TELKOM
NIKA
, Vol.13, No
.3, Septembe
r 2015, pp. 7
94~805
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v13i3.1808
794
Re
cei
v
ed Ap
ril 3, 2015; Re
vised Ma
y 28
, 2015; Accep
t
ed Jun
e
14, 2015
Scanning-fluorescence Reader Based on Embedded
System
Zhonglong Z
h
ao
1
, Xiaoping Min*
1,2
, S
h
engxiang Ge
2
, Ningshao
Xia
2
1
Departme
n
t of Computer Sci
ence, Xiam
en Univers
i
t
y
,
Xi
ame
n
36
100
5, Chin
a
2
Nation
al Institute of Diag
nost
i
cs and Vacc
in
e Deve
lopm
ent
in Infectious D
i
seas
es, Xi
ame
n
Univ
ersit
y
,
Xi
ame
n
36
100
5, Chin
a
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: mxp@
xm
u.e
du.cn
A
b
st
r
a
ct
T
o
measur
e t
he co
nce
n
trati
on of
C-reacti
ve
prot
ein (
C
RP) in s
e
ru
m,
a p
o
rtabl
e, s
c
ann
ing-
fluoresc
ence
r
ead
er b
a
sed
on ti
me-res
o
lve
d
flu
o
roi
m
mu
no
assays
w
a
s deve
l
op
e
d
. T
he sca
nn
ing-
fluoresc
ence
r
ead
er i
n
tegr
ates w
i
th the
A
D
77
07 c
onv
erter, w
h
ich
perf
o
rms
at a
hi
g
h
acc
u
racy. T
he
photos
ens
itive
diod
e acts as the phot
o
e
le
ctric conversi
o
n
devic
e, an
optica
l
modu
le
based o
n
opt
ica
l
fibers, w
h
ich is
able to co
nce
n
trate the excit
a
tion l
i
g
h
t from an LED i
n
to a
line-s
h
a
pe b
e
a
m
, w
a
s desig
n
e
d
to
se
nd
an
d re
ce
i
v
e th
e op
ti
ca
l si
gn
a
l
. Th
e d
e
vi
ce
su
b
s
eq
uen
tl
y ad
d
r
e
sse
s
wa
ve
fo
rm d
a
t
a u
s
in
g a
g
r
ad
i
e
n
t
,
smo
o
thi
ng, a
n
d
bi
nari
z
at
io
n
meth
od. W
h
e
n
the dev
ice
measur
es the
C
R
P fluor
escen
c
e test strip, the
results ex
hi
bite
d a
go
od
lin
ea
rity (0.99
998)
and
the
CVs (
c
oefficie
n
t of v
a
riati
on) w
e
re
bel
ow
5%, w
h
i
c
h
indic
a
te hi
gh a
ccuracy. At the same ti
me
th
e
system is low
cost and s
m
al
l si
z
e
.
Ke
y
w
ords
:
fluoresc
ence test strip, tim
e
-resolved
fluor
oimm
unoass
a
y
,
em
bedded system
,
scanning-
fluoresc
ence r
ead
er
Copy
right
©
2015 Un
ive
r
sita
s Ah
mad
Dah
l
an
. All rig
h
t
s r
ese
rved
.
1. Introduc
tion
At pre
s
ent, a
promi
nent fe
a
t
ure of In
Vitro
Di
agno
si
s i
s
its tend
en
cy of pola
r
ization. On
e
scale of op
e
r
ation i
s
larg
e automate
d
instru
ment
s use
d
in me
dicin
e
; the other i
s
miniat
ure
POCT (p
oint
-of-care testi
ng) devi
c
e
s
. POCT
instruments a
r
e
conve
n
ient, fast and effici
ent,
gainin
g
incre
a
sin
g
re
cog
n
i
t
ion in the market [1-3].
Lateral flo
w
strip dete
c
tio
n
techn
o
logy
pl
ays impo
rtant roles in
healthcare,
dise
ase
diagn
osi
s
an
d di
sea
s
e
p
r
e
v
ention [4]. T
he first g
ene
ration late
ral
flow-stri
p
d
e
te
ction te
ch
nol
ogy
coul
d only a
c
hieve q
ualit
ative detecti
on; the
se
cond ge
neration reali
z
e
d
semi
-qu
antita
t
ive
detectio
n
usi
ng a dete
c
to
r; the third g
enerati
on
co
uld dete
c
t q
uantitatively,
prod
uci
ng m
o
re
accurate information and p
e
rform
ed reli
ably durin
g cli
n
ical di
agno
si
s.
Lateral
flow
strip ma
rkers
mainly in
clud
e colloid
gold
,
particulate
carbo
n
, latex p
a
rticle
s,
magneti
c
pa
rticle
s, and fluo
resce
n
t parti
cles. With
the
developm
ent of lateral flow-stri
p
dete
c
tio
n
techn
o
logy, tech
niqu
es to
analyze re
sults also
d
e
velope
d from qualitative de
tection ba
sed
on
image p
r
o
c
e
s
sing to qu
anti
t
ative detection ba
sed o
n
fluore
s
cen
c
e
monitori
ng an
d pro
c
e
ssi
ng
[5,
6]. Becau
s
e
of the more
pre
c
ise con
c
entration i
n
fo
rmation
and
more
accu
rat
e
and
relia
bl
e
evaluation
ba
sis for cli
n
ical
diag
no
sis,
q
uantitativ
e de
tecting i
n
st
ru
ments with
flu
o
re
scent l
a
teral-
flow strip
s
are becoming
an importa
nt POCT dete
c
tion platform in many cou
n
tries. A few of
quantitative immuno
assay
detection me
thods h
a
ve
been re
porte
d [7-9]. For exa
m
ple, Li and
et
al. have d
e
scribed
2-D o
p
tical
scan
syst
em to
q
uantif
y and im
age
up
conve
r
ting
pho
sph
o
r (UCP)
particl
es and haveutilized this system
to develop an IFN-
γ
imm
unoa
ssay wit
h
UCP lab
e
l
s
.
A
1.2W 9
80nm
fiber-co
uple
d
laser di
od
e wa
s u
s
ed
to excite the UCP pa
rticles th
rou
g
h
a
transpa
rent
microfluidi
c
chip cover, a
n
d
a ph
otomul
tiplier (PM
T
)
wa
s u
s
ed fo
r detectio
n
. T
he
scann
er take
s 8 minute
s
to finish 2-D strip sc
anni
ng
in both X and
Y direction
s
[7]. Yan and et
al. have deve
l
oped a bi
ose
n
so
r u
s
ed for quantitative
detectio
n
of Yersi
n
ia pe
st
is based on
LF
immuno
assa
y with a
com
p
licate
d
a
nd
bulky
optical
module, whi
c
h
al
so
i
n
cl
ud
es an
i
n
fra
r
e
d
(I
R)
lase
r and a p
hotomultiplie
r (PMT), wa
s
built with co
n
s
ide
r
ably hig
h
co
st and large si
ze [8].
In This
pap
er, a scanni
ng-fluore
s
cen
c
e
read
er
syste
m
wa
s
d
e
veloped. Combi
n
ing wit
h
time resolved
fluoroimm
un
oassay techn
o
logy for biol
ogical mea
s
u
r
eme
n
ts, the
s
e ARM-b
a
sed
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
9
30
Scanni
ng
-fluo
r
esce
nce Re
ader Ba
se
d o
n
Em
bedded
System
(Zho
nglon
g
Zhao
)
795
device
s
u
s
in
g AD77
07 a
nalog
-digital
conve
r
te
r
chi
p
s to conve
r
t the fluore
s
cen
c
e si
gn
al
captured by the high light
-sensitive tube into
a
digital signal. A custom
illumi
nation module
was
built to fo
cu
s
the ultraviolet
LED light i
n
to a
line
-
shap
e excitatio
n
b
e
am. Strip
scannin
g
wa
s d
o
n
e
by controllin
g a line
a
r
stage a
c
tuate
d
by a
ste
p
-
motor with
high resoluti
on. Wh
en t
h
e
fluore
s
cen
c
e
intensity wa
s colle
cted, an
algorithm
wa
s de
sign
ed to
pro
c
e
ss the
test data. Th
e
results of ex
perim
ents
sh
ow that scan
ning-f
lu
orescence rea
d
e
r
s have advant
a
ge
s of a high
detectio
n
sp
e
ed, high repe
atability,
and good lin
earity
,
at the same
ti
me the syst
em is lo
we
r cost
and sm
alle
r size.
2. Biological Method
s
2.1. Fluoresc
ence Immun
o
assay
CRP from h
u
man
pla
s
m
a
wa
s
used
as a
n
imm
u
n
o
gen fo
r the
pro
d
u
c
tion
of mou
s
e
mono
clon
al antibody (m
Ab).
Imm
uni
zation
s, cell
fusion, and
scree
n
ing
of
hybri
d
oma
cell
s
prod
uci
ng ant
i-CRP-mAb
were cond
ucte
d ac
cording t
o
a stand
ard
method [10
-
1
4
].
To lab
e
l the
anti-CRP-mAb with a
fluoresce
nt p
a
rti
c
l
e
, the EDC
method
wa
s
use
d
. A
spe
c
ime
n
wa
s ad
ded to th
e dete
c
tor
bu
ffer at 100 ti
mes
dilution,
and the
sp
eci
m
en
was loa
ded
into the sam
p
le well o
n
the ca
rtrid
ge. After 20
min incu
bation, th
e cart
ridg
e was in
serte
d
a
nd
scann
ed with
the scanni
ng-fluore
s
cen
c
e
read
er.
A spe
c
ime
n
wa
s ad
ded t
o
the dete
c
to
r buffer at 10
0 times
diluti
on, and th
e
specim
en
wa
s load
ed i
n
to the sam
p
le well on th
e ca
rtridg
e. After 20 min incu
bation, th
e ca
rtridg
e was
inse
rted an
d scann
ed with
the scanni
ng-fluore
s
cen
c
e
read
er.
2.2. Time-res
olv
e
d Fluoroimmunoassa
y
Several fluorescen
c
e imm
unoa
ssay technolo
g
ies a
r
e
available on
the market: enzym
e
immuno
assa
y, chemilumi
ne sce
n
ce i
mmuno
assay
,
and time-resolve
d
fluoroimmun
o
a
s
say.
Time-re
s
olve
d fluoroimm
u
noa
ssay is a
biologica
l d
e
tection te
ch
nology which
is use
d
by the
scanni
ng-flu
o
r
esce
nce re
a
der.
Gene
ral fluo
resce
n
t markers have
short fl
uorescence lifetimes, but a few of the
lanthani
de el
ements
(such as Eu, Tb,
Sm, and Dy
) displ
a
y a long-live
d
fluore
scen
ce. T
he
cha
r
a
c
teri
stics of
the
s
e l
a
n
t
hanide
elem
ents
are a
s
f
o
llows: n
o
int
e
rferen
ce
occurs b
e
twe
en t
h
e
exc
i
tation light and emitted light
bec
a
us
e
of a large Stok
es
s
h
ift; time-resolved
fluoroimm
uno
assay i
s
hig
h
l
y
sen
s
itivity beca
u
se the
spectrum
of th
e excitation
li
ght is
wid
e
, b
u
t
the spe
c
tru
m
of the emitted light is narro
w [15].
In time-resolved a
fluoroim
m
unoa
ssay t
e
ch
nolo
g
y, if we
delay th
e
measurement
of the
fluore
s
cent m
a
rkers after e
x
cita
tion, then the sen
s
itivity can be improved. Thi
s
delay can al
so
redu
ce
the i
n
terferen
ce fro
m
ba
ckgroun
d fluoresce
nce an
d e
nhan
ce
the
stability
and
spe
c
ificit
y.
Additionally, this te
chni
que
has a
wide
r l
i
near ra
nge,
and the
re
ag
ents h
a
ve a l
onge
r lifetime
.
In
addition, tim
e
-resolved fl
uoroi
mmun
o
a
s
says h
a
ve
both hi
gh
er
sen
s
itivities a
nd
ea
sier
automation th
an en
zyme immuno
assay
s
and
chemil
umine sce
n
ce immuno
assays [16, 17].
2.2.Time-res
olv
e
d fluoroimmunoassa
y
The lateral flow strip
prin
cip
l
e for the fluorometer i
s
sh
o
w
n in Figu
re
1.
Figure 1. Strip detectio
n
principl
e of fluoromete
r
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 3, September 20
15 : 794 – 805
796
To fix the
structure
of the
antigen
-antib
ody,
we
po
st
a layer of
nitrocell
u
lo
se m
e
mbran
e
on the
su
rface of the
strip
.
The
fro
n
t o
f
the stri
p is
the co
ntrol li
ne, whi
c
h
is
con
s
tru
c
ted
by
coatin
g goat anti-mo
use a
n
tibodie
s
and
monocl
onal
antibodi
es wi
th fluoresce
n
t
micro
s
ph
eres
on the surfa
c
e. Mono
clon
a
l
antibodie
s
can be
re
cogni
zed
by coatin
g goat anti
-
m
ouse antibo
d
i
e
s
automatically to form a
sta
b
le st
ru
cture.
Therefore, t
he fluoresce
nce inten
s
ity of the co
ntrol li
n
e
is su
bsta
ntiall
y fixed to perform qu
ality control fun
c
tio
n
s.
The test lin
e is con
s
tituted by co
ati
ng mono
clonal antibo
d
ies an
d n
on-coatin
g
mono
clon
al antibodi
es. Unli
ke
th
e control
lin
e, these a
n
tibodie
s
are
not
re
cog
n
ize
d
automatically, but
both
may co
mbin
e
with an
antig
en
to form
a
san
d
wi
ch
co
nst
r
u
c
tion
of
coati
n
g
mono
clon
al a
n
tibody + anti
gen + mo
no
cl
onal antibo
d
y. The more a
n
tigen
s, the more fluo
re
scent
microspheres will be stably fix
ed on the nitrocellul
o
se
memb
rane, whi
c
h
can measure t
he
antigen
de
nsi
t
y by determi
ning th
e fluo
rescen
ce
inte
nsity. The
st
ructure
of
the
test strip use
d
by
scanni
ng-flu
o
r
esce
nce re
a
der is
sho
w
n i
n
Figure 2.
Figure 2. Structure of the te
st strip
The intern
al stru
cture of the
strip i
s
sh
own ab
ove. The sa
mple
pad (p
art 1) i
s
used for
loadin
g
samp
les. Th
e conj
ugate p
ad (p
art 2) i
s
pl
aced in ex
ce
ssi
ve antibodi
es and
compl
e
xes.
The lami
nation layer
(part
4)
contains t
w
o lines: te
st line (line 3)
with fixed sol
ubility antibodies
and th
e
cont
rol line
(li
ne
5
)
with q
uantit
ative antibo
d
i
e
s [1
8]. Figu
re 3
shows a
numbe
r
of ti
me
-
resolved fluo
rescen
ce
stri
p
s
un
de
r ultra
v
iolet (UV
)
lig
ht, in whi
c
h t
he inten
s
ity o
f
the co
ntrol li
ne
is fixed whe
r
e
a
s the de
nsit
y of the test li
ne increa
se
s
grad
ually fro
m
wea
k
to strong.
Figure 3. Fluore
s
cent stri
p
under ult
r
avi
o
let (UV) lig
ht
3. Sy
stem Design
In the scan
n
i
ng-fluo
re
sce
n
ce
re
ade
r, t
he processo
r S3c244
0 i
s
respon
sibl
e for all
module
s
i
n
cl
uding th
e o
p
tical m
odul
e, motor
m
odule, di
gital
-
anal
og
co
n
v
ersio
n
mo
d
u
le,
perip
he
rals m
odule, po
we
r module
s
, and
softwa
r
e de
sign modul
e.
S3C24
40 is
a powe
r
ful proce
s
sor h
ad
been wi
dely use
d
in emb
edde
d syste
m
[19-22].
In
o
u
r
s
y
s
t
em, a
ll h
a
r
d
w
a
r
e
w
o
rk
is
per
fo
r
m
ed
by th
e S3c244
0 proce
s
sor. Fi
rst, the pro
c
e
ssor
controls the
motor to
start, accel
e
ration, re
verse,
stop an
d othe
r operation
s
.
Then, S3
c24
4
0
determi
ne
s the batch nu
m
ber of the test strip th
roug
h a built-in scanne
r. To en
sure the pre
c
ise
positio
ning of
the motor d
u
ring the m
e
asu
r
em
ent, motor op
erat
es un
der the
optoco
uple
r
state
judge
ci
rcula
r
ly whe
n
runn
ing. Durin
g
t
he te
st
st
rip
that is
drive
n
by the
moto
r
controlled
b
y
S3c24
40, the photodi
ode
s se
nd the
real-time fluo
rescen
ce info
rmation to th
e analo
g
/digi
t
al
conve
r
ter m
o
dule for p
r
o
c
essing. Fin
a
ll
y, all dat
a are
pro
c
e
s
sed b
y
the S3c244
0 pro
c
e
s
sor,
and
the final resul
t
s are di
splay
ed on th
e screen. Th
e ov
e
r
all stru
cture of
the
inst
rum
ent
is sho
w
n in
Figure 4.
In develo
p
ing
a
scanni
ng-fl
uore
s
cen
c
e
reade
r, t
he
co
re fun
c
tion
m
u
st o
b
tain
sta
b
le a
n
d
accurate flu
o
rescen
ce
info
rmation
and
convert th
e d
a
t
a into
a di
gital si
gnal
for p
r
ocessin
g
. Th
e
entire
pro
c
e
s
sing i
n
volved
inclu
d
e
s
the
optoc
oupl
er,
steppi
ng m
o
tor, LED l
a
m
p
, AD77
07, a
nd
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Scanni
ng
-fluo
r
esce
nce Re
ader Ba
se
d o
n
Em
bedded
System
(Zho
nglon
g Zhao
)
797
other
sp
ecifi
c
h
a
rd
wa
re.
The flo
w
-ch
a
t of the d
e
t
ailed dete
c
ti
on me
ch
ani
sm is
sh
own
in
Figure 5.
Figure 4. Overall struct
u
r
e
of the instrum
ent
Figure 5. Flow-chat of the
detectio
n
Difficulty re
mains in
ho
w to
ca
pture the i
n
formation, tra
n
s
mit the
flu
o
re
scen
ce
informatio
n, and tran
sform the captu
r
e
d
analo
g
sig
nal
into a digital sign
al that ca
n be processe
d
by the S3c24
40 processo
r. Ther
efo
r
e, we sele
cted
a high-se
nsit
ivity photose
n
sit
i
ve tube and
a
glass fiber to
address the p
r
oble
m
.
3.1. Optical Module
The
optical
module
is th
e core
mo
dul
e of a
sca
nni
ng-fluo
re
scen
ce
rea
d
e
r
; th
e opti
c
al
module
controls the inst
ru
mental sig
nal
s and dete
r
m
i
nes the final
result
s. An unsta
ble opti
c
al
module
will cause large di
fferences
am
ong in
stru
m
e
nts and
re
stri
cts the
a
c
cu
racy, se
nsitivity
and othe
r ke
y prope
rties.
Therefo
r
e,
whe
n
we
d
e
s
ign a
n
optical module, th
e stability of the
excitation li
gh
t sou
r
ce
sh
ou
ld be
imp
r
ove
d
, and
the
attenuatio
n
rate
of the
excitat
i
on lig
ht
shou
ld
be
redu
ced. Hen
c
e, we selecte
d
a
hi
g
h
-sen
si
tivity photo
s
en
sitive tube
and
glass fibe
r.
The
detail of optical module i
s
as sho
w
n in
Figure 6.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 3, September 20
15 : 794 – 805
798
Figure 6. De
sign of the optical mod
u
le
Figure 6
sho
w
s th
at light
sou
r
ce 1 i
s
resp
on
sible fo
r the g
ene
rat
i
on of the
ultraviolet
light. Filter
3
filters the e
m
itted light.
Light d
e
tecti
on u
n
it 2
re
ceives th
e flu
o
re
scen
ce
si
gnal
reflecte
d by fluore
s
cen
c
e
strip 7. Filter 4 filters
the received optical
sign
al in which the shap
e of
the optical fib
e
r ap
pea
rs
as a 'Y'.
This
sh
ape of the o
p
tical fibe
r is
re
spo
n
si
ble for
conve
r
ting a
n
d
transmitting the light sig
nal
. The det
ail is as sh
own in Figure 7.
Figure 7. Cro
s
s-sectio
n of the optical fib
e
r
From Fi
gure
7, the left masking a
p
e
r
tu
re 2
con
nect
s
to the LET
lights to tran
smit the
ultraviolet ligh
t
emitted by
the LED lamp.
At
the ri
ght most part, cyl
i
ndri
c
al light is tran
sform
e
d
to
strip li
ght (wit
h the le
ngth
equali
ng the
width
of the
strip test li
ne)
and i
rra
diate
d
to the te
st
strip
throug
h the ri
ght masking
apertu
re 2.
Und
e
r the i
r
radiatio
n of ultraviolet lig
h
t, the fluorescent pa
rticles on th
e strip will
fluore
s
ce; rig
h
t maskin
g a
pertu
re
1
will
tran
smit this fluore
s
ce to
the left ma
sking
ape
rture
1.
The light-se
n
s
itive tube wil
l
t
hen receive
this fluore
s
cence.
A sca
nning
-fl
uore
s
cen
c
e
reade
r usi
ng
ODA-6W
B
-
50
0M photodi
o
de as o
p
tical
receiver
module i
s
a light-sen
sitive photo re
si
stor. Ther
efore, photodio
d
e
s
can ch
ange th
e curre
n
t in the
circuit sen
s
itively by judging the inten
s
ity diffe
ren
c
es of the lig
ht, thus dete
c
ting the lig
ht
intensity. An amplificatio
n circuit in the light-sen
sitive tube amplifies the re
ceive
d
optical si
gn
al.
Therefore,
the
sen
s
itivity of
dete
c
tion in
crea
se
s, captu
r
in
g wea
k
sig
nal
s.
Th
e re
ce
ived
fluore
s
cent would be p
r
o
c
essed by the
light-s
e
n
sitiv
e
tube and t
hen tran
smitt
ed to AD770
7
(anal
og to dig
i
tal conversio
n
chip
) to perf
o
rm the next step.
3.2. Digital-a
n
alog Conv
ersion Modul
e
ADI's AD770
7 is a typical
16-bit
Σ
-
∆
A
DC
chi
p
. This chip m
a
y re
ceive both a hi
gh- a
n
d
low-l
e
vel si
g
nal from th
e
sen
s
o
r
s
directly. AD7707
can
reg
u
late
from zero an
d sp
an throu
gh
instru
ction
s
, i
n
clu
d
ing
self
-re
gulatio
n a
nd sy
ste
m
regulatio
n. When
data tra
n
sfers b
e
twe
e
n
AD770
7 a
nd
MCU, th
e
chi
p
s
ca
n conn
ect to
each o
t
her u
s
in
g a
SPI serial
bu
s inte
rface. T
h
i
s
con
n
e
c
tion g
r
eatly sim
p
lifies the
co
mpl
e
xity of t
he circuit inte
rfa
c
e d
e
si
gn. T
he detail
of this
desi
gn is a
s
a
s
sh
own in Figure 8.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Scanni
ng
-fluo
r
esce
nce Re
ader Ba
se
d o
n
Em
bedded
System
(Zho
nglon
g Zhao
)
799
Figure 8. AD7707 inte
rface
The A
D
conv
ersi
on
pro
c
e
s
s i
s
a
s
follo
ws: t
he li
ght-sensitive tub
e
receives an
d
amplifies
the fluore
s
ce
nt sign
al on t
he st
rip; the t
ube tra
n
smits the sig
nal to
AD770
7; the
chip tran
smit
s
the digital sig
nal to the S3c2440 p
r
o
c
e
s
sor thro
ugh th
e SPI bus.
4. Data Pro
c
essing
The fluoresce
nce d
e
tectio
n
reage
nts ha
ve a te
st and
control lin
e containin
g
fluoresce
nt
particl
es.
Thu
s
, the
dete
c
tion reag
ent m
a
y emit fluo
rescen
ce
un
d
e
r ult
r
aviolet
irra
diation. Af
ter
the test, the i
n
stru
ment
ca
n mea
s
u
r
e t
w
o fluo
re
sc
e
n
ce
pea
ks, reco
rd the
are
a
ratio of the
two
waveforms, a
nd use a sta
ndard test st
rip to per
fo
rm
large
qua
ntities of biol
ogi
cal experim
ent
s.
The followi
ng
mathematica
l
model is e
s
tablished:
"AT/ACrea
gent concentratio
n
of test". Finally,
this mathem
a
t
ical model
ca
n cal
c
ulate th
e unk
no
wn concentratio
n
of the detecti
on rea
gent.
For a
contin
u
ous m
a
thema
t
ical mod
e
l, we use
a de
rivation and
an
alyze the e
q
u
a
tion to
obtain the extreme p
o
ints o
f
a curve.
Acco
rdi
ng to
this process, on
ce
a
te
s
t
is
c
o
mple
te
d
,
we
ob
ta
in
12
00
fluore
s
cen
c
e
data p
o
ints (Origin
[12
00]) after
anal
og
to digital
co
n
v
ersio
n
. As
shown in
Fig
u
re 9,
we pe
rform th
e followin
g
proce
s
sing.
Figure 9. Orig
in data
For th
e
colle
cted
Origi
n
[
1200],
we p
e
r
form
a g
r
adi
ent (simila
r t
o
de
rivation)
to obtain
the Gra
d
ient [1200]. The d
e
tails of this
p
r
ocess a
r
e a
s
as sh
own in Figure 10.
Figure 10. Pri
n
cipl
e of deri
v
ation
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 3, September 20
15 : 794 – 805
800
Whe
r
e
,
,
,
are a
d
j
a
ce
nt points,
and
∆x
1
,
∆
y
,
′
∆
/
∆
The Gradie
n
t [1200] after the gra
d
ient
is as as
sho
w
n
in Figure 11.
Figure 11. Gradient data
From
Figu
re 11,
num
erou
s sa
wtooth p
a
tte
rns a
r
e
n
o
ted in
the
g
r
adie
n
t data.
The
s
e
pattern
s are
not
suitable
for analysi
s
a
nd
p
r
o
c
e
ssi
n
g
.
The
r
efo
r
e, we mu
st
pe
rform a smo
o
thing
operation for the Gra
d
ient
[1200]. This pro
c
e
ss
o
b
tains 5
point
s close to the
current p
o
int
on
each
side
(1
1 in total
)
an
d u
s
e
s
the
a
v
erage
value
of these p
o
i
n
ts a
s
the
n
e
w valu
e of t
he
c
u
rrent point.
,
5
1195
⋯
,
(
1
)
Gradi
ent [120
0] after the smoothing o
p
e
r
ation a
r
e a
s
as sho
w
n in
Figure 12.
Figure 12. Da
ta after the smoothing o
p
e
r
ation
Theo
retically, the poi
nt wi
th its de
rivation eq
ualin
g
0 is th
e extreme p
o
int. F
o
r the
gradi
ent of
discrete
poi
n
t
s, however,
Smooth[120
0]
≠
0 i
s
po
ssible. The
r
efo
r
e, we ne
ed
a
binari
z
atio
n o
peratio
n. Let the
thre
shol
d be X, then we
have:
1,
0,
1,
(
2
)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Scanni
ng
-fluo
r
esce
nce Re
ader Ba
se
d o
n
Em
bedded
System
(Zho
nglon
g Zhao
)
801
The Final
data
[
1200] after the bina
rizatio
n
operation a
r
e as a
s
sho
w
n in Figu
re
13.
Figure 13. Final data after
binari
z
atio
n
From Fig
u
re
13, we obtai
n
the attr
ibutes of the two peaks cle
a
rly:
Cre
s
t:
1
1
0
(
3
)
Start point:
0
1
1
(
4
)
End point:
1
1
0
(
5
)
Width:
_
_
(
6
)
Area:
1
(
7
)
The im
prove
d
alg
o
rithm
is obtain
ed
by
comp
ari
ng Fi
gure
9
and
Fi
gure
13,
Equ
a
tion
(3
)
s
h
ows
that c
r
es
t is
loc
a
ted
to the left of
the actu
al po
sition. Assum
e
the following:
Point X1:
0
1
1
Point X2:
1
1
0
The cre
s
t mu
st be located
in the rang
e o
f
,
, and the cre
s
t is the maxi
mum value in
betwe
en this i
n
tervals. Th
u
s
, the crest is:
,⋯,
,
9
The
gra
d
ient,
sm
oothing,
a
nd bi
nari
z
atio
n op
er
atio
ns
above
are
for the id
eal
ca
se. After
compl
e
ting th
e instrume
nt desi
gn, difference am
o
ng instru
ment
s are
p
o
ssibl
e
becau
se
of
t
h
e
differen
c
e
s
among
the mech
ani
cal pro
c
e
ssi
ng, hard
w
a
r
e ci
rcuitry,
an
d optical mod
u
l
es.
Variation
s
in the pro
d
u
c
tion
of test strips
may al
so inf
l
uen
ce t
he de
v
i
ces.
I
n
act
u
al t
e
st
s,
sev
e
ra
l
small pe
aks a
ppea
r. Therefore, a
thre
sh
old must be e
s
tabli
s
he
d to filter the interferen
ce p
e
a
k
s.
Figure 14. Re
sult of pea
k d
e
tection
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 3, September 20
15 : 794 – 805
802
As sho
w
n in
Figure 14, se
arching in thi
s
crest re
sult
ed in the values di
splaye
d in table
1.Both test peaks an
d co
n
t
rol pea
ks h
a
v
e been a
c
cu
rately detecte
d.
Table 1. Re
sult of peak re
cog
n
ition
Peak 1
Peak 2
Peak Start
202
706
Peak End
600
1029
Coordinate
421
859
Peak Value
4055.3
1910.6
Peak Area
657959
343575
5.Resul
t
s an
d Analy
s
is
T
h
e
pr
e
v
io
us
wo
rk
s
u
cc
es
s
f
u
lly des
ig
n
ed a
scan
ning
-fluorescen
c
e
re
a
der
and
compl
e
ted th
e prep
aratio
n
of the hardware an
d
software pro
g
ra
mming. The followin
g
se
cti
ons
will show
sev
e
ral important
test
s of performance indicators.
The te
st st
rip
is a
s
follo
ws (Figure 15
), fo
r
which the
control li
ne
(C
line) i
s
th
e go
at anti-
mouse p
o
lycl
onal
antibodi
es, the
test li
ne (T lin
e)
i
s
the 7
D
9
antib
ody (fo
r
coati
ng), the
sam
p
le
in the pad
s is the 10
C5
antibody
(containin
g
fluore
s
cent pa
rticles, for m
a
rki
n
g
)
, and
the
measured o
b
j
ect i
s
the
CRP (C-re
a
cti
v
e protei
n)
.
Whe
n
serum
wa
s
adde
d, the effe
ct of
the
absorb
ent pa
d, CRP, 10C5 antibody, and 7D9 a
n
ti
b
ody in the blood are com
b
ined a
s
a st
able
stru
cture at the T line; the
105
C and
g
oat anti-m
o
u
s
e polycl
onal
ant
ibody al
so
bind
s in a
stable
st
ru
ct
ur
e.
Figure 15. Te
st strip
We obtai
ned
the CRP con
c
entration in
the se
rum by
using the
scannin
g
-fluo
r
e
s
cen
c
e
read
er
be
cau
s
e the
CRP concentratio
n
i
n
a he
alth
y body is lo
w (<5 mg/L); h
o
wever, if the b
ody
is infected by bacteria or t
he ti
ssue is i
n
jured, the co
ncentration
will increase si
gnificantly. Thus,
we can dete
r
mine the heal
th of the blood. We used this
test stri
p in all of the followin
g tests; the
detailed te
sts and data are
as follows.
5.1. Linear Relationship
Whe
n
we used stand
ard
strip
s
,
fixed sample
s of
kn
own
con
c
entration
were di
luted i
n
different p
r
op
ortion
s an
d th
e fluore
s
cen
c
e inten
s
ity wa
s kno
w
n. The
read
er m
e
a
s
ured th
e valu
e
of the are
a
o
f
the control l
i
ne an
d test l
i
ne an
d dete
r
mined the li
n
ear
relatio
n
ship of the d
a
ta.
The data
covering the fixe
d con
c
e
n
trati
on sam
p
le a
r
e sho
w
n b
e
lo
w (Ta
b
le 2):
Table 2. Fluo
resce
n
ce inte
nsity and AT/AC relatio
n
sh
ip
No.
Fluorescence
intensity(a.u.)
AT AC
A
T/AC
1 34.375
1969.20
97561.4
0.020
2 68.75
4662.80
99203.3
0.047
3 137.5
10699.9
92682.6
0.115
4 275
24025.7
106667
0.225
5 550
52429.5
101452
0.517
6 1100
106998
103624
1.033
7 2200
189900
90705
2.094
8 4400
399811
95550.3
4.184
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Scanni
ng
-fluo
r
esce
nce Re
ader Ba
se
d o
n
Em
bedded
System
(Zho
nglon
g Zhao
)
803
By adding
tre
nd-lin
es, fo
rm
ulas an
d
co
rrelati
on
co
efficient
s in
excel, we
an
alyzed the
s
e
data and o
b
ta
ined fluorescence inten
s
ity AT/AC
linear relation
ship
a
s
sh
own in Figure 1
6
.
Figure 16. Fluore
s
cen
c
e i
n
tensity AT/AC linea
r rel
a
tionship
Figure 16 sh
ows that the linear relationshi
p bet
we
en the fluore
s
cen
c
e inten
s
ity and
AT/AC is
y
=
0.
00
10x
-
0
.
0
181
, the co
rrelation
coeffici
ent
2
R
=
0
.
99
99
8
illustrating the excell
ent
linear
rel
a
tion
ship. T
herefo
r
e, we can u
s
e thi
s
line
a
r relation
shi
p
as the fo
und
ation for fu
rther
measurement
s of the test strip
with un
kn
own
con
c
e
n
trations.
5.2. Reading Repeatability
Dete
cting th
e
identi
c
al te
st strip
multipl
e
time
s verifi
es th
e
stabili
ty of the
sca
nning
-
fluore
s
cen
c
e
read
er. Figu
re 17 displays
the results
fro
m
7 repe
ated
trials of the identical strip.
Figure 17. Re
sults of
repea
tability detection
Figure 17 di
splay
s
high repro
d
u
c
ibility.
A fine coincide
nce wa
s found amon
g these
detectio
n
s. Among them, a small atten
uation wa
s
n
o
ted at the crest; enlarging
this portion
data
(from
370 to
470 (101 i
n
total))
re
sults i
n
Figu
re
1
8
. From Fi
gure
18, the fluore
s
cen
c
e inte
n
s
ity
of
ea
ch
te
st has
a slig
ht attenuation. Whe
n
ul
traviolet excite
s
a fluor
esce
nt su
bsta
nce, th
e
fluore
s
cent substa
nce ab
sorb
s en
ergy,
jumping
from
the gro
und
state to an e
x
cited state a
n
d
then ba
ck to
grou
nd
stat
e throu
gh
ra
diation.
Whe
n
excitation
stop
s, fluore
s
cent oxidati
on
grad
ually we
ake
n
s th
e intensity of the mole
cul
a
r fluorescen
c
e emi
s
sion.
Therefore,
the
deviation
s in
Figure 1
8
a
ppea
red. By con
d
u
c
ting a
large nu
mb
er of experi
m
ental analy
s
e
s
,
small attenu
a
t
ion of fluorescen
c
e inte
nsit
y has no effe
ct on the test results.
Evaluation Warning : The document was created with Spire.PDF for Python.