OPTICAL SPECTROSCOPY

AND OBSTACLES BY NON-INVASIVE DETECTION OF

GLUCOSE CONCENTRATION BY HOME MONITORING

O. Abdallah, Q. Qananwah, A. Bolz

Institute of Biomedical Engineering IBT, Karlsruhe Institute of Technology KIT, Karlsruhe, Germany

J. Hansmann, H. Walles, T. Hirth

Fraunhofer Institute for Interfacial Engineering and Biotechnology IGB, University of Stuttgart, Stuttgart, Germany

Keywords: Glucose Management, Raman Scattering IR Spectroscopy, Fluorescence Spectroscopy, Blood Components

Concentration Monitoring, High Signal to Noise Ratio, Parameters Affecting In-vivo Glucose

Measurement.

Abstract: Tight glycemic monitoring and control is the main goal in successful diabetes management to avoid its

complications. Frequent blood glucose measurements with a combination of regimented diet, exercise and

insulin administration can accomplish this task. Different methods are applied for non-invasive

measurement of blood glucose concentration. Despite the great interest and the intensive research in this

field since 1980s, there is no convenient device at the market that can measure the glucose concentration

non-invasively in an easy manner. This paper discusses the different methods for detecting the glucose

concentration. Elastic and inelastic (Raman) scattering as well as fluorescence and IR Spectroscopy

measurements well be shown and discussed for the development of a compact non-invasive device for home

monitoring. In conclusions, an optical multi-sensor measuring the fluorescence and light scattering in the

tissue optical window in and around visible range (360 nm – 1200 nm) taking the perturbation factors into

account is promising and under development.

1 INTRODUCTION

Diabetes risk lies in its complications like heart

diseases and infarcts, stroke, blindness, kidney

disease, nerve disease, diabetic foot and amputation.

The current applied invasive methods are

intermittent, inconvenient and painful, having

infection risk, blood loss and time delay, need

consumables materials, needles and strips. The

invasive method cannot be applied continuously, and

hence hypo- or hyperglycemia may be not detected.

An easy accessible and low-cost method for

continuous glucose concentration monitoring and

diabetes management will be a great help for more

than 250 millions of diabetic patients worldwide to

avoid the risks and the complications caused by

hyper- or hypoglycaemia. Standard treatment

includes lifestyle changes, medication and frequent

monitoring of blood glucose levels.

Insulin and other

diabetes medications are designed to lower the blood

sugar level when diet and exercise alone aren't

sufficient for managing diabetes. By the

development of a compact system, different LASER

diodes (LD`s) or light emitting diodes (LED`s) in

the range of UV and NIR will be used. Light

scattering and fluorescence spectroscopy can be

applied for non-invasive measurement of blood

components like glucose concentration or the early

detection of pathological variations like melanoma.

The complications of diabetes are largely

avoidable and may be reversed by strict control of

blood sugars through medication and diet. Patients

with type 2 diabetes mellitus are at increased risk for

macrovascular disease complications (Gaster, 1998;

Mezzetti, 2000; Pambianco, 2006). Detection of

blood contents like Glucose non-invasively in an

easy manner can reduce morbidity and mortality by

diabetics.

291

Abdallah O., Qananwah Q., Bolz A., Hansmann J., Walles H. and Hirth T..

OPTICAL SPECTROSCOPY AND OBSTACLES BY NON-INVASIVE DETECTION OF GLUCOSE CONCENTRATION BY HOME MONITORING.

DOI: 10.5220/0003796902910296

In Proceedings of the International Conference on Bio-inspired Systems and Signal Processing (BIOSIGNALS-2012), pages 291-296

ISBN: 978-989-8425-89-8

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

1.1 Non-invasive Methods for

Detecting Glucose Concentration

In-vitro Glucose concentration measurements

depend on chemical principles. The non invasive

monitoring methods are much more difficult to

apply with the required accuracy. Glucose

concentrations in blood are very low compared for

example with that of hemoglobin. The effect of

glucose on the measured signals is too low and

hence a high amplification is needed, which means

that other background and surrounding noises will

also be amplified and hence the measurement will be

very sensible for tiny perturbations.

The different methods for non-invasive glucose

concentration montoring are discussed in diverse

literature (Khalil, 2004; Yamakoshi, 2006; Maruo,

2003; Tura, 2010). Raman and fluorescence

spectroscopy as examples of promising optical

methods will be briefly discussed in the next section.

1.2 Optical Spectroscopy

The optical methods rely on the interaction between

light and tissue. They are widely used by different

research groups and companies.

1.2.1 Raman Spectroscopy

Raman spectroscopy measures scattered light that

has been influenced by the oscillation and rotation of

the scattered molecules. Various Raman techniques

have been attempted in blood, water, serum, plasma

solutions and the eye, but multiple problems remain

before human studies can be accomplished.

Analytical problems include instability in the laser

wavelength and intensity, errors due to other

chemicals in the tissue sample and long spectral

acquisition times. The applying of special types of

Raman spectroscopy can greatly enhance the signal

noise ratio, the resolution and sensitivity. Resonance

Raman (RR) scattering, surface enhanced Raman

spectroscopy (SERS)

, coherent anti-Stokes Raman

scattering spectroscopy (CARS), and Stimulated

Raman scattering (SRS) are examples of the Raman

enhancement methods.

1.2.2 Fluorescence Spectroscopy

Fluorescence spectroscopy and time resolved

fluorescence are dominant methodologies and used

extensively not only in biochemistry and biophysics,

but also in biotechnology, medical dioagnostics and

genetic analysis (Moschou, 2004, Pickup, 2005,

Lakovics, 2006). The technique is extremely

sensitive. There are increasing examples of even

single-molecule detection using fluorescence

methods. Many studies indicate that fluorescent

technology has real sensitivity especially in low

glucose ranges. In addition, since near-infrared light

passes through several centimeters of tissue, with the

appropriate choice of fluorophore, molecules can in

theory be excited and the emission interrogated from

outside the body providing the potential for

completely non-invasive sensing. A convenient way

of classifying fluorescence-based glucose sensors

that involve measurements of fluorescence is either

according to the type of molecular receptor for

glucose, or whether cells or tissues are used to signal

glucose concentrations and/or glucose metabolism.

A review of the principles of operation and current

status of the various approaches to fluorescence-

based glucose sensing are described in D’Auria,

1999.

In DMEM solution a glucose dependent

autofluoresence can be observed. The fluorescence

differs from the process of Raman Effect in that the

incident light is completely absorbed and the system

is transferred to an excited state from which it can

go to various lower states only after a certain

resonance lifetime.

2 APPARATUS AND METHOD

Our measurements were obtained using a micro

Raman spectrometer, based on Olympus IX71

microscope by Fraunhofer Institute in Stuttgart. The

separation of spectrums is achieved using holo-

graphic grill in spectrograpic Holospec f/1.8 (Kaiser

Optical Systems). Spectrum detection attained using

a CCD for NIR (DU420A-BR-DD, 1024x256 Pixel

von Andor). The measurements are done using a

glas bottom dish Willco Welles.

The method discussed here can be applied for

invasive and non-invasive measurement. Photo-

diodes or phototransistors for light scattering and

fluorescence detection in the visible and NIR

spectrum are used. Light emitting diodes LED and

LASER diodes as light sources are applied. Variable

frequency and duty cycle can be adjusted for time

resolved fluorescence signal detection.

The developed system is flexible and can be used

for the development purposes, where different

parameters have to be adjusted. Light intensities,

duty cycle, different LASER types and variable

amplification can be acquired using this system.

BIOSIGNALS 2012 - International Conference on Bio-inspired Systems and Signal Processing

292

In the meantime we are developing a time

resolved spectroscopy system in order to take more

parameters into account by the calculation of

glucose with a multisensor. By applying a

multisensor for the detection of glucose

concentration a great attention has to be given for

the calibration method.

As we discussed in other papers (Abdallah,

2010) and as we can see in diverse literature the

signal to noise ratio has to be kept as high as

possible, but the glucose signal is too low, so that in

addition to diminish the noise and considering all

parameters affecting the measurements an adequate

calibration is necessary for accurate glucose

measurement non-invasively. The scattered signals

may come mainly from deep tissue and blood

glucose may be taken as reference. Moreover

hemoglobin concentration and oxygen saturation

have to be taken in to account by these

measurements. Direct invasive measured values can

be taken as reference. But fluorescence signals may

come mainly from intestinal skin fluid ISF. Transfer

of glucose from the blood to the ISF compartment

occurs by passive diffusion through an established

concentration gradient. The mass transfer rate is

affected by several variables, such as the blood flow

rate to the site, rate of glucose uptake by the

surrounding tissue, and capillary permeability.

Nevertheless, as discussed in the literature [Barman

2010], a simple mass transfer model can be written

for the ISF volume (VISF).

The review by Ziegler (Zierler, 1999) describes

major factors that, singly or together, influence the

concentration and distribution of D-glucose in

humans, with emphasis on rest, physical activity, and

alimentation. It identifies areas of uncertainty:

distribution and concentrations of glucose in

interstitial fluid, kinetics and mechanism of

transcapillary glucose transport, kinetics and

mechanism of glucose transport via its transporters

into cells, detailed mechanisms by which hormones,

exercise, and hypoxia affect glucose movement across

cell membranes, whether translocation of glucose

transporters to the cell membrane accounts

completely, or even mainly, for insulin-stimulated

glucose uptake, whether exercise stimulates release of

a circulating insulinomimetic factor, and the relation

between muscle glucose uptake and muscle blood

flow. It was pointed out that there is no compartment

of glucose in the body at which all glucose has the

same concentration, and that models of glucose

metabolism, including effects of insulin on glucose

metabolism based on assumptions of concentration

homogeneity, cannot be entirely correct.

3 RESULTS AND DISCUSSION

The results obtained by applying a simple system

with costs effective components and using DMEM

solutions show that the measurements are

reproducible under the same conditions. Using

different light sources (LASER, LED, IR-emitter)

having wavelengths in the visible and IR and using

photodiodes and thermopile as detectors have shown

the same tendency by measurements. A few

wavelengths have demonstrated more dependency

on the glucose concentrations. As an example of the

measurements, figure 1 shows the high dependency

of the detected light from glucose concentrations in

glucose DMEM solutions by IR around the wave

number 3200 cm

-1

and between 1000 and 1700 cm

-1

Figure 1: IR-Spectroscopy by DMEM glucose solution.

The tissue light absorption in the IR-range is too

high due to the high water absorption, so that the

light penetration in tissue is too small. Light emitting

diodes and IR-detectors are also too expensive in

this range. The non-invasive glucose measurement

in this range is then too difficult and very expensive

for home monitoring. Using an IR-emitter as an

example of the results by measurements in DMEM

solution and a thermopile as detector of scattered IR

radiation between 9000 nm and 10000 nm have

shown good results. Instead of IR-Emitter and

thermopile, Laser sources and pyroelectric infrared

sensors can be utilized.

By Raman scattering without enhancement the

resolution of the measurements was not enough for

the glucose concentrations monitoring. Increasing

the measuring time will increase the resolution and

the detection threshold. Figure 2 shows that the

increase of the detection time increases the signal

quality and the resolution; where as the resulting

error will be reduced. The measuring time in Figure

2 by measurements an DMEM-solutions having

different glucose concentrations is five minutes. As

shown in the non-processed signal, the detection

OPTICAL SPECTROSCOPY AND OBSTACLES BY NON-INVASIVE DETECTION OF GLUCOSE

CONCENTRATION BY HOME MONITORING

293

threshold is ca. 30 mg/dl and the resolution seems to

be around 10 mg/dl. Increasing the measuring time

may be possible by in-vivo measurements. The

motion artifact will cause large perturbations to the

detected signals, which can be minimized by

applying an adaptive filter. The enhancement of the

Raman scattering by applying the previous

mentioned methods can produce results having a

high resolution and the detection time may be

reduced.

Figure 2: Raman scattering measured using a LASER

diode with wavelength 785 nm and 80 mW power. From

top: 300 s measuring time; a. 10mg/dl, b. 20 mg/dl, c. 30

mg/dl, d. 40 mg/dl, e. 50 mg/dl, f. reference solution with

a very high glucose concentration.

The realization of a compact simple device for

home monitoring using Raman spectroscopy seems

to be very difficult, but it may be possible using

miniaturized components and intelligent methods.

Figure 3: Scattered measured signal humidity for different

glucose concentrations al parameters.

The relation between the scattered measured

signal and humidity for different glucose

concentrations are shown in Figure 3. The results

obtained by using a fluorescence spectrometer show

the emitted light by stimulation of a DMEM solution

with different glucose concentrations. The detected

signal with 465 nm by the stimulation in the UV

light at the wavelength of 360 nm is not highly

correlated with the glucose concentration. But an

increasing tendency of the emitted light with the

increasing glucose concentration is registered.

The detected signal at 535 nm shows a high

correlation with glucose concentration when

stimulated with a light by the wavelength 485 nm. In

the contrary to the detected signals at 465 nm

mentioned above, a decreasing tendency of the

emitted light with the increasing glucose

concentration is shown. Despite of the variations of

the detected signals around the wavelengths 465 nm

when stimulated with the UV-light at the wavelength

360 nm, non-invasive measurement may deliver

very good results due to other fluorophores in the

skin.

A high correlation of the detected signal at

535 nm with known glucose concentration when

stimulated with 485 nm or 430 nm at different

glucose concentrations was obtained in DMEM-

solutions.

The detected glucose signals are too small and

should be processed carefully. Also the in vivo

measurements are subjected to more noise and

motion artifacts. An adaptive filtering will be needed

for eliminating these perturbations. A noise

reference signal is generated by means of a

Synthesizer or piezoelectric element and will be

adjusted as much as possible to the real noise

contained in the corresponding measurement by the

adaptive filter based on the least mean square

optimization algorithm. This algorithm has delivered

very good results by testing it for the non-invasive

calculations of oxygen saturation by artificial

vibrations of the hand, where a pulse oximeter

sensor is applied at the finger subjected to these

artifacts. We are applying a multisensor technology

that overcomes the obstacles others have faced

trying to measure blood glucose optically through

the skin.

4 CONCLUSIONS

AND FUTURE WORK

Optical methods are valuable and promising for the

non-invasive detection of blood glucose. Raman

scattering can have a very high sensitivity and

resolution, but the development of an in-vivo simple

device till now, seems to be very difficult. IR-

spectroscopy is promising for the development of a

cost effective sensor for home monitoring. Because

of the fact that the detected glucose signals are too

small and subjected to a lot of disturbances from the

surroundings and from the background of the

measured locations due to tissue alteration and

physiological parameter variations, a high signal to

noise ratio measuring system is essential for this

difficult task. Tiny perturbations such as

BIOSIGNALS 2012 - International Conference on Bio-inspired Systems and Signal Processing

294

temperature, humidity and applied pressure

variations can adulterate the measurements. Also

small drift of the characteristics of the electronic and

optical components can cause great disturbances to

the measurements reducing the accuracy or even

yielding invalid measurements. In addition to use a

robust hardware and apply advanced signal

processing methods by the glucose detection, all

these factors have to be taken into account.

The results obtained using a fluorescence

spectrometer having the stimulation/emission wave-

lengths of 360 nm /465 nm, 430 nm / 535 nm and

485 nm / 535 nm (Abdallah, 2011) as well as our

further fluorescence measurements at these and

further wavelengths have shown that fluorescence

spectroscopy is a very promising method. We have

already developed a compact sensible system and

sensor for that aim.

The integration of further parameters can

enhance the accuracy, but the system complexity, its

size and costs have to be minimized to enable the

applying of the device for home monitoring. Also

the number of the measured parameters has to be

minimized in order to reduce the resulting error

caused by the measurements variations..

Also by applying light sources with wavelengths

in IR over 1400 nm the penetration depth of light in

the tissue will be very small because water has a

high absorption of IR-light. This will be important

by detecting glucose or cholesterol using IR-

spectroscopy. The reflective sensors can be applied

proximal in order to avoid the perfusion problem by

applying sensors (in case of transmission) distal to

body extremity like fingers or earlobe. Light

reflected from the tissues and detected by

photodetectors and then the findings can be

interpreted by the software in the sensor. The

reflection sensor can be applied on forehead, back,

breast etc., and hence diagnose the central parts of

the body.

For the detection of glucose concentration

noninvasively using various optical methods, the

interaction between light and definite glucose

solutions was studied. IR-spectroscopy has the

potential for the development of a simple cost

effective sensor for glucose monitoring that can be

used for home care.

Problems with existing methods have

encouraged alternative approaches to glucose

sensing, and those based on multiparameter like

scattering, fluorescence intensity and lifetime have

special advantages, including sensitivity and the

potential for non-invasive measurement when UV,

visible or NIR light is used (Yamakoshi, 2006;

Evans, 2005; Evans, 2003; Pickupa, 2005). The

fluorescence signals using UV light as stimulus and

detection of fluorescence at violet or blue have

shown a very good correlation with the glucose

concentrations in DMEM solution. Light stimulation

with blue light and the detection of fluorescence by

green region shows also a high correlation with the

glucose concentrations. The detected glucose signals

will be subjected to perturbations from the

surroundings and from the background of the

measured locations due to tissue alteration and

physiological parameter variations. All perturbations

such as temperature, humidity and applied pressure

variations have to be considered by the calculations,

as illustrated by Figure 4. The drift of the

characteristics of the system components may cause

high disturbances to the measurements.

Figure 4: Schematic of a multisensor for non-invasive

detection of blood glucose, hemoglobin concentration, and

fractional oxygen saturation.

There is no doubt that the multiparametic

measurement depending on scattering, absorption

and fluorescence technologies have considerable

promise for glucose sensing.

As a future work, all developed sensors will be

integrated in one system that enables the

simultenous processing of the detected signals

(Caduff, 2009). Other blood components like total

hemoglobin concentrations and fractional oxygen

saturation measured non-invasively have to be taken

as parameters by the glucose calculations. The

suitable locations for measurements may be earlobe

for transmission measurements. For reflection

measurements forehead as well as abdomen or arm

can be chosen. Applying the Twersky theory or

diffusion theory by the calculations are our next

perspectives. After that a clinical study for non-

invasive measurements will be conducted. Applying

the neural fuzzy techniques, the results and the

system will be optimized to obtain the required

resolution and accuracy.

OPTICAL SPECTROSCOPY AND OBSTACLES BY NON-INVASIVE DETECTION OF GLUCOSE

CONCENTRATION BY HOME MONITORING

295

ACKNOWLEDGEMENTS

This work is a part of the project „System for Non-

invasive Detection of Glucose “supported by the

Foundation Baden-Württemberg Stiftung by

Research Program: Microsystem technology for the

life sciences.

REFERENCES

Abdallah O., Qananwah Q., Bolz, A. A., Hansmann J.,

Hinderer S., and Mertsching H., 2011. Fluorescence

Spectroscopy by Detection of Glucose Concentrations

in DMEM-Solutions and its Perspectives for Non-

invasive Measurement BIODEVICES 2011 part of

BIOSTEC, The International Joint Conference on

Biomedical Engineering Systems and Technologies,

Rome, Italy

Abdallah O., Hansmann J., Bolz A. and Mertsching H.,

2010. "IR spectroscopy vs. Raman scattering by

measurement of glucose concentration", Proc. SPIE

7376, 73760B; doi:10.1117/12.871469

Barman Ishan, Chae-Ryon Kong, Gajendra P. Singh,

Ramachandra R. Dasari, and Michael S. Feld, 2010:

Accurate Spectroscopic Calibration for Noninvasive

Glucose Monitoring by Modeling the Physiological

Glucose Dynamics. Analytical Chemistry 2010, Vol.

82, No. 14, pp. 6104–6114

Caduff, A., Talary, M., Mueller, M., Dewarrat, F., Klisic,

J., Donath, M., Heinemann, L., Stahel, W., 2009. Non-

invasive glucose monitoring in patients with Type 1

diabetes: A Multisensor system combining sensors for

dielectric and optical characterisation of skin. Bio-

sensors and Bioelectronics Vol. 24, pp. 2778– 2784.

D’Auria, S., Herman, P., Rossi, M., Lakowicz, J., 1999.

The Fluorescence Emission of the Apo-glucose

Oxidase from Aspergillusniger as Probe to Estimate

Glucose Concentrations. Biochemical and Biophysical

Research Communications, Vol. 263, pp. 550-553.

Evans, N., Gundi, L., Rolinski, O., Birch, D., Pickup, J.,

2003. Non-invasive glucose monitoring by NAD(P)H

autofluorescence spectroscopy in Fibroblasts and

adipocytes: A model for skin glucose sensing.

Diabetes Technology & Therapeutics, Vol. 5, No. 5.

Evans, N.D., Gnudi, L., Rolinski, O.J., Birch, D.J., Pickup,

J.C., 2005. Glucose-dependent changes in NAD(P)H-

related fluorescence lifetime of adipocytes and

fibroblasts in vitro: Potential for non-invasive glucose

sensing in diabetes mellitus. J. of Photochemistry and

Photobiology B: Biology, Vol. 80, pp. 122-129.

Gaster, Barak, Hirsch, MD; Irl B., MD, 1998. The Effects

of Improved Glycemic Controlon Complications in

Type 2 Diabetes, Arch Intern Med/VOL 158

Harman-Boehm Ilana, Avner Gal, Alexander M.

Raykhman, Jeffrey D. Zahn, Eugene Naidis, and

Yulia Mayzel, 2009. Noninvasive Glucose

Monitoring: A Novel Approach, Journal of Diabetes

Science and Technology Volume 3, Issue 2

Khalil, O.S., 2004. Non-invasive glucose measurement

technologies: An update from 1999 to the dawn of the

new millennium. Diabetes Technol. Ther., 6, 660-697

Lakovics, J., 2006. Principles of Fluorescence. Springer.

Maruo K., M. Tsurugi, J.Chin, T. Ota, H. Arimoto, Y.

Yamada, M. Tamura, M. Ishii, and Y. Ozaki, 2003.

"Noninvasive Blood Glucose Assay Using a Newly

Developed Near-Infrared System," IEEE Journal Of

Selected Topics In Quantum Electronics, 9 (2)

Mezzetti Andrea, Cipollone Francesco, Cuccurullo

Franco, 2000. Oxidative stress and cardiovascular

complications in diabetes: Isoprostanesas new markers

on an old paradigm, Cardiovasc. Research 47, 475–

488

Moschou, E., Sharma, P., Deo, S., Daunert, S., 2004.

Fluorescence Glucose Detection: Advances Toward

the Ideal In Vivo Biosensor. Journal of Fluorescence,

Vol. 14, No. 5.

Pambianco Georgia, Tina Costacou, Demetrius Ellis,

Dorothy J. Becker, Ronald Klein, andTrevor J.

Orchard, 2006. The 30-Year Natural History of Type 1

Diabetes Complications: The Pittsburgh Epidemiology

of Diabetes ComplicationsStudy Experience,

DIABETES, VOL. 55

Pickup, C. John, Faeiza Hussain, Nicholas D. Evans, Olaf

J. Rolinski, David J.S. Birch: Review: Fluorescence-

based glucose sensors, Biosensors and Bioelectronics

20, 2555–2565 (2005)

Smith John L., 2011, The Pursuit of Noninvasive Glucose:

“Hunting the Deceitful Turkey” Second Edition

Tura Andrea, Stefano Sbrignadello, Domenico

Cianciavicchia, Giovanni Pacini and Paolo Ravazzani,

2010. A Low Frequency Electromagnetic Sensor for

Indirect Measurement of Glucose Concentration: In

Vitro Experiments in Different Conductive Solutions

Sensors 10, pp. 5346-5358;

Wang Chuji and Sahay Peeyush, 2009, Breath Analysis

Using Laser Spectroscopic Techniques: Breath

Biomarkers, Spectral Fingerprints, and Detection

Limits, Sensors 2009, 9, PP. 8230-8262

Yamakoshi, K. Y. Yamakoshi: Pulse Glucometry, 2006. A

New Approach for Non-invasive Blood Glucose

Measurement Using Instantaneous Differential Near

Infrared Spectrophotometry, Journal of Biomedical

Optics 11(5), 1-11

Zierler, K., 1999. Whole body glucose metabolism. Am J

Physiol, 276(3 Pt 1):E409-E426

BIOSIGNALS 2012 - International Conference on Bio-inspired Systems and Signal Processing

296