Non-invasive blood glucose monitoring using NIR light has been suffered from the variety of optical background that is
mainly caused by the change of human body, such as the change of temperature, water concentration, and so on. In
order to eliminate these internal influence and external interference a so called floating-reference method has been
proposed to provide an internal reference. From the analysis of the diffuse reflectance spectrum, a position has been
found where diffuse reflection of light is not sensitive to the glucose concentrations. Our previous work has proved the
existence of reference position using diffusion equation. However, since glucose monitoring generally use the NIR light
in region of 1000-2000nm, diffusion equation is not valid because of the high absorption coefficient and small source-detector
separations. In this paper, steady-state high-order approximate model is used to further investigate the existence
of the floating reference position in semi-infinite medium. Based on the analysis of different optical parameters on the
impact of spatially resolved reflectance of light, we find that the existence of the floating-reference position is the result
of the interaction of optical parameters. Comparing to the results of Monte Carlo simulation, the applicable region of
diffusion approximation and higher-order approximation for the calculation of floating-reference position is discussed at
the wavelength of 1000nm-1800nm, using the intralipid solution of different concentrations. The results indicate that
when the reduced albedo is greater than 0.93, diffusion approximation results are more close to simulation results,
otherwise the high order approximation is more applicable.
The technologies of high sensitivity optical spectroscopy analysis on turbid media play an important part in scientific
research and biomedical applications. The optical path in which photons travel inside the turbid media generally brings
information of the components of the media. This paper introduces a novel method to study some of the properties of
turbid media by measuring and analyzing the differences of optic paths of wavelength modulated laser beams
experienced in the media. The operating principle to accomplish detecting media information in specified optical length
is theoretical analyzed. The Experiments and measurements on the multiple scattering properties in transparent media
(water, and air) and turbid media (simulation tissue fluid) are reported in the paper as well.
A reference position where the diffuse reflectance light intensity is insensitive to the variation of glucose concentration
exists in the radial detection space for glucose measurement in the scattering medium such as skin. The signal measured
in this position could be used as an inside reference to evaluate the influence on spectrum caused by other interferential
factors. The relationship between the position of radial reference point and the skin tissue property is studied in this paper.
Three-layer skin models with different optical parameters are designed to get sample sets at 1200~1700nm. In these
sets, μa, μs and g of dermis varies respectively, so does the depth of epidermis or dermis. The distribution rule of
dispersion of diffuse reflectance light intensity in the radial space is confirmed with the glucose concentration changes.
And the distribution property of the radial reference position in every sample set is obtained through Monte Carlo
simulation. The result shows that the distance of radial reference position from light source is insensitive to the variation
of absorption coefficient or the depth of dermis, but an increased scattering coefficient will shorten the distance; an
increased anisotropy coefficient or depth of epidermis will lengthen it. On the basis of that, the optical probes with
different structures are designed according to the skin tissue properties. So they could be used for the measurement of
corresponding patients, which enhances the practicability of floating reference method greatly.
KEYWORDS: Glucose, Temperature metrology, Diffuse reflectance spectroscopy, Monte Carlo methods, Absorbance, Blood, Tissue optics, Near infrared spectroscopy, Spectroscopy, Transmittance
The use of near-infrared spectroscopy for the monitoring of blood glucose concentration is limited by many ambiguous factors, which leads to the prediction precision is not satisfied. Due to the weak interested signal and the difficulty to quantify the physiological noise directly, the absorbance induced by glucose concentration and temperature was analyzed based on Beer-Lambert Law and displacement between glucose and water. Then the transmittance of glucose aqueous solution in different temperatures was measured by spectrometer to investigate the influence of glucose concentration and temperature. As it's difficult to distinguish the influence of temperature from the diffuse reflectance, the Monte Carlo simulation was used to compute the light intensity induced by the change in glucose concentration and physiological temperature. Finally, the influence of actual physiological temperature on the prediction model of glucose concentration was estimated based on the oral glucose tolerance tests of two diabetics. The result showed that, near the normal physiological temperature, the intensity of diffuse reflectance caused by -0.1 °C change in temperature was equivalent to that caused by 2.7 mmol/L change in glucose concentration. Moreover, the proportion of prediction error induced by temperature to the total error was more than 50%.
In the noninvasive blood glucose sensing using near-infrared spectroscopy, the physiological noise was one of the
biggest challenges. In order to efficiently reduce the influence of the physiological background variations on the diffuse
reflectance spectra, the floating-reference method was used by differentially processing two signals from reference point
and measuring point. In this paper, the wavelength-dependent characteristic of the floating-reference point was discussed
by simulation and primary experiment. First, the wavelength-dependent characteristic of intralipid- 5% solution was
investigated in the wavelength range of 1300-1600nm. And source-detector distance for reference point in the
wavelength of 1300nm was conducted by different concentration of scatter media including 2%, 5% and 10% intralipid
solution. Then the single-layer and three layers skin model were built to investigate the wavelength characteristic of
reference point. The water displacement coefficients and relative large change in glucose concentration were considered
in the simulation. Finally, the primary experiment of intralipid model was conducted to validate the wavelength
dependence of reference point. The result showed that, the floating reference will not exist in the strong absorption
region (near 1450nm) and the region where the change of absorption coefficient is positive (high than 1525nm) due to
the corporate influence of scattering and absorption coefficient. And the wavelength-dependent characteristic is
consistent for intralipid solution and the skin model.
Noninvasive and minimally invasive blood glucose sensing is one of most interesting research fields. For the noninvasive measurement using near-infrared spectroscopy, the optical signal was impaired by the uncertain physiological noise and systematical drift. A floating reference method by differentially processing two signals from reference point and measuring point was used to deal with these uncertain noises. For the minimally invasive measurement, interstitial fluid extracted by ultrasound and vacuum is investigated. Low-frequency ultrasound was applied to enhance the skin permeability to interstitial fluid by disrupting the stratum corneum lipid bilayers. In this paper, a kind of protein absorbing the glucose specifically called D-galactose/D-glucose Binding Protein (GGBP) was introduced to construct a novel surface plasmon resonance (SPR) measuring system. By immobilizing GGBP onto the surface of the SPR sensor, a new detecting system for glucose testing in mixed solution was developed. The experimental result indicated that, the SPR system succeeded in distinguishing glucose resolution of 0. 1 mg/L, and had linear relationship between 0.5 mg/L and 5 mg/L.
KEYWORDS: Glucose, Blood, Diffuse reflectance spectroscopy, Monte Carlo methods, Skin, Data processing, Tissue optics, Light, Scattering, Signal processing
A new method, floating-reference, can efficiently reduce the influence of the physiological background variations on the
blood glucose measurement by differentially processing two signals from reference point and measuring point.
In this paper, the reference point, where the diffuse reflected light is not sensitive to the variation of glucose
concentration, is theoretically proved. Then Monte Carlo simulations are applied to study the radial distribution of the
diffuse reflectance at different glucose concentrations for the skin. Moreover, the experiments are constructed to measure
the radial distribution of the diffuse reflectance by the intralipid solutions with different glucose concentrations. Both the
results from simulation and experiment validate the existence of floating-reference point.
By theoretically analyzing the background noises and their disturbing mode on the blood glucose detection, a novel data
processing method based on the reference point is proposed to effectively extract the blood glucose information. And it is
found from our preliminary experiments with intralipid samples that, this data processing method can reduce the
influence of background variation on the extraction of real glucose signal and thus enhance the resolving capability on
glucose concentration.
In the area of noninvasive human blood glucose concentration detecting, it has always been a critical task to extract the
glucose-specific signal from the highly overlapped and disturbed near-infrared spectrum. In this paper, the methodology
of effective glucose-specific signal extraction in complicated non-scattering sample is studied. By analyzing the impact
of water displacement upon dissolution of glucose, the relationship between glucose concentration and absorption
coefficient of the sample is deduced. Then, the reference wavelength where the absorption coefficient is insensitive to the
changes of glucose concentration is put forward theoretically. Accordingly, the validating experiments in aqueous
glucose solutions are executed. Both the theoretical and laboratorial results show that the reference wavelength of
glucose appears at 1525nm. Based on the reference wavelength, an effective method for extracting the glucose-specific
signal in complicated non-scattering samples is proposed and the corresponding validating experiments are constructed
with different glucose and albumin concentration. Two different methods, traditional and the novel reference wavelength
method are used to extract glucose signal and the corresponding root mean square error of prediction are 19.86mg/dl and
9.87mg/dl respectively. The experiment results indicate that the reference wavelength method can effectively eliminate
the influence of various noises on the glucose-specific signal extraction, and thus can remarkably improve the measuring
precision in noninvasive near-infrared glucose detecting.
In this paper, to find the quantitative errors of aqueous glucose induced by the temperature change at every wave point
ranging from 1200 to 1700 nm, the calibration curve is calculated and shown. During the measurement the temperature
varies from 30°C to 40°C, at a 2°C interval, and aqueous glucose concentration ranges from 100 mg/dL to 500 mg/dL, at
a interval of 100 mg/dL. The absorption of aqueous glucose decreases with the increasing of temperature, also the
absorbance decreases. In addition, only 1°C change in the temperature induces about -7xlO-3and -4x10-3 errors in the
absorbance of the aqueous glucose at the wavelength of 1550 and 1610 nm respectively. To decrease or even eliminate
the error caused by the temperature, two methods are put up in this paper.
We have performed transcutaneous measurement and vessel bypass measurement to obtain the skin spectra and the blood vessel spectra respectively over the 1100-1700nm in the animal trial. The aim of this study is to validate the feasibility of the near-infrared (NIR) spectroscopy as a non-invasive blood glucose monitoring method, in particular during clinically relevant fluctuation in blood glucose. Two steps are adopted to evaluate the correlation between the skin diffusion spectra and the blood vessel transmission spectra. First, the variation tendencies of the skin and the blood vessel spectra were evaluated, and the partial least square (PLS) regression was adopted to establish the calibration model between the skin spectra, the vessel spectra and the corresponding concentration respectively. Then, the correlation analysis method is used to describe the relationship between the two kinds of spectra mentioned above. The correlation between the skin and the vessel spectra will be a powerful proof to demonstrate the correlation between the skin spectra and the blood glucose concentration.
KEYWORDS: Glucose, Photons, Absorption, Scattering, Diffuse reflectance spectroscopy, Blood, Monte Carlo methods, Tissue optics, Near infrared spectroscopy, Near infrared
Non-invasive glucose monitoring with optical methods has obtained increasing interest, in that the methods have shown great benefit for diabetes. Nevertheless, low sensitivity and signal to noise ratio (ratio of effective photons) are two major difficulties in non-invasively NIR spectral monitoring of blood glucose concentration. Designing the optical probe properly is one of the effective ways to improve measuring sensitivity and ratio of effective photons. In this paper, definition about ratio of effective photons in measurement of glucose is introduced. And then effect of glucose on optical properties of human skin is analyzed, based on this, two kind of sensitivities for diffuse reflectance, namely sensitivity to absorption and that to scattering, is derived. To investigate the ratio of effective photons and sensitivities, Monte Carlo simulations have been performed on a three-layered media with optical parameters similar to those of human skin. The results have shown that (1) high ratio of effective photons, even as high as 60%, can be obtained by choosing proper the separation between source and detector; (2) sensitivity of diffuse reflectance to absorption and scattering has different dependence on source-detector separation, which enables one can have different options by making use of different effect from glucose level changing. In the end, some suggestions have been put forward to improve precision of measurement of blood glucose.
A non-invasive and continuous blood glucose monitoring would be of great advantage for diabetic patients. Many techniques have been proposed for the purpose. But so far, none of these methods has been proven to be reliable and precise enough for in vivo monitoring. In non-invasive glucose measurement using near-infrared (NIR) spectroscopy, the difficulty is that the spectral variations due to the glucose concentration are extremely small compared with other sources of variations. Therefore extracting the variation signal of glucose in complicated background is challenging. We investigated the relationship between sample complexity and prediction accuracy, which was the fundamental research of non-invasive sensing and a kind of method to determine whether the OGTT or non-invasive sensing can achieve the required accuracy of clinic. A series of in vitro experiments had been conducted with different complex samples and same measurement system to analyze the relation between the sample complexity and the prediction accuracy, and some conclusions had been drawn. In general, the increase of sample complexity doesn’t lead to the distinct increase of prediction error.
Non-invasive glucose measurement by near-infrared spectroscopy is mainly based on the absorption of glucose. However, for non-invasive blood glucose measurement, the diffuse reflectance spectra are influenced not only by the absorption coefficient, but also by the scattering coefficient, anisotropy factor, and refractive index, which are normally nonlinear with the glucose concentrations. Furthermore, the variations of spectra depend on the relative changing direction of the absorption coefficient and scattering coefficient. In this paper, using the simulated samples of human tissues with different glucose concentrations in different conditions, we discussed the rules of how the glucose concentrations affected the absorption coefficient and scattering coefficient, respectively. The relations between the diffuse reflectance spectra and the absorption coefficient, as well as and the scattering coefficient were also investigated. Thus, we confirmed which of the optical parameters and measurement conditions would affect the diffuse reflectance spectra significantly. Based on the above results, proper methods could be selected to measure blood glucose concentration non-invasively according to different conditions, then the information of glucose absorption would be extracted more effectively, and higher measurement precision would be expected.
KEYWORDS: Glucose, Signal analyzers, Blood, Absorption, In vivo imaging, Absorbance, Diffuse reflectance spectroscopy, Calibration, In vitro testing, Near infrared
Near infrared spectroscopy has been proposed as an effective way for measuring blood glucose non-invasively. However the change of spectrum due to an increase in glucose level is very small compared to the changes due to other variations such as absorption of major blood components, skin surface reflectance, temperature and pressure and so on. So the complexity of spectrum makes it difficult to identify unique glucose information. In this paper, the effect of background correction is discussed firstly. Then a simple substitution is proposed to compute the net analyte signal of glucose using the subspace spanned by the background spectra. For the in vitro experiment, the net analyte signals of glucose using the traditional methods and the subspace spanned by background have the same peaks in the absorption peaks of glucose for the glucose aqueous solution. For in vivo experiment, there is significant spectral difference between the subject who took OGTT test and the subject who took no glucose or water. And the net analyte signal of glucose is computed for OGTT test based on the subspace spanned by the spectra of subject who didn’t take glucose. Results show that, the spectral information induced by glucose taking is quite significant but it does not have the same peak at the absorption peak of glucose in near-infrared region.
Concept of penetration depth of diffused photons migrating in turbid medium is introduced, and subsequently, distribution of penetration depth and mean penetration depth in three-layered media are investigated by using of Monte Carlo simulation technique. An optimal source-detector separation is derived from the mean penetration depth referring to monitoring the change of chromophore concentration of the sandwiched layer. In order to verify the separation, we perform Monte Carlo simulations 80 times with varied absorption coefficient of the sandwiched layer. All these diffuse reflectance of 80 times of Monte Carlo simulations are used to construct a calibration model with the method of PLS. High correlation coefficients and low RMSEP at the optimal separation have conformed correctness of the selection.
KEYWORDS: Skin, Reflectivity, Tissues, In vivo imaging, In vitro testing, Near infrared, Blood, Glucose, Reflectance spectroscopy, Near infrared spectroscopy
In non-invasive blood sensing with near-infrared (NIR) reflectance spectroscopy, optical probe usually directly contacts skin to eliminate specular reflection. Due to the direct contact, changes in contact pressure can lead to changes in internal structure and components distribution of the measured site, and thus introduces great interference into the final results. In this paper, we use self-made AOTF spectrophotometer to investigate the changes of reflectance spectrum with changing contact status for tissues in vitro (fresh porcine skin) and in vivo (two volunteers' left palms) at wavelengths ranging from 1100 nm to 1700 nm. The results show that with increasing degree of contact, energy of reflectance spectrum gradually decreases and the trend goes stable with time. However, the decreasing degree is related to wavelengths, which potentially suggests an indirect relevance with changes of components in tissues. Meanwhile, the results provide a practical solution to determining the optimum contact status between probe and skin.
KEYWORDS: Glucose, Absorption, Near infrared, Near infrared spectroscopy, Photons, Spectroscopy, Monte Carlo methods, Blood, Tissues, Mathematical modeling
Utilizing Near-infrared Spectroscopy for non-invasive glucose concentration sensing has been a focusing topic in biomedical optics applications. In this paper study on measuring conditions of spectroscopy on human body is carried out and a series of experiments on glucose concentration sensing are conducted. First, Monte Carlo method is applied to simulate and calculate photons’ penetration depth within skin tissues at 1600 nm. The simulation results indicate that applying our designed optical probe, the detected photons can penetrate epidermis of the palm and meet the glucose sensing requirements within the dermis. Second, we analyze the influence of the measured position variations and the contact pressure between the optical fiber probe and the measured position on the measured spectrum during spectroscopic measurement of a human body. And, a measurement conditions reproduction system is introduced to enhance the measurement repeatability. Furthermore, through a series of transmittance experiments on glucose aqueous solutions sensing from simple to complex we found that though some absorption variation information of glucose can be obtained from measurements using NIR spectroscopy, while under the same measuring conditions and with the same modeling method, choices toward measured components reduce when complication degree of components increases, and this causes a decreased prediction accuracy. Finally, OGTT experiments were performed, and a PLS (Partial Least Square) mathematical model for a single experiment was built. We can easily get a prediction expressed as RMSEP (Root Mean Square Error of Prediction) with a value of 0.5-0.8mmol/dl. But the model’s extended application and reliability need more investigation.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.