2012 IEEE Conference on Sustainable Utilization and Development in Engineering and Technology (STUDENT)

Universiti Tunku Abdul Rahman, Kuala Lumpur, Malaysia. 6 - 9 October 2012

Non-invasive blood glucose measurement

Application of near infrared optical measurement

Anas, M.N1, Syafirah, A.2, A.N. Norali, Normahira, M.

Biomedical Electronic Engineering,

School of Mechatronic Engineering,
Universiti Malaysia Perlis.
E-mail: anasmohdnoor@gmail.com1
syafirah.azahari@gmail.com2

Abstract—In clinical practice currently, blood glucose is
measured using an invasive techniques which is unsuitable and
having risky of infection in the patient since the finger is pricked
three to five times a day. In this study, near infrared optical
measurement is applied to overcome the invasive issues. The
designed device consists of an infrared emitter which placed over
the fingertip for measurement of blood glucose optically. The
near-infrared light will be sensed by the photo diode which
having a certain wavelength. The intensity of received light is
depending on the glucose molecules inside the blood. Tensubjects
have been used during the experiment set. Early result is
presented that blood-glucose level range starting from 6.3 to
11.5mol/L during normal non-fasting, voltage measured from the
design device having of 1.4V to 1.66V and during fasting, the
voltage range is about 1.38V to 1.61 V with reading of 4.7 to 5.5
mol/L. The prediction of blood glucose reading for 10 subjects
also is presented.
Keywords- Near Infrared, blood glucose,optical measurement
I. INTRODUCTION
Diabetes mellitus has become the serious medical
condition in which it may cause blindness, obesity, heart
disease, renal failure and the stroke [1-3]. These killer disease
effects on the ability of human bodies in producing and
utilization of the insulin. It is a hormone which is essential in
processing the blood glucose. The World Health Organization
estimated that there will be 177 million who suffered diabetes
in 2000. By the year 2025, the value will increase up to 300
million minimally [4-6].
Currently, the blood glucose has been measured by
pricking the patient's finger using a lancet to extract a small
quantity of blood at blood capillary [7]. Blood extracted will
be placed on the disposable test strip and insert into the blood-
glucose meter. Next, a series of chemical reaction will take
place. The reaction between the glucose and the substances on
the test strip will produce potassium ferrocyanide; a
compound consists of an anion [Fe(CN)6]4- and cations K+.
Potassium Ferro cyanide will react with the metals on the
electrode layer and caused the electric current to flow through
the electrodes. The more glucose from the blood, the more
gluconic acid is produced thus, more Ferro cyanide anions will
be formed, which increases the electric current. The electric
current flows through the glucose meter and the
strength of this current is consumed to compute
the amount of glucose from the blood [8].
The blood-glucose meter will convert it to the digital
value and give the blood-glucose reading. In order to monitor
the disease, the recommended self-test for the type 2 diabetic
patient is 4 to 7 times per day. However, the average diabetic
patients test the blood glucose is less than twice a day on the
average of 1.6 times each day. The main issue for the invasive
finger pricking for glucose tests are painful. Where the
requirement needs them to prick at least once each time the
glucose test is done [8]. Consequently, the pin prick may be a
resultant in bruising and the high risk of the infection. For the
last decade, a non-invasive technique has become an interest
to overcome the conventional measurement issues such as
pain, risk of infection and high cost. There many new methods
in order to improve the blood-glucose measurement and the
optical method has become interest in the most research
studies. It is more reliability and has effective cost [9]. The
optical method of blood-glucose measurement is the most
popular method which has been studied and research in regard
to provide a non-invasive measurement. There are a variety of
the optical methods for the non-invasive technique, including
near-infrared (NIR), Raman's spectroscopy, photo acoustic
spectroscopy, polarization technique, polarimetry and light
scattering [10,11].
In the Near Infrared Spectroscopy (NIR), glucose as
the biological component in the human blood will produce the
weakest NIR absorption signals. In measuring the glucose
level, the NIR spectroscopy enables the exploration of tissue
depth within the range of 1 up to 100 millimeters with
decrease in penetration depth as an increase in wavelength
value [12].
II. METHODOLOGY
A. Experiment set up
In the process to develop the system, kindly it is divided into
several main parts and need crucial designs in each stage.

978-1-4673-1705-4/12/$31.00 ©2012 IEEE 258

 The experimental circuit set up using near infrared (NIR)
spectral range to measure the blood glucose. The infrared
Electronic Circuit
light-emitting diode, LED (NIR LED) used has peaked at
wavelengths of 940nm. The height of NIR LED is
Data Acquisition
34mm±0.3mm with the diameter is 5.9mm±0.3mm.
Meanwhile, the lead spacing is 2.54mm On the other hands,
with the same characteristics with the NIR LED, the photo
GUI (LabVIEW) diode detection is in the spectral range of 700nm to 1050nm.
The peak wavelength for this receiver is 940nm. The data
recorded show differences of voltage value related to their
Regression analysis blood-glucose alterations. In addition, the metabolism
rates in everyone not same. By considered the metabolism rate
with the fluctuation of the blood glucose, obviously the data
Glucose prediction obtained is not similar to each other. Figure 2 shows the
receiver's circuit for converting current generated by a photo
diode into a voltage signal for advance amplification [4]. The
Figure 1: The block diagram of the design system.
output signal of the trans-impedance amplifier, Vout is
connected to the data acquisition for interfacing to computer.
Figure 1 show the experiment set-up of non-invasive
technique to measure the blood glucose by sensing the change
in the blood glucose in the fingertip.

B. Electronic Circuit
Basically, the non-invasive technique designed is
based on the principle of the absorbance transmittance
photometry. The absorption value of light energy is dependent
on the number of molecules present in absorbing material.
Hence, the intensity of the light energy when leaved the Figure 2: Current to voltage converter circuit
absorbing material is consumed as an indication of
concentration of that particular material. In quantitative, the
absorbance is expressed by the Beer Lambert Law C. Graphical User Interface
T = I / I0 (1) The designing system required the LabVIEW to
recording and save the subject’s data in a programmed placed.
The transmittance of the sample is directly measured by taking The system is continuously record for 3 minutes for each
the strength of the wavelength measured and dividing it by the recording. Each subjects will undergoes 2 times. The signal
initial strength. The absorbance is computed as recorded is keep in a particular file which is instructed by the
users.
A = -log10 (T) (2) D. Measurement Procedure and Data analysis
The experimental procedure established is to have an
In addition, the absorbance is equal to the abc, in which the
accurate and valid data among the subjects. Figure 3 shows the
absorptively coefficient, a multiplied by the path length, b,
block diagram on procedure of experiment. The subjects
multiplied by the concentration, c. In the meantime; the actual
consider healthy. All 10 subjects are female. There are some
glucose level will be measured against a baseline wavelength
assumptions made during the experiment as below
which changes a little with glucose levels (λ). Wavelength
with the glucose levels is (λ'). So, the actual concentration will
i. The skin temperature for each subject is almost the
be
same, 36.5°C.
A(λ) - A(λ') = bc(a-α) (3)
ii. The index finger diameter is approximately same
among the subjects, 10.1mm.
Assume that the glucose level is known, so (a-?) can
iii. The fingertip diameter is assumed same among the
be computed from the equation (3). Once it is known, all the
subjects.
other concentrations measured is determined as a ratio of the
iv. The skin color is light intermediate among the
initial concentration, where, a-? Both are constants. Then, the
subjects which is depends on the presence of the melanin
ration is checked against the known concentration of the
in the skin.
glucose solutions to describe the characteristics equation
as the ratio to the actual concentration.

259
 Subject fasting for 8 Hours
Measurement using hand-prick and designed device
Subject break-fasting, having 50g of glucose drinking
After 30 minutes, subjects will be measure again using
hand-prick and designed device
Data obtained will be analyze and prediction of blood
glucose using the design device
Figure 3: The measurement procedure.
Data obtain will be analyzed using simple polynomial
regression. Regression analysis is the statistical technique for
the purpose to identify the relationship between two or more
quantitative variables. The relationship is focusing at a
dependent variable whose value is to be predicted with the
independent variable. This analysis is used to determine and
predict the relations between an independent variable X
(Voltage measured, (V)) and a dependent variable Y (Blood
glucose concentration, (mol/L)).
III. RESULTS
The purpose of this design system is to determine a
method for the prediction of blood-glucose level for human.
This invented system would enable the monitoring blood
glucose level continuously and noninvasively. When the use
of the noninvasively technique coupled with the several
advantages such as the absence of pain, it will
offer securely a technique to the user.
The graph plotted presents the designed circuit can distinguish
the reading between the fasting and non- fasting subject. The
voltage value for each subject has a huge different among
them, and only several data has been slightly different.
However, it still determines the designed system capable of
give a reading for blood glucose between fasting and the non-
fasting people. Figure 4 illustrates the difference's voltage
measured from fasting and non-fasting subjects.
Figure 4: Data obtain from the design device
For the fasting subjects, the voltage is altering from -1.619 V
to -1.381 V. By the way, the non-fasting subjects exhibit the
voltage varying from -1.663 V to -1.406 V. It is concluded in
the table below. The voltage measured between the fasting and
non-fasting were statistically significant.
Table 1: Voltage measured for the subjects
Voltage measured
Subjects
Min. (V) Max. (V)
Fasting -1.619 -1.381
Non fasting -1.663 -1.406
Next, the prediction of the glucose is determined. By using the
2nd order of the polynomial regression, the prediction is
accomplished by using the equation generated
mol/L = 97.82 * X^2 + 298.8 * X + 233.8 (4)
Among the orders in the polynomial used in this study, the 2nd
order of the polynomial regression provided higher correlation
of the blood glucose concentration reading (mol/L). The
experiment data are inputted into the regression analysis and
modeled as Figure 5. The curve demonstrates the relationship
between the independent variables X (Voltage measured, (V))
and the dependent variables Y (Blood glucose concentration,
(mol/L)).
Figure 5: The 2nd order polynomial regression analysis
260
 [5] WHO. Definition, diagnosis and classification of
The corresponding voltages measured and the blood diabetes mellitus and its complications. Geneva.
glucose concentrations are tabulated. Next, the prediction for 1999.
blood glucose concentration is performed. The voltage value is [6] V. Ashok, A. Nirmalkumar and N. Jeyashanti. A
chosen and substitute in the formula. As a result, it will induce novel method for blood glucose measurement by
a prediction for the glucose prediction. Table 2 summarized noninvasive technique using laser, 2011.
the correlation between the voltage values with the estimation [7] M. B. Savage. et al., Development of a non-invasive
of the blood glucose concentration. blood glucose monitor: application of artificial neural
networks for signal processing, 2000.
Table 2: Summary of the predicted blood glucose concentration
[8] S. Oshima. and Y. Sankai. , Optical measurement of
(mol/L)
blood hematocrit on medical tubing with dual
Voltage (V) Blood glucose wavelength and detector model, 2009.
concentration (mol/L) [9] Peled, N., Wong, D. and Gwalani, S.L. , Comparison
of glucose level in capillary blood samples obtained
-1.555 6
from a variety of blood sites, Diabetes technology
-1.411 7 and therapeutics, 2002.
-1.364 8 [10] Eun-Hyung Yoo, Soo-youn Lee., Glucose biosensors:
An overview of use in clinical practice.” Sensors,
-1.287 11 2010.
-1.222 15 [11] Goetz, M.J. et al., Application of a multivariate
technique to Raman spectra for quantification of
The investigation between the voltage value and the body chemicals.” IEEE Trans. Biomed. Eng, vol. 42,
concentration of the blood glucose is summarized. Smaller pp.728-731, 1995.
negative voltage indicates higher concentration of blood [12] Zuomin Zhao., Pulsed photoacoustic techniques and
glucose. glucose determination in human blood and tissue,
Academic dissertation, Faculty of Technology,
IV. CONCLUSIONS University of Oulu, Finland, 2002.

The experiment result for prediction of the blood

glucose level using the correlation between the voltage
measured. The difference in voltage measured between fasting
and non-fasting could be differentiated. On the other hand,
many parameters for improvement could be made to increase
accuracy and precision of measurement using the device.

REFERENCES

[1] Waynant, R.W. and Chenault V.M. , Overview of non-

invasive fluid glucose measurement using optical
techniques to maintain glucose control in diabetes
mellitus, Vol. 12/2, 1998.
[2] Jaspreet et al., Non invasive blood glucose measurement
using optical method, feasibility study, and design issues,
ICOP 2009-International Conference on Optics and
Photonics, 2009.
[3] Nakamachi, E. ,Development of automatic operated
blood sampling system for portable type self-monitoring
blood glucose device, 32nd Annual International
Conference of the IEEE EMBS, Buenos Aires,
Argentina, pp. 335-338,2010.
[4] Anitha Soni. , Diabetes Management: Tests and
treatments among the 18 and older U.S.Civilian non
institutionalized population in 2003, pp.1, December,
2005

261