
r
e
obtaine
d
f
or
10
puls
e
s
,
the
s
e
c
ond
de
r
ivativ
e
of
P
P
G
,
the
S
DPP
G
s
ignal
is
ge
ne
r
a
ted
us
ing
t
he
dif
f
(
)
f
unc
ti
on
in
the
M
AT
L
AB
S
of
twa
r
e
.
T
he
obtaine
d
s
ignal
S
DPP
G
c
ontain
5
wa
ve
s
c
a
ll
e
d
,
,
,
,
a
nd
.
I
n
our
wor
k
/
r
a
ti
o
is
de
r
ived
us
ing
(
3
)
.
One
s
hould
k
now
that
the
a
mpl
it
ude
o
f
a
nd
that
of
a
r
e
obtaine
d
us
ing
the
pe
a
k
f
unc
ti
on
f
inder
a
s
s
hown
in
F
igu
r
e
s
5(
a
)
,
(
b)
,
a
nd
(
c
)
.
/
=
(
)
/
(
)
(
3)
2.
3.
2.
Re
gr
e
s
s
ion
m
od
e
Af
ter
de
nois
ing
the
P
P
G
s
ignal
us
ing
f
ini
te
im
puls
ion
r
e
s
pons
e
f
il
ter
.
P
P
G
f
e
a
tur
e
s
we
r
e
de
tec
ted.
A
r
e
gr
e
s
s
ion
model
be
twe
e
n
the
da
ta
e
xtr
a
c
ted
f
r
om
the
P
P
G
a
nd
or
igi
na
l
va
lues
of
b
lood
gluc
os
e
wa
s
c
ons
tr
uc
ted.
T
his
pr
oc
e
s
s
is
il
lus
tr
a
ted
in
F
igu
r
e
6.
P
r
e
vious
s
tudi
e
s
ha
ve
noted
s
igni
f
ica
nt
dif
f
e
r
e
nc
e
s
in
the
c
ha
r
a
c
ter
is
ti
c
s
of
photopl
e
thys
mogr
a
phy
(
P
P
G)
wa
ve
f
or
ms
be
twe
e
n
he
a
lt
hy
indi
viduals
a
nd
thos
e
with
diabe
tes
.
Nir
a
la
e
t
al
.
[
20]
highl
ight
e
d
the
f
ir
s
t
a
nd
s
e
c
ond
e
igenva
lues
de
r
ived
f
r
om
the
f
ir
s
t
de
r
ivative
of
the
P
P
G
s
ignal
a
s
ke
y
f
e
a
tur
e
s
f
or
id
e
nti
f
ying
type
2
d
iabe
tes
melli
tus
(
T
2DM
)
.
F
ur
the
r
mor
e
,
r
e
s
e
a
r
c
h
[
21]
ha
s
s
hown
that
e
leva
ted
glucos
e
leve
l
s
in
the
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
E
lec
&
C
omp
E
ng
I
S
S
N:
2088
-
8708
B
lood
glucos
e
pr
e
diction
us
ing
non
-
invas
ive
opti
c
al
s
y
s
tem
bas
e
d
on
…
(
M
ohamm
e
d
A
ne
s
B
e
r
e
k
s
i
R
e
guig
)
5203
blood
a
lt
e
r
the
s
ha
pe
of
the
P
P
G
s
ignal,
whic
h
c
a
n
be
us
e
d
to
e
s
ti
mate
blood
glucos
e
leve
ls
in
dir
e
c
tl
y.
How
e
ve
r
,
the
s
ys
tol
ic
a
mpl
it
ude
a
nd
b/a
r
a
ti
o
a
r
e
obtaine
d
f
r
o
m
the
pe
a
ks
o
f
P
P
G
a
nd
S
DPP
G
da
t
a
,
due
to
the
s
igni
f
ica
nc
e
of
it
s
puls
a
ti
le
c
omponents
in
r
e
f
lec
ti
ng
va
r
iations
in
the
blood
f
low
ins
ide
a
r
ter
i
e
s
,
thes
e
c
ha
r
a
c
ter
is
ti
c
s
a
r
e
pr
ovided
.
Dia
be
tes
im
pa
c
ts
blo
od
volum
e
tr
ic
c
ha
nge
s
via
a
lt
e
r
ing
blood
pr
opa
ga
t
ion.
T
he
a
c
c
umul
a
ti
on
of
a
ther
os
c
ler
os
is
a
nd
diabe
te
s
in
the
a
r
ter
ies
is
a
s
s
um
e
d
to
be
the
or
igi
n
of
thi
s
c
ondit
ion.
C
ha
nge
s
in
blood
c
ir
c
ulation
c
a
u
s
e
the
mor
phology
of
P
P
G
to
be
c
ome
r
ounde
r
a
s
a
r
e
s
ult
,
diabe
tes
a
lt
e
r
s
the
a
r
ter
y
wa
ll
’
s
e
las
ti
c
it
y,
whic
h
is
thought
to
c
ha
nge
the
P
P
G’
s
mo
r
phology
[
22]
.
T
h
r
ough
thi
s
wo
r
k
t
wo
P
P
G
indi
c
e
s
a
r
e
e
xa
nim
a
ted,
s
ys
tol
ic
a
mpl
it
ude
S
a
a
nd
b/a
r
a
ti
o
a
r
e
obtaine
d
f
r
o
m
ten
P
P
G
a
nd
S
DPP
G
r
e
c
or
ding
f
or
the
s
a
me
s
ubjec
t.
I
n
the
f
ir
s
t
s
tep
thes
e
de
r
ive
d
va
lues
of
S
a
a
nd
the
b
/a
r
a
ti
on
a
r
e
s
e
t
in
two
ve
c
tor
s
(
4
)
a
nd
(
5)
whe
r
e
a
s
the
invas
ive
B
G
L
s
im
ult
a
ne
ou
s
l
y
mea
s
ur
e
d
va
lues
a
r
e
s
e
t
in
a
nother
ve
c
tor
(
6)
.
I
n
s
e
c
ond
s
tep
the
mos
t
a
c
c
ur
a
te
li
ne
a
r
r
e
gr
e
s
s
ion
models
th
a
t
de
pict
the
c
ha
nge
in
s
ys
tol
ic
a
mpl
it
ude
a
r
e
then
s
e
lec
ted.
E
qua
ti
on
(7
)
a
nd
b
/a
r
a
ti
o
(
8)
with
B
GL
a
r
e
de
ter
m
ined.
=
[
(
1
)
.
.
.
.
.
.
(
10
)
]
(
4)
′
/
=
[
′
(
1
)
.
.
.
.
.
.
′
(
10
)
]
(
5)
=
[
′
(
1
)
.
.
.
.
.
.
′
(
10
)
]
(
6)
W
it
h
is
s
ys
tol
ic
a
mpl
it
ude
ve
c
tor
,
’
/
is
/
r
a
ti
o
ve
c
tor
,
a
nd
is
blood
glucos
e
leve
l
ve
c
tor
.
T
he
e
qua
ti
on
s
(
7)
a
nd
(
8
)
r
e
pr
e
s
e
nt
the
li
ne
a
r
r
e
gr
e
s
s
ion
model
.
=
+
(
7)
′
=
′
′
+
′
(
8)
W
it
h
is
the
e
s
ti
mate
d
va
lue
o
f
blood
glucos
e
c
onc
e
ntr
a
ti
on
f
r
o
m
s
ys
tol
ic
a
mpl
it
ude
in
P
P
G
s
ignal
a
n
d
’
is
t
he
e
s
ti
mate
d
va
lue
blood
glucos
e
c
onc
e
ntr
a
ti
on
f
r
om
/
r
a
ti
o
f
e
a
tur
e
in
S
DPP
G
s
ignal.
T
he
,
’
,
,
a
nd
’
a
r
e
the
r
e
gr
e
s
s
ion
c
oe
f
f
icie
nts
.
(
a
)
(
b)
F
igur
e
4.
T
he
identif
ica
ti
on
pe
a
ks
a
nd
va
ll
e
y
in
P
P
G
s
ignal:
(
a
)
s
ys
tol
ic
a
mpl
it
ude
in
P
P
G
s
ignal
a
nd
(
b)
v
a
ll
e
y
point
in
P
P
G
s
ignal
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2088
-
8708
I
nt
J
E
lec
&
C
omp
E
ng
,
Vol
.
1
4
,
No.
5
,
Oc
tober
2
02
4
:
5200
-
5208
5204
(
a
)
(
b)
(
c
)
F
igur
e
5
.
T
he
de
tec
ti
on
of
dif
f
e
r
e
nt
wa
ve
s
in
s
e
c
on
d
de
r
ivative
of
P
P
G
s
ignal:
(
a
)
s
e
c
ond
de
r
ivative
o
f
P
P
G
s
ignal
(
S
DPP
G)
,
(
b
)
‘
b
’
wa
ve
de
tec
ti
on
in
S
DPP
G,
a
nd
(
c
)
‘
a
’
wa
ve
de
tec
ti
on
in
S
DPP
G
F
igur
e
6.
T
he
buil
ding
pr
oc
e
s
s
of
our
c
a
li
br
a
ti
on
m
ode
l
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
E
lec
&
C
omp
E
ng
I
S
S
N:
2088
-
8708
B
lood
glucos
e
pr
e
diction
us
ing
non
-
invas
ive
opti
c
al
s
y
s
tem
bas
e
d
on
…
(
M
ohamm
e
d
A
ne
s
B
e
r
e
k
s
i
R
e
guig
)
5205
2.
4.
T
h
e
m
e
as
u
r
e
m
e
n
t
p
r
ot
oc
ol
I
n
thi
s
th
is
s
tudy,
we
r
e
c
r
uit
e
d
10
male
s
he
a
lt
hy
s
ubjec
ts
,
T
he
s
e
pa
r
ti
c
ipants
,
who
r
a
nge
d
in
a
ge
f
r
om
20
to
38,
we
r
e
a
l
l
in
good
phys
ica
l
he
a
lt
h
ove
r
a
ll
a
nd
a
gr
e
e
d
in
wr
it
ing
to
pa
r
ti
c
ipate
in
th
e
s
tudy.
How
e
ve
r
,
the
blood
glucos
e
leve
l
will
be
c
he
c
ke
d
a
t
s
e
ve
r
a
l
mom
e
nts
thr
oughout
the
dur
a
ti
on
of
ma
ny
da
ys
,
both
be
f
or
e
a
nd
a
f
ter
mea
ls
,
whic
h
wil
l
r
e
s
ult
in
va
r
ying
qua
nti
ti
e
s
of
s
uga
r
leve
ls
in
the
blood
.
3.
RE
S
UL
T
S
AN
D
DI
S
CU
S
S
I
ON
I
n
thi
s
s
tudy
,
s
e
ve
r
a
l
mea
s
ur
e
s
we
r
e
made
to
bui
ld
our
r
e
gr
e
s
s
ion
model
in
or
de
r
to
pr
e
dict
the
B
GL
us
ing
a
noninvas
ive
P
P
G
s
e
ns
or
.
Us
ing
a
gr
oup
o
f
ten
male
volun
tee
r
s
a
c
r
os
s
a
r
a
nge
of
a
ge
s
,
the
pr
opos
e
d
s
ys
tem
wa
s
a
s
s
e
s
s
e
d.
F
oll
owing
s
ys
tem
c
a
li
br
a
ti
on,
the
pe
r
f
o
r
manc
e
of
the
non
-
invas
ive
blood
glucos
e
mea
s
ur
e
ment
s
ys
tem
w
a
s
e
va
luate
d
u
s
ing
the
va
l
idation
da
ta
s
e
ts
f
or
e
a
c
h
indi
vidual.
F
igur
e
s
7(
a
)
a
nd
(
b)
pr
e
s
e
nt
a
n
e
xa
mpl
e
of
r
e
gr
e
s
s
ion
a
na
lys
is
f
or
a
s
ubjec
t,
de
mons
tr
a
ti
ng
that
both
S
a
a
nd
the
b/a
r
a
ti
o
P
P
G
f
e
a
tur
e
s
incr
e
a
s
e
a
s
the
a
c
tual
b
lood
glucos
e
le
ve
l
(
B
GL
)
r
is
e
s
.
Af
ter
the
s
ys
tem
wa
s
f
ull
y
c
a
li
br
a
ted,
va
li
da
ti
on
da
ta
s
e
ts
f
or
e
a
c
h
s
ubjec
t
w
e
r
e
us
e
d
to
e
va
luate
the
pe
r
f
or
manc
e
of
the
mea
s
ur
e
ment
model.
T
he
pr
opos
e
d
non
-
invas
ive
glucos
e
s
e
ns
or
pr
otot
ype
’
s
pe
r
f
or
manc
e
is
e
va
luate
d
us
ing
c
or
r
e
lation
c
oe
f
f
icie
nts
(
R
²)
,
r
oot
mea
n
s
qua
r
e
s
e
r
r
or
(
R
M
S
E
)
a
nd
P
e
a
r
s
on’
s
c
or
r
e
lation
c
oe
f
f
icie
nt
.
2
=
1
−
∑
(
−
)
2
∑
(
−
′
)
2
(
9)
=
√
∑
(
−
)
²
(
10)
=
∑
(
−
′
)
(
−
′
)
√
∑
(
−
′
)
²
∑
(
−
′
)
²
(
11)
w
he
r
e
is
the
pr
e
dicte
d
blood
glucos
e
leve
l
,
is
the
r
e
a
l
GC
B
r
e
a
ding
,
’
is
the
mea
n
of
r
e
a
l
GC
B
r
e
a
ding
,
a
nd
’
is
the
mea
n
of
pr
e
dicte
d
blood
glucos
e
leve
l
.
(
a
)
(
b)
F
igur
e
7.
T
he
r
e
gr
e
s
s
ion
models
be
twe
e
n
B
GL
a
nd
P
P
G
s
ignal
f
e
a
tur
e
s
:
(
a
)
T
he
r
e
gr
e
s
s
ion
models
be
twe
e
n
B
GL
a
nd
s
ys
tol
ic
a
mpl
it
ude
in
P
P
G
s
ignal
,
(
b
)
T
h
e
r
e
gr
e
s
s
ion
models
be
twe
e
n
B
GL
a
nd
b/a
r
a
ti
o
in
S
DPP
G
T
a
ble
1
r
e
s
umes
the
obtaine
d
r
e
s
ult
s
of
B
GL
,
f
r
o
m
c
omm
e
r
c
ially
invas
ive
glucome
ter
a
nd
ou
r
non
-
invas
ive
P
P
G
s
ys
tem
us
ing
r
a
ti
o
f
e
a
tur
e
.
Ac
c
or
ding
to
the
r
e
s
e
a
r
c
h
of
[
11]
,
[
23]
r
e
lating
non
-
invas
ive
B
GL
a
na
lys
is
,
W
e
c
a
n
ve
r
if
y
the
a
c
c
ur
a
c
y
o
f
th
e
pr
e
dict
non
-
invas
ive
B
G
va
lue
obtaine
d
r
e
s
ult
s
.
I
n
f
a
c
t,
ba
s
e
d
on
thi
s
r
e
s
e
a
r
c
h
a
nd
the
objec
t
on
whic
h
t
he
mea
s
ur
e
ments
we
r
e
take
n,
we
dis
c
ove
r
that,
t
he
mea
n
va
lue
of
the
c
or
r
e
lation
c
oe
f
f
icie
nts
(
R
²)
is
:
0
.
80
37.
a
nd
P
e
a
r
s
on’
s
c
oe
f
f
icie
nt
is
:
0.
98.
I
t
is
e
vid
e
nt
that
ther
e
is
a
s
tr
ong
c
o
r
r
e
lation
be
twe
e
n
the
B
GL
mea
s
ur
e
ment
us
ing
invas
ive
de
vice
,
a
nd
the
B
GL
mea
s
ur
e
ment
us
ing
noninvas
ive
P
P
G
s
e
ns
or
a
c
c
or
ding
to
f
e
a
tu
r
e
.
T
hr
ough
thi
s
s
tudy
,
we
us
e
d
a
ls
o
the
b/a
r
a
t
io
in
the
P
P
G
s
ignal
to
pr
e
dict
B
GL
to
f
ur
ther
e
va
luate
the
li
nk
be
twe
e
n
the
c
omm
e
r
c
ially
a
va
il
a
ble
invas
ive
glucome
ter
a
nd
our
s
ugge
s
ted
non
-
invas
ive
P
P
G
s
e
ns
or
.
T
a
ble
2
r
e
s
umes
the
ob
taine
d
r
e
s
ult
s
o
f
B
G
L
,
f
r
o
m
c
omm
e
r
c
ially
invas
ive
glucome
ter
a
nd
ou
r
non
-
inv
a
s
ive
P
P
G
s
ys
tem
us
ing
b/a
r
a
ti
o
f
e
a
tur
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2088
-
8708
I
nt
J
E
lec
&
C
omp
E
ng
,
Vol
.
1
4
,
No.
5
,
Oc
tober
2
02
4
:
5200
-
5208
5206
F
r
om
T
a
ble
2
we
c
a
n
noti
c
e
that
ther
e
is
a
s
tr
ong
c
or
r
e
lation
be
twe
e
n
the
blood
g
lucos
e
leve
l
mea
s
ur
e
ment
us
ing
invas
ive
de
vice
,
a
nd
the
pr
e
dict
blood
glucos
e
leve
l
us
ing
noninvas
ive
P
P
G
s
e
ns
or
a
c
c
or
ding
to
the
b/a
r
a
ti
o
.
B
a
s
e
d
on
ou
r
r
e
s
e
a
r
c
h
a
nd
the
s
pe
c
if
ic
s
ubjec
t
of
our
mea
s
ur
e
ments
,
we
f
o
und
that
the
a
ve
r
a
ge
c
or
r
e
lation
c
oe
f
f
icie
nt
(
R
²)
is
0.
9314
,
a
nd
P
e
a
r
s
on’
s
c
oe
f
f
icie
nt
is
:
0.
99
.
T
o
f
u
r
ther
a
s
s
e
s
s
the
a
gr
e
e
ment
be
twe
e
n
blood
glucos
e
leve
ls
(
B
GL
)
mea
s
ur
e
d
us
ing
a
n
invas
ive
de
vice
a
nd
our
d
e
ve
loped
s
ys
tem,
we
c
onduc
ted
a
B
land
-
Altm
a
n
a
na
lys
is
a
s
s
hown
in
F
igur
e
8
(
a
)
a
nd
(
b
)
.
T
a
ble
1
.
S
umm
a
r
ize
s
the
B
GL
va
lues
of
both
a
c
o
mm
e
r
c
ially
a
va
il
a
ble
invas
ive
glucome
ter
a
nd
our
pr
opos
e
d
non
-
invas
ive
P
P
G
s
e
ns
or
us
ing
the
s
ys
tol
ic
a
mpl
it
ude
(
)
f
e
a
tur
e
S
ubj
e
c
t
1
2
3
4
5
B
G
L
us
in
g
in
va
s
iv
e
gl
uc
ome
te
r
(
mg/
dl
)
100
106
105
110
.
6
107
.
6
B
G
L
us
in
g
noni
nva
s
iv
e
P
P
G
s
e
ns
or
(
mg/
dl
)
119
.
16
111
.
12
109
.
36
108
.
98
105
.
97
C
or
r
e
la
ti
on (
R
²)
0
.
6356
0
.
89
0
.
7637
0
.
8552
0
.
8743
R
M
S
E
(
mg/
dl
)
20
.
24
5
.
83
4
.
61
2
.
4
7
.
03
P
e
a
r
s
on’
s
r
c
or
r
e
la
ti
on c
oe
f
f
ic
ie
nt
0
.
98
0
.
95
0
.
99
0
.
99
0
.
99
T
a
ble
2
.
R
e
s
ult
s
f
r
om
ou
r
s
ugge
s
ted
non
-
invas
ive
P
P
G
s
e
ns
or
us
ing
b/a
f
e
a
tu
r
e
a
nd
thos
e
f
r
om
a
c
ur
r
e
ntl
y
a
va
il
a
ble
invas
ive
glucome
ter
S
ubj
e
c
t
1
2
3
4
5
B
G
L
us
in
g
in
va
s
iv
e
gl
uc
ome
te
r
(
mg/
dl
)
100
106
105
110
.
6
107
.
6
B
G
L
us
in
g
noni
nva
s
iv
e
P
P
G
s
e
ns
or
(
mg/
dl
)
107
.
2
102
104
.
88
110
.
14
108
.
8
C
or
r
e
la
ti
on (
R
²)
0
.
9069
0
.
9692
0
.
9117
0
.
9775
0
.
8904
R
M
S
E
(
mg/
dl
)
6
.
08
3
.
94
1
.
28
1
.
59
4
.
52
P
e
a
r
s
on’
s
r
c
or
r
e
la
ti
on c
oe
f
f
ic
ie
nt
0
.
99
0
.
993
0
.
998
0
.
999
0
.
992
(
a
)
(
b)
F
igur
e
8.
B
land
-
Altm
a
n
plot
s
f
or
pr
e
dicte
d
va
lues
a
nd
r
e
f
e
r
e
nc
e
va
lues
(
a
)
us
ing
s
ys
tol
ic
a
mpl
it
ude
i
n
P
P
G
s
ignal
a
nd
(
b)
us
ing
b/a
in
S
DPP
G
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
E
lec
&
C
omp
E
ng
I
S
S
N:
2088
-
8708
B
lood
glucos
e
pr
e
diction
us
ing
non
-
invas
ive
opti
c
al
s
y
s
tem
bas
e
d
on
…
(
M
ohamm
e
d
A
ne
s
B
e
r
e
k
s
i
R
e
guig
)
5207
I
n
thi
s
f
igur
e
,
the
blue
l
ine
r
e
pr
e
s
e
nts
the
mea
n
dif
f
e
r
e
nc
e
(
)
be
twe
e
n
the
two
s
e
ts
of
mea
s
ur
e
ments
:
the
pr
e
dicte
d
glucos
e
va
lues
f
r
o
m
our
method
a
nd
the
glucos
e
va
lues
obtaine
d
us
ing
the
invas
ive
de
vice
.
T
he
r
e
d
l
ines
r
e
pr
e
s
e
nt
the
r
a
nge
of
±
1.
96
ti
mes
the
s
tanda
r
d
de
viation
(
)
.
T
he
B
land
-
Altm
a
n
plot
il
lus
tr
a
tes
that
the
mea
n
dif
f
e
r
e
nc
e
(
)
be
twe
e
n
the
two
mea
s
ur
e
ment
s
e
t
s
is
c
ons
i
s
tently
c
los
e
to
z
e
r
o
in
a
ll
c
a
s
e
s
,
whic
h
is
highl
y
f
a
vo
r
a
ble.
C
ons
e
que
ntl
y,
be
yond
the
c
or
r
e
lation
c
oe
f
f
icie
nt,
we
c
a
n
c
onf
idently
c
onc
lude
that
ou
r
pr
opos
e
d
method
yields
r
e
li
a
ble
pr
e
dictions
f
o
r
B
GL
,
a
s
s
uppor
te
d
by
the
B
land
-
Altm
a
n
a
na
lys
i
s
.
Ac
c
or
ding
to
the
r
e
s
ult
s
o
btaine
d
in
thi
s
wor
k
a
nd
c
ompar
ing
to
o
ther
s
wor
k
s
,
we
c
a
n
ve
r
if
y
that
the
s
ugge
s
ted
method
is
a
c
c
ur
a
te.
T
a
b
le
3
s
ynthes
ize
a
c
ompar
is
on
of
the
s
ugge
s
ted
s
y
s
tem
a
nd
s
ome
of
the
other
pr
opos
e
d
s
ys
tems
in
the
domain
s
c
ientif
ic
li
ter
a
tur
e
.
T
he
di
f
f
e
r
e
nc
e
va
lue
of
B
GL
be
twe
e
n
invas
ive
s
ys
tems
a
nd
the
non
-
invas
ive
s
ys
tems
is
lowe
r
than
mos
t
of
the
pr
opos
e
d
s
ys
tems
in
the
domain
s
c
ientif
ic
li
ter
a
tu
r
e
whe
r
e
a
s
the
c
or
r
e
lation
c
oe
f
f
icie
nt
is
gr
e
a
ter
.
T
his
pr
ove
s
a
n
a
melior
a
ti
on
in
the
p
r
e
c
is
ion
of
th
e
r
e
s
ult
s
due
to
our
de
ve
loped
indi
vidual
c
a
li
br
a
ti
on
model
f
or
e
a
c
h
s
ubjec
t
indepe
nde
ntl
y
of
the
other
s
.
T
his
a
ppr
oa
c
h
mi
ti
ga
tes
the
c
ha
ll
e
nge
s
a
r
is
ing
f
r
om
int
e
r
-
indi
vidual
va
r
iabili
ty
in
P
P
G
mea
s
ur
e
ments
.
F
a
c
to
r
s
s
uc
h
a
s
f
inger
c
ir
c
um
f
e
r
e
nc
e
,
s
kin
r
oughne
s
s
lea
ding
to
li
ght
s
c
a
tt
e
r
ing,
va
r
iations
in
body
f
lui
d
c
onc
e
ntr
a
ti
on,
int
e
r
f
e
r
e
nc
e
f
r
om
other
biol
ogica
l
c
ons
ti
tue
nts
,
s
kin
tone,
a
nd
mot
ion
a
r
ti
f
a
c
ts
c
a
n
s
igni
f
ica
ntl
y
inf
luenc
e
P
P
G
r
e
a
dings
.
B
y
tailor
ing
r
e
gr
e
s
s
ion
models
to
e
a
c
h
indi
vidual,
we
a
ddr
e
s
s
thes
e
c
ha
ll
e
nge
s
e
f
f
e
c
ti
v
e
ly,
e
ns
ur
ing
mor
e
a
c
c
ur
a
te
a
nd
r
e
li
a
ble
blood
glucos
e
pr
e
dictions
.
T
his
indi
vidualize
d
a
ppr
oa
c
h
e
nha
nc
e
s
the
pr
e
c
is
ion
of
non
-
inva
s
ive
glucos
e
mo
nit
or
ing,
pr
ovidi
ng
pe
r
s
ona
li
z
e
d
ins
ight
s
int
o
glucos
e
dyna
mi
c
s
while
mi
nim
izing
the
im
pa
c
t
of
c
onf
oundin
g
f
a
c
tor
s
inher
e
nt
in
P
P
G
mea
s
ur
e
ments
.
T
he
s
tudy
c
a
r
r
ied
out
in
thi
s
wor
k
wa
s
on
10
s
ubjec
ts
;
howe
ve
r
,
in
t
he
f
utur
e
s
c
ope
we
e
nvis
a
ge
to
a
dd
mor
e
s
ubjec
ts
to
the
e
xpe
r
im
e
nt
to
ge
ne
r
a
li
z
e
thes
e
obtaine
d
r
e
s
ult
s
.
T
a
ble
3
.
C
ompar
is
ons
be
twe
e
n
the
s
ys
tems
we
’
ve
pr
opos
e
d
a
nd
thos
e
that
ha
ve
be
e
n
put
out
in
the
r
e
leva
nt
s
c
ientif
ic
li
ter
a
tur
e
W
or
ks
i
n t
he
doma
in
T
e
c
hni
que
s
us
e
d
S
pe
c
tr
os
c
opy te
c
hni
que
S
ubj
e
c
ts
W
a
ve
le
ngt
h (
nm)
P
P
G
f
e
a
tu
r
e
s
e
xt
r
a
c
ti
on
R²
R
a
c
hi
m a
nd C
hung [
23]
N
ot
me
nt
io
ne
d
12
530,660,850,950
01
0.85
O
guns
a
nya
a
nd D
a
r
a
mol
a
[
24]
T
r
a
ns
mi
s
s
io
n/
G
S
M
modul
e
40
1550
01
0.95
Q
a
w
qz
e
h
e
t
al
.
[
22]
.
T
r
a
ns
mi
s
s
io
n
587
950
03
0.70
Al
-
dha
he
r
i
e
t
al
.
[
25]
T
r
a
ns
mi
s
s
io
n
10
940
01
0.839
J
a
in
, J
os
hi
,
a
nd M
oha
nt
y [
16]
A
bs
or
pt
io
n a
nd
r
e
fl
e
c
ta
nc
e
97
940
01
0.900
O
ur
pr
opos
e
d me
th
od
T
r
a
ns
mi
s
s
io
n
10
940
02
0.9775
4.
CONC
L
USI
ON
I
n
thi
s
wor
k,
a
n
in
f
r
a
r
e
d
opti
c
a
l
s
e
ns
or
is
de
ve
loped
to
non
-
invas
ive
mea
s
ur
e
ment
of
blood
glucos
e
leve
l.
T
his
wa
s
a
c
c
ompl
is
he
d
us
ing
s
of
twa
r
e
pr
ogr
a
ms
whic
h
we
r
e
de
ve
loped
to
f
ir
s
t
pr
oc
e
s
s
the
photopl
e
thys
mogr
a
m
s
ignal
a
nd
then
e
xt
r
a
c
t
two
i
mpor
tant
f
e
a
tur
e
s
f
r
om
P
P
G
s
ignal
mor
phology
.
T
o
model
the
li
nk
be
twe
e
n
P
P
G
s
ignals
a
nd
B
GL
,
a
li
ne
a
r
r
e
gr
e
s
s
ion
model
wa
s
de
ve
loped.
T
he
obtaine
d
r
e
s
ult
on
the
dif
f
e
r
e
nt
tes
ts
a
ppli
e
d
on
di
f
f
e
r
e
nt
s
ubjec
ts
s
how
that
ther
e
is
a
s
tr
ong
c
o
r
r
e
lation
be
twe
e
n
the
c
o
mm
e
r
c
ial
glucome
ter
mea
s
ur
e
ments
a
nd
our
pr
opos
e
d
s
ys
tem
mea
s
ur
e
ments
.
S
im
il
a
r
ly;
a
c
c
or
ding
to
obtaine
d
r
e
s
ult
s
a
nd
c
ompar
e
d
with
other
s
tudi
e
s
,
we
c
a
n
c
onf
i
r
m
the
f
e
a
s
ibi
li
ty
of
us
ing
our
de
ve
loped
inf
r
a
r
e
d
opti
c
a
l
s
e
ns
or
li
ke
a
non
-
invas
iv
e
blood
glucos
e
mea
s
ur
e
ment
tec
hnique.
I
n
the
f
utur
e
,
we
int
e
nd
to
e
x
pa
nd
the
number
of
pa
r
ti
c
ipants
in
our
s
tudi
e
s
.
And
a
ls
o,
we
e
nvis
a
ge
to
tes
t
mor
e
P
P
G
f
e
a
tu
r
e
s
with
dif
f
e
r
e
nt
r
e
gr
e
s
s
ion
models
to
incr
e
a
s
e
the
pe
r
f
o
r
manc
e
of
o
ur
non
-
invas
ive
glucome
ter
.
RE
F
E
RE
NC
E
S
[
1]
L
.
R
il
e
y,
“
M
e
a
n
f
a
s
ti
ng
bl
ood
gl
uc
o
s
e
,”
W
or
ld
H
e
al
th
O
r
gani
z
at
io
n
.
ht
tp
s
:/
/ww
w
.w
ho.i
nt
/d
a
ta
/g
ho/
in
di
c
a
to
r
-
me
ta
da
ta
-
r
e
gi
s
tr
y/
im
r
-
de
ta
il
s
/2
380 (
a
c
c
e
s
s
e
d
J
un. 16, 2024)
.
[
2]
P
.
J
a
in
,
A
.
M
.
J
os
hi
,
a
nd
S
.
P
.
M
oha
nt
y,
“
i
G
L
U
1.1:
to
w
a
r
ds
a
gl
uc
os
e
-
in
s
ul
in
mode
l
ba
s
e
d
c
lo
s
e
d
lo
op
I
oM
T
f
r
a
me
w
or
k
f
or
a
ut
oma
ti
c
in
s
ul
in
c
ont
r
ol
of
di
a
be
ti
c
pa
ti
e
nt
s
,”
in
2020
I
E
E
E
6t
h
W
or
ld
F
or
um
on
I
nt
e
r
ne
t
of
T
hi
ngs
(
W
F
-
I
oT
)
,
J
un.
2020,
pp. 1
–
6, doi:
10.1109/WF
-
I
oT
48130.2020.9221132.
[
3]
R
.
P
e
r
iy
a
s
a
my
a
nd
S
.
A
na
nd,
“
A
s
tu
dy
on
non
-
in
va
s
iv
e
bl
ood
gl
uc
os
e
e
s
ti
ma
ti
on
—
a
n
a
ppr
oa
c
h
us
in
g
c
a
pa
c
it
a
nc
e
me
a
s
ur
e
me
nt
te
c
hni
que
,”
in
2016
I
nt
e
r
nat
io
nal
C
onf
e
r
e
n
c
e
on
Si
gnal
P
r
oc
e
s
s
in
g,
C
o
m
m
uni
c
at
io
n,
P
ow
e
r
and
E
m
be
dde
d
Sy
s
te
m
(
SC
O
P
E
S)
,
O
c
t.
2016, pp. 847
–
850, doi:
10.1109/S
C
O
P
E
S
.2016.7955561.
[
4]
F
. A
gui
r
r
e
e
t
al
.
,
I
D
F
di
abe
te
s
at
la
s
, S
ix
th
e
di
t.
I
nt
e
r
na
ti
ona
l
D
i
a
be
te
s
F
e
de
r
a
ti
on
, 2013.
[
5]
S
.
R
a
ma
s
a
ha
y
a
m,
L
.
A
r
or
a
,
S
.
R
.
C
how
dhur
y,
a
nd
M
.
A
numukonda,
“
F
P
G
A
ba
s
e
d
s
ys
te
m
f
or
bl
ood
gl
uc
os
e
s
e
n
s
in
g
u
s
in
g
phot
opl
e
th
ys
mogr
a
phy
a
nd
onl
in
e
mot
io
n
a
r
ti
f
a
c
t
c
or
r
e
c
ti
on
us
in
g
a
da
li
ne
,”
in
2015
9t
h
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
Se
n
s
in
g
T
e
c
hnol
ogy
(
I
C
ST
)
, D
e
c
. 2015, pp. 22
–
27, doi:
10.1109/I
C
S
e
ns
T
.2015.7438358.
[
6]
D
.
H
e
,
Z
.
W
u,
B
.
C
ui
,
Z
.
J
in
,
a
nd
E
.
X
u,
“
A
f
lu
or
ome
tr
ic
me
t
hod
f
or
a
pt
a
me
r
-
ba
s
e
d
s
im
ul
ta
ne
ous
de
te
r
mi
na
ti
on
of
two
ki
nds
of
th
e
f
us
a
r
iu
m
myc
ot
oxi
ns
z
e
a
r
a
le
none
a
nd
f
umoni
s
in
B
1
ma
ki
ng
us
e
of
gol
d
na
nor
ods
a
nd
upc
onv
e
r
s
io
n
na
nop
a
r
ti
c
le
s
,
”
M
ic
r
oc
hi
m
ic
a A
c
ta
, vol
. 187, no. 4, Apr
. 2020, doi:
10.1007/s
00604
-
020
-
04236
-
4.
[
7]
N
.
M
.
Z
hi
lo
,
P
.
A
.
R
ude
nko,
a
nd
A
.
N
.
Z
hi
ga
yl
o,
“
D
e
ve
lo
pme
nt
of
ha
r
dw
a
r
e
-
s
of
twa
r
e
te
s
t
be
nc
h
f
or
opt
ic
a
l
non
-
in
va
s
iv
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2088
-
8708
I
nt
J
E
lec
&
C
omp
E
ng
,
Vol
.
1
4
,
No.
5
,
Oc
tober
2
02
4
:
5200
-
5208
5208
gl
uc
ome
te
r
im
pr
ove
me
nt
,”
in
2017
I
E
E
E
C
onf
e
r
e
nc
e
of
R
us
s
ia
n
Y
oung
R
e
s
e
ar
c
he
r
s
in
E
le
c
tr
ic
al
and
E
le
c
tr
oni
c
E
ngi
ne
e
r
in
g
(
E
I
C
onR
us
)
, 2017, pp. 89
–
90, doi:
10.1109/
E
I
C
onR
us
.2017.7910500.
[
8]
S
.
I
.
G
us
e
v,
A
.
A
.
S
im
onova
,
P
.
S
.
D
e
mc
h
e
nko,
M
.
K
.
K
hodz
i
ts
ky,
a
nd
O
.
P
.
C
he
r
ka
s
ov
a
,
“
B
lo
od
gl
uc
os
e
c
onc
e
nt
r
a
ti
on
s
e
ns
in
g
us
in
g
bi
ol
ogi
c
a
l
mol
e
c
ul
e
s
r
e
la
x
a
ti
on
ti
me
s
de
te
r
mi
na
ti
on,”
i
n
2017
I
E
E
E
I
nt
e
r
nat
io
nal
Sy
m
pos
iu
m
on
M
e
di
c
al
M
e
as
u
r
e
m
e
nt
s
and A
ppl
ic
at
io
ns
(
M
e
M
e
A
)
, M
a
y 2017, pp. 458
–
463, doi:
10.11
09/
M
e
M
e
A
.2017.7985920.
[
9]
M
.
W
.
S
a
r
i
a
nd M
.
L
ut
hf
i,
“
D
e
s
ig
n a
nd
a
na
ly
s
is
of
non
-
in
va
s
iv
e
bl
ood
gl
uc
os
e
le
ve
ls
moni
to
r
in
g,”
in
2016
I
nt
e
r
nat
io
nal
Se
m
i
nar
on
A
ppl
ic
at
io
n
fo
r
T
e
c
hnol
ogy
of
I
nf
or
m
at
io
n
and
C
om
m
uni
c
at
io
n
(
I
Se
m
ant
ic
)
,
A
ug.
2016,
pp.
134
–
137,
doi
:
10.1109/I
S
E
M
A
N
T
I
C
.2016.7873825.
[
10]
S
.
L
e
kha
a
nd
M
.
S
uc
he
th
a
,
“
N
on
-
in
va
s
iv
e
di
a
b
e
te
s
d
e
te
c
ti
on
a
nd
c
la
s
s
if
ic
a
ti
on
us
in
g
br
e
a
th
a
n
a
ly
s
is
,
”
in
2015
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on C
om
m
uni
c
at
io
n and Signal
P
r
oc
e
s
s
in
g, I
C
C
SP
2015
, A
pr
. 2015, pp. 955
–
958, doi:
10.1109/I
C
C
S
P
.2015.7322
639.
[
11]
K
.
F
e
ll
a
h
A
r
bi
,
S
.
S
oul
i
ma
ne
,
a
nd
F
.
S
a
f
f
ih
,
“
N
on
-
in
va
s
iv
e
me
th
od
f
or
bl
ood
gl
uc
os
e
moni
to
r
in
g
us
in
g
E
C
G
s
ig
na
l,
”
P
o
li
s
h
J
our
nal
of
M
e
di
c
al
P
hy
s
ic
s
and E
ngi
ne
e
r
in
g
, vol
. 29, no. 1, pp. 1
–
9, F
e
b. 2023, doi:
10.2478/pj
mpe
-
2023
-
0001.
[
12]
J
.
C
hu,
W
.
T
.
Y
a
ng,
W
.
R
.
L
u,
Y
.
T
.
C
h
a
ng,
T
.
H
.
H
s
ie
h,
a
nd
F
.
L
.
Y
a
ng,
“
90%
a
c
c
ur
a
c
y
f
or
phot
opl
e
th
ys
mogr
a
phy‐
ba
s
e
d
n
on‐
in
va
s
iv
e
bl
ood
gl
uc
o
s
e
pr
e
di
c
ti
on
by
d
e
e
p
le
a
r
ni
ng
w
it
h
c
oho
r
t
a
r
r
a
nge
me
nt
a
nd
qua
r
te
r
ly
me
a
s
ur
e
d
H
ba
1C
,”
Se
ns
o
r
s
,
vol
.
21,
no. 23, Nov. 2021, d
oi
:
10.3390/s
21237815.
[
13]
S
.
H
a
bbu,
M
.
D
a
le
,
a
nd
R
.
G
honga
de
,
“
E
s
ti
ma
ti
on
of
bl
ood
gl
uc
os
e
by
non
-
in
va
s
iv
e
m
e
th
od
us
in
g
phot
opl
e
th
ys
mogr
a
p
hy,”
Sadhana
-
A
c
ade
m
y
P
r
oc
e
e
di
ngs
i
n E
ngi
ne
e
r
in
g S
c
ie
nc
e
s
, vol
. 44, no. 6, M
a
y 2019, doi:
10.1007/s
12046
-
019
-
1118
-
9.
[
14]
S
.
S
a
tt
e
r
,
M
.
S
.
T
ur
ja
,
T
.
-
H
.
K
w
on,
a
nd
K
.
-
D
.
K
im
,
“
E
M
D
-
ba
s
e
d
noni
nva
s
iv
e
bl
ood
gl
uc
os
e
e
s
ti
ma
ti
on
f
r
om
P
P
G
s
ig
na
ls
us
in
g
ma
c
hi
ne
l
e
a
r
ni
ng a
lg
or
it
hms
,”
A
ppl
ie
d Sc
ie
nc
e
s
, vol
. 14, no. 4,
F
e
b. 2024, doi:
10.3390/app140414
06.
[
15]
M
.
S
.
T
ur
ja
,
T
.
-
H
.
K
w
on,
H
.
K
im
,
a
nd
K
.
-
D
.
K
im
,
“
N
oni
nva
s
iv
e
in
vi
vo
e
s
ti
ma
ti
on
of
H
bA
1c
ba
s
e
d
on
th
e
be
e
r
–
la
mbe
r
t
mode
l
f
r
om phot
opl
e
th
ys
mogr
a
m us
in
g only t
w
o w
a
ve
le
ngt
hs
,”
A
ppl
ie
d Sc
ie
nc
e
s
, vol
. 13, no. 6, M
a
r
. 2023, doi:
10.3390/app130636
26.
[
16]
P
.
J
a
in
,
A
.
M
.
J
os
hi
,
a
nd
S
.
P
.
M
oha
nt
y,
“
iG
L
U
1.0:
a
n
a
c
c
ur
a
te
non
-
in
va
s
iv
e
ne
a
r
-
in
f
r
a
r
e
d
dua
l
s
hor
t
w
a
ve
le
ngt
hs
s
pe
c
tr
os
c
opy
ba
s
e
d gl
uc
ome
te
r
f
or
s
ma
r
t
he
a
lt
hc
a
r
e
,”
a
r
X
iv
:
1911.04471
,
N
ov. 2019, doi:
10.1109/M
C
E
.2019.2940855.
[
17]
M
.
S
ho
kr
e
k
ho
da
e
i
,
D
. P
.
C
is
to
la
,
R
.
C
.
R
o
b
e
r
t
s
,
a
n
d
S
.
Q
ui
no
ne
s
,
“
N
o
n
-
i
nv
a
s
i
v
e
gl
uc
o
s
e
m
on
it
or
i
ng
u
s
in
g
opt
i
c
a
l
s
e
n
s
or
a
nd
m
a
c
h
in
e
l
e
a
r
ni
n
g
t
e
c
hn
iq
ue
s
f
or
di
a
b
e
te
s
a
pp
li
c
a
t
io
ns
,
”
I
E
E
E
A
c
c
e
s
s
,
vo
l.
9
,
p
p.
7
30
29
–
7
30
45
,
20
21
,
doi
:
10
.1
10
9/
A
C
C
E
S
S
.2
02
1.
30
79
18
2.
[
18]
“
B
lo
od gluc
os
e
me
t
e
r
,”
(
in
F
r
e
nc
h)
,
D
ia
gno
-
C
he
c
k
®
Sm
ar
t
. ht
tp
s
:/
/d
ia
gno
-
c
he
c
k.c
om/
f
r
/
(
a
c
c
e
s
s
e
d M
a
r
. 18, 2023)
.
[
19]
“
M
A
T
L
A
B
,
‘
F
in
d
lo
c
a
l
ma
xi
ma
-
M
A
T
L
A
B
f
in
dpe
a
ks
,’
”
M
a
th
W
or
ks
,
2020
.
A
c
c
e
s
s
e
d
:
M
a
r
18,
2023.
[
O
nl
in
e
]
.
A
va
il
a
bl
e
:
ht
tp
s
:/
/ww
w
.ma
th
w
or
ks
.c
om/
he
lp
/s
ig
na
l/
r
e
f
/f
in
dpe
a
ks
.ht
ml
[
20]
N
.
N
ir
a
la
,
R
.
P
e
r
iy
a
s
a
my,
B
.
K
.
S
in
gh,
a
nd
A
.
K
uma
r
,
“
D
e
te
c
ti
on
of
ty
pe
-
2
di
a
be
te
s
us
in
g
c
h
a
r
a
c
te
r
is
ti
c
s
of
to
e
phot
opl
e
th
ys
mogr
a
m
by
a
ppl
yi
ng
s
uppor
t
ve
c
to
r
ma
c
hi
ne
,”
B
io
c
y
be
r
ne
ti
c
s
and
B
io
m
e
di
c
al
E
ngi
ne
e
r
in
g
,
vol
.
39,
no.
1,
pp. 38
–
51, J
a
n. 2019, doi:
10.1016/j
.bbe
.2018.09.007.
[
21]
G
.
H
a
mm
our
a
nd
D
.
P
.
M
a
ndi
c
,
“
A
n
in
-
e
a
r
P
P
G
-
ba
s
e
d
bl
ood
gl
uc
os
e
moni
to
r
:
a
pr
oof
-
of
-
c
onc
e
pt
s
tu
dy,”
Se
ns
or
s
,
vol
.
23,
no
.
6,
M
a
r
. 2023, doi:
10.3390/s
23063319.
[
22]
Y
.
K
. Q
a
w
qz
e
h, A
.
S
. B
a
j
a
h
z
a
r
, M
.
J
e
mm
a
l
i,
M
.
M
. O
t
oo
m,
a
nd A
. T
ha
lj
a
o
ui
, “
C
l
a
s
s
if
ic
a
t
io
n of
di
a
be
te
s
u
s
in
g ph
ot
o
pl
e
t
hy
s
mo
gr
a
m
(
P
P
G
)
w
a
v
e
f
or
m
a
na
ly
s
i
s
:
lo
gi
s
ti
c
r
e
gr
e
s
s
io
n
mo
d
e
li
ng
,
”
B
io
M
e
d
R
e
s
e
a
r
c
h
I
nt
e
r
n
at
io
na
l
,
v
ol
.
20
20
,
pp
.
1
–
6,
A
ug
.
202
0
,
d
oi
:
10
.1
15
5/
20
20
/3
76
46
53
.
[
23]
V
.
P
.
R
a
c
hi
m
a
nd
W
.
Y
.
C
hung,
“
W
e
a
r
a
bl
e
-
ba
nd
ty
pe
vi
s
ib
l
e
-
ne
a
r
in
f
r
a
r
e
d
opt
ic
a
l
bi
os
e
ns
or
f
or
non
-
in
va
s
iv
e
bl
ood
gl
uc
os
e
moni
to
r
in
g,”
Se
ns
or
s
and A
c
tu
at
or
s
, B
:
C
he
m
i
c
al
, vol
. 286, pp. 173
–
180, M
a
y 2019, doi:
10.1016/j
.s
nb.2019.01.121.
[
24]
A
.
O
.
O
guns
a
nya
a
nd
D
.
O
.
D
a
r
a
mol
a
,
“
D
e
s
ig
n
a
nd
de
ve
lo
pm
e
nt
of
a
non
-
in
va
s
iv
e
gl
uc
ome
te
r
s
ys
te
m,”
W
e
s
t
I
ndi
an
J
our
nal
of
E
ngi
ne
e
r
in
g
, vol
. 44, no. 2, pp. 70
–
79, J
a
n. 2022, doi:
10.4741
2/
ic
pw
8648.
[
25]
M
. A
. A
l
-
D
ha
he
r
i,
N
. E
. M
e
kka
ki
a
-
M
a
a
z
a
, H
. M
ouha
dj
e
r
, a
nd
A
. L
a
khda
r
i,
“
N
oni
nva
s
iv
e
bl
ood gluc
os
e
moni
to
r
in
g s
ys
te
m b
a
s
e
d
on
ne
a
r
-
in
f
r
a
r
e
d
me
th
od,”
I
nt
e
r
nat
io
nal
J
our
nal
of
E
le
c
tr
ic
al
and
C
om
put
e
r
E
ngi
ne
e
r
in
g
,
vol
.
10,
no.
2,
pp.
1736
–
1746,
A
pr
.
2020, doi:
10.11591/i
je
c
e
.v10i2.pp1736
-
1746.
B
I
OG
RA
P
HI
E
S
OF
AU
T
HO
RS
M
o
ha
m
m
ed
A
nes
Bereks
i
R
eg
u
i
g
recei
v
ed
t
h
e
d
o
ct
o
r’s
d
eg
ree
i
n
b
i
o
me
d
i
cal
i
n
s
t
ru
m
en
t
at
i
o
n
i
n
2
0
1
8
fro
m
t
h
e
A
b
o
u
Be
k
r
Bel
k
ai
d
U
n
i
v
er
s
i
t
y
,
T
l
emcen
,
A
l
g
eri
a
.
H
e
i
s
cu
rren
t
l
y
l
ect
u
rer
i
n
Bi
o
me
d
i
cal
E
n
g
i
n
eer
i
n
g
D
e
p
art
me
n
t
,
T
l
emcen
U
n
i
v
ers
i
t
y
an
d
memb
er
o
f
Mi
cro
s
y
s
t
em
s
an
d
Co
n
n
ect
e
d
O
b
j
ect
s
Fi
el
d
G
ro
u
p
o
f
t
h
e
“Bi
o
med
i
ca
l
E
n
g
i
n
eer
i
n
g
Res
earch
L
ab
o
rat
o
ry
”.
H
i
s
recen
t
w
o
rk
f
o
cu
s
es
o
n
t
rea
t
men
t
an
d
an
al
y
s
i
s
o
f
p
h
y
s
i
o
l
o
g
i
cal
s
i
g
n
a
l
s
.
H
e
can
b
e
co
n
t
ac
t
ed
at
ema
i
l
:
mo
h
ammed
_
an
e
s
.
b
erek
s
i
_
reg
u
i
g
@
u
mmt
o
.
d
z
.
N
a
s
s
i
m
a
La
bde
l
l
i
recei
v
ed
t
h
e
d
o
c
t
o
r’
s
d
eg
ree
i
n
el
ect
r
o
n
i
cs
i
n
s
t
r
u
men
t
a
t
i
o
n
in
2
0
2
0
fro
m
t
h
e
A
b
o
u
Bek
r
Bel
k
ai
d
U
n
i
v
er
s
i
t
y
,
T
l
emce
n
,
A
l
g
er
i
a
.
s
h
e
i
s
cu
rren
t
l
y
l
ect
u
rer
i
n
Sch
o
o
l
o
f
A
p
p
l
i
e
d
Sc
i
en
ce
s
T
l
emcen
an
d
memb
er
o
f
Mi
cro
s
y
s
t
em
s
an
d
Co
n
n
ec
t
ed
O
b
j
ec
t
s
Fi
el
d
G
ro
u
p
o
f
t
h
e
“
B
i
o
me
d
i
ca
l
E
n
g
i
n
eer
i
n
g
Res
ear
ch
L
ab
o
rat
o
ry
”.
h
er
re
s
earch
i
n
t
eres
t
i
n
c
l
u
d
es
m
o
d
e
l
l
i
n
g
an
d
co
m
p
u
t
at
i
o
n
o
f
b
i
o
med
i
cal
m
i
cro
s
y
s
t
em
s
.
Sh
e
can
b
e
co
n
t
act
e
d
at
emai
l
:
n
as
s
i
ma
l
ab
d
el
l
i
@
g
ma
i
l
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.