back to main pageOf
ISTC 2259p
May
2004
This work was supported financially by Korea (Samsung Electronics Co., Ltd.by and through R&D Center, Samsung Advanced Institute of Technology) and under the contract to the International Science and Technology Center (ISTC), Moscow.
Investigation of Dielectric
Properties of Sugar Solutions
In the
Millimeter –Wave Range and the Design of Devices for the Control of Sugar
Content in Blood
(From 1 November 2003 to 30 April 2004 for 6
months)
Viacheslav Viacheslavovich Meriakri (Project
Manager)
Institute of Radio Engineering and Electronics
Russian Academy of Sciences
Fryazino Branch
The objective of this project is to investigate the dielectric properties of sugar solutions in water, in an imitator of blood, and in blood, as well as to investigate the possibility of using millimeter waves for sugar (glucose) control in human blood, including non-invasive control.
A method for the
non-destructive measurement (by measuring the reflection only) of glucose
content in such solutions is elaborated and realized.
Dielectric properties
of glucose solutions with concentrations of W from 5 to 0.25% wt. are
measured in the frequency range 30–93 GHz and at 10 GHz.
Dielectric properties
of blood (using one drop) are investigated in vivo at frequencies of 42
and 66 GHz. The measurements have shown that the method developed in this
project can be used for the real-time determination of the glucose content in
blood without strips used in optical invasive glucometer after oral glucose
tolerance test (OGTT).
Dielectric properties
of skin in the frequency range from 30 to 80 GHz are determined. It is shown
that results of the project allow, in principle, to create a non- invasive
device for the glucose content in blood determination, at least after OGTT.
Recommendations for
the continuation of the project and the practical application of the results
obtained are given.
Keywords: millimeter waves, dielectric properties,
glucose solutions, sugar in blood determination
141190, Fryazino Moscow Region, Vvedenski sq.1, Russia
Phone =7 (095) 5269266, Fax + 7 (095) 7029572
E-mail: meriakri@ms.ire.rssi.ru
The objective of this project is to investigate the dielectric properties of sugar solutions in water, in an imitator of blood, and in blood, as well as to investigate the possibility of using millimeter waves for sugar (glucose) control in human blood, including non-invasive control.
These investigations are fundamental since they
allow one to obtain valuable information about the interaction between
electromagnetic waves and liquids, on the other hand, they are related to
applied investigations since the knowledge of the properties of these liquids
allows one to develop methods and devices for the nondestructive (noninvasive)
real-time control of their composition. One of such practical problems is the
measurement of sugar (glucose) content in various liquids, including blood. The
importance of such investigations as applied to the problem of diabetes is
obvious; they are of interest also for other problems, for example, for the
determination of water and sugar content in wine and some other liquids.
The
measurement methods were chosen so that, subsequently, one could use them for
the noninvasive control of the glucose content in blood (i.e., the methods are
based on the measurement of the reflection of an electromagnetic wave, without
penetrating into the medium). To determine the e˘ and e˘˘ of a medium, one has to measure two parameters of the reflected wave.
In noninvasive methods one usually
employs a sophisticated and expensive vector network analyzers and measures the
modulus |r| and phase j of the reflection coefficient |r| eij (|r|2 = R is the power reflection coefficient).
We developed a sufficiently simple metod and scheme for determining e˘ и e˘˘,
which consists in measuring the minimum of the
reflection coefficient R(f ) = Rmin(fmin)
and frequency fmin corresponding
to this minimum from the following structure: a specially chosen plane-parallel
matching plate made of a low-loss dielectric—a medium under measurement with
high losses (water, solution, blood, skin).
This method was realized in frequency range 28- 150 GHz and at
frequencies near 10 GHz and following measurements were conducted.
1. Investigation of solutions.
Dielectric properties of glucose solutions in water and in a blood imitator have been measured for the first time on the equipment of the executors of the project in the frequency range from 28 to 93 GHz (and at a frequency of about 10 GHz) for glucose concentrations W ranging from 5 to 0.5% wt. (an even up to .25% in some cases). Sensitivities of 2.2 dB per 0.5% wt. of glucose concentration in water and 0.9 dB per 0.2% wt. in physiological solution have been obtained. Extrapolation of these results to smaller values of W shows that an increase in the dynamic range of the measurement device may increase the sensitivity to 0.04% wt. (2 mmoles/l). This can be achieved by increasing the signal-to-noise ratio of the directional couplers available and by using ferrite circulators. The measurement scheme developed by the executors of the project may serve as a basis for the design of a laboratory or industrial equipment for controlling small concentrations of glucose (sugar) in water and in physiological solution.
2.
Investigation of blood.
Dielectric properties of blood are measured in vivo (without preservatives) for the first time with a sufficiently high accuracy (the measurement accuracy of that e˘ and e˘˘ is ±0.2) at frequencies of 42 and 66 GHz.
The method developed in the project
allows a real-time determination of glucose content in blood by a single drop
of blood. After a series of experiments in a thermostatically controlled
chamber (unfortunately, temperature in the chamber was stabilized only to
within ±1°C for the present), we established the
following.
(a) There is a clear correlation
between W and the measured value of Rmin as W increases after an oral glucose tolerance test (OGTT) on an
empty stomach. The functions Rmin(W) show individual behavior for each test person.
(b) The decrease in W after the maximum and the increase in
W after the minimum are not so clearly correlated with the variations in
Rmin. This may be attributed to some presently unknown
changes in blood within this time interval after the maximum of W, which
may be associated with physical activities, digestion process, etc. Our
experiments have shown that these changes cannot be related to the content of
hemoglobin in blood, because the latter was virtually invariant during the
measurements. As for the content of cholesterol, its concentration decreased in
all test persons during the whole period of measurements (i.e., when W
increased, decreased, and again increased). This may affect the behavior of Rmin during the second rise
of W. The determination of reasons for such behavior of blood properties
requires additional analysis, including medico-biological investigations.
Thus, one can carry out real-time
invasive measurements of W after oral glucose tolerance test (OGTT)
without using expensive strips for optical glucometers. After appropriate
modifications (the thermostatting of the measurement plate and the surrounding
medium to at least within 0.5°C), such a measurement device can be supplied to
hospitals for systematic measurements of W. These measurements, combined
with medico-biological tests, may give a final decision about the practical
applicability of the device.
3. Investigation of skin.
The reflectivity of skin on various parts of
human body has been measured on a number of test persons at different frequency
intervals in the millimeter-wave band. It has been established that, for close
values of W, fingertips,
on the one hand, and forearms and earlobes, on the other, have substantially
different values of Rmin
and fmin. Moreover, they
have individual values for different test persons. The values of e˘ and e˘˘ have been measured near the elbow joint in the
frequency range from 30 to 80 GHz. Above 40 GHz, these data have been obtained
for the first time, whereas, at frequencies of 30—40 GHz, the values of e˘˘ obtained in our experiments are higher
than those available in the literature.
As for the determination of the glucose
concentration W by measuring Rmin and fmin, the series of experiments carried out within the
project (including those performed with specially fabricated matching plates
that provide Rmin < -30 dB for skin) show a correlation between
W and Rmin in the region where W increases after taking glucose on an empty stomach at frequencies
of about 40 and 60 GHz; however, a further variation in W (when W decreases,
especially in the afternoon) does not display a clear correlation between W and Rmin. This is likely to be attributed to the physical
and physiological activities of test persons and requires further investigations.
We did not observe any effect of the content of hemoglobin. As for the content
of cholesterol, it may manifest itself in the region of decreasing W, upon reaching the maximum of W. However, the main contribution to the
error in measuring small variations of W
is made by the dependence of Rmin
on the position and the pressure of the measurement plate to the skin. This
error has been reduced in the device described in Appendix 2, where the contact
area between the plate and the skin was increased and the measurements were
carried out under the conditions of OGTT in the absence of physical activities.
As a result, we obtained a good correlation between W and the output signal of the device.
Possible continuation of the project.
(a) As
regards the investigation of solutions with low concentration of glucose, it is
possible to design a device with increased signal-to-noise ratio, which will
allow the measurements of W < 0.05% wt.
(b) As regards the invasive control
of blood without using expendable measurement strips for an optical glucometer,
our investigations show possiblity to
design a device with temperature-controlled measurement plate for measuring W under clinical conditions. This device
can be designed on a compact semiconductor oscillator and will allow real-time
measurements of W.
(c) As for the
noninvasive control of W through
skin, the equipment elaborated in the
Project should be modified (with the application of a semiconductor oscillator,
a special means providing a constant pressure of the measurement plate to the
skin, and with increased signal-to-noise ratio). In this case, it is possible
to design a device for the noninvasive control of W in human blood, at least under conditions of OGTT, in the absence
of physical activities, and with individual adjustment of the device for each
test person. The problem about the design of a universal device requires
additional investigations.
back to main page
back to project page