Conference Paper

Single chip AWG-based Raman spectroscopy for continuous glucose monitoring

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

In this paper, we introduce an innovative proposal for a low cost micro Raman spectrometer for non-invasive glucose monitoring. In this design, an Array Waveguide Grating (AWG) chip integrated with PIN photodiodes and a prism coupling are proposed to replace fiber micro alignment in packaging and to reduce system cost. Experiments were designed and conducted to evaluate the impact of thickness of the waveguides and the incident angle of the laser beam on the efficiency of the prism coupling. Considerable light coupling was observed when the waveguide core is 2mm thick. An updated fabrication process was designed and test chip was also fabricated to determine the critical feature size to solve the problem of air bubble resulted in trench filling.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

Supplementary resource (1)

... Integrated photonic sensors based on arrayed waveguide gratings (AWGs) have gained significant attention in recent years because of their high performances, low footprints, low energy consumption, and low cost. Several AWG-based Raman spectrometers have been reported to determine early-stage tooth decay in humans [1] and monitor glucose non-invasively by integration with PIN photodiodes and a prism coupler [2]. Such integrated components are also applied in spectral domain optical coherence tomography (SD-OCT). ...
Article
Full-text available
We present the design and simulations of a novel integrated device concept enabling a frequency conversion of a broad signal. The solution is based on a hybrid silicon–graphene photonic chip, which could be used for controlled spectrometry in low-cost devices. The device is based on a silicon-on-insulator (SOI) platform on which an arrayed waveguide grating (AWG) is designed for operation at the center wavelength of λ = 1800 nm. The AWG is spectrally separating one broad input signal to thirty-two-output channels with a channel spacing of 2.72 nm. The output signals are well separated and uniform with the extinction ratio and the standard deviation of 10.00 dB and 0.04, respectively. The 3 dB channel width is 1.34 nm, which is suitable for sensing applications with significant accuracy. After spacial and spectral separation, each output signal is then converted to one signal at 1480 nm wavelength through a graphene-based saturable absorber scheme. Therefore, the device allows the detection of each separated signal with a simple near-infrared camera on which the outputs are imaged using conventional optics, leading to a classical pixel/wavelength correspondence. Crossed-waveguide couplers are designed to combine the controlling signal at 1480 nm to each channel waveguide of the AWG. The combination of the signals saturates the graphene layer at the output waveguides, allowing the pass of the controlling wavelength. This device can be applied as a spectrometer in environmental sensing and monitoring with high efficiency and low cost.
Article
Full-text available
An integrated arrayed-waveguide grating fabricated in silicon-oxynitride technology is applied to Raman spectroscopy. After its validation by reproducing the well-known spectrum of cyclohexane, polarized Raman spectra are measured of extracted human teeth containing localized initial carious lesions. Excellent agreement is obtained between the spectra of healthy and carious tooth enamel measured with our integrated device and spectra recorded using a conventional Raman spectrometer. Our results represent a step toward the realization of compact, hand-held, integrated spectrometers, e.g. for the detection of dental caries at an early stage.
Article
Full-text available
In this report, we detail our current work towards developing a surface-enhanced Raman spectroscopy (SERS) based sensor for in vivo glucose detection. Despite years of innovations in the development of blood glucose monitors, there remains a need for accurate continuous glucose sensors to provide care to rising numbers of diagnosed diabetes patients and mitigate secondary health complications associated with this metabolic disorder. SERS is a highly specific and sensitive optical technique suitable for direct detection of glucose. The SERS effect is highly distance dependent, thus the glucose molecules need to be within a few nanometers or adsorbed to an SERS-active surface. In our sensor, this is achieved with a self-assembled monolayer (SAM) that facilitates reversible interactions between glucose molecules and the surface. The amount of glucose near the surface is proportional to its concentration in the surrounding environment. We determined that the SAM-functionalized surface is stable for at least 10 days and provides rapid, reversible partitioning. In vitro experiments in bovine plasma as well as in vivo experiments in rats demonstrated quantitative detection. We show successful use of the SERS glucose sensor in rats, making it the first in vivo SERS sensor. Furthermore, we demonstrate free space transdermal detection of a SERS signal through the rat's skin as an initial step toward developing a transcutaneous sensor.
Article
Full-text available
To evaluate the precision and accuracy of a new advanced prototype of a noninvasive blood glucose monitor across a wide range of serum glucose concentrations. An advanced handheld noninvasive glucose monitor prototype was calibrated and tested using patients recruited by the General Research Center of the University of Connecticut Health Center. The monitor, developed by Infratec, uses principles of thermal emission spectroscopy. The noninvasive measurement of tympanic membrane glucose concentration was calibrated to the serum glucose concentration using 432 paired measurements from 20 subjects with insulin-requiring diabetes. This calibration was subsequently tested (results of power analyses) in a blind fashion with 126 paired measurements from six diabetic subjects who require insulin. In vivo measurements demonstrated the reproducibility of the methodology of the noninvasive glucose monitor. Based on the calibration model, predicted glucose concentrations for six subjects were as follows (for 126 data points): SD = 32 mg/dl, mean absolute relative error (%MARE) = 11.6, with a correlation coefficient of r = 0.87. Noninvasive glucose results were also compared with laboratory reference measurements using an error-in-variables method. Clark error grid analysis showed that 100% of the measurements fell within zones A and B (90% in zone A and 10% in zone B). The SD for all noninvasive measured concentrations was 27 mg/dl, %MARE was 8.6, and the correlation coefficient was r = 0.94. This first independent clinical study of an advanced noninvasive blood glucose prototype based on thermal emission in the mid-infrared spectral region has demonstrated glucose measurements with clinically acceptable accuracy but without the necessity of individual daily calibration.
Article
Full-text available
Tight glucose monitoring is essential for the reduction of diabetic complications. This research investigated the changes of absorption spectra observed in serum at three prominent glucose absorption peaks in the middle infrared using a demountable liquid, transmission cell. Two frequencies of light were used to determine the glucose absorption: one at 1193 cm(-1 ) to determine the background water absorption and the other at one of the characteristic peaks (1035, 1080, and 1109 cm(-1)). The peak at 1035 cm(-1) was best for quantitative determination with a standard of error of 20.6 mg/dl (1.1 mmol/L). While interference from other serum constituents could cause problems, urea and albumin-two constituents known to have close absorption peaks-were determined to have no effect on the ability to determine the glucose levels at 1035 cm( -1).
Article
Full-text available
We report the first successful study of the use of Raman spectroscopy for quantitative, noninvasive ("transcutaneous") measurement of blood analytes, using glucose as an example. As an initial evaluation of the ability of Raman spectroscopy to measure glucose transcutaneously, we studied 17 healthy human subjects whose blood glucose levels were elevated over a period of 2-3 h using a standard glucose tolerance test protocol. During the test, 461 Raman spectra were collected transcutaneously along with glucose reference values provided by standard capillary blood analysis. A partial least squares calibration was created from the data from each subject and validated using leave-one-out cross validation. The mean absolute errors for each subject were 7.8%+/-1.8% (mean+/-std) with R2 values of 0.83+/-0.10. We provide spectral evidence that the glucose spectrum is an important part of the calibrations by analysis of the calibration regression vectors.
Article
Full-text available
Tissue modulated Raman spectroscopy was used noninvasively to measure blood glucose concentration in people with type I and type II diabetes with HemoCue fingerstick measurements being used as reference. Including all of the 49 measurements, a Clarke error grid analysis of the noninvasive measurements showed that 72% were A range, i.e., clinically accurate, 20% were B range, i.e., clinically benign, with the remaining 8% of measurements being essentially erroneous, i.e., C, D, or E range. Rejection of 11 outliers gave a correlation coefficient of 0.80, a standard deviation of 22 mg/dL with p<0.0001 for N=38 and places all but one of the measurements in the A and B ranges. The distribution of deviations of the noninvasive glucose measurements from the fingerstick glucose measurements is consistent with the suggestion that there are at least two systematic components in addition to the random noise associated with shot noise, charge coupled device spiking, and human factors. One component is consistent with the known variation of fingerstick glucose concentration measurements from laboratory reference measurements made using plasma or whole blood. A weak but significant correlation between the deviations of noninvasive measurements from fingerstick glucose measurements and the test subject's hemoglobin concentration was also observed.
Article
For non-invasive in vivo glucose determinations by means of near-infrared spectroscopy, the anterior chamber of the human eye is a promising site. An optical set-up for the non-invasive glucose determination in the human eye precisely in the anterior chamber with a beam reflected from the surface of the eye lens is presented here. As the anterior chamber has a depth of 3.13±0.50mm, the beam follows an optical path of 5.3–7.3mm depending on the angle of incidence, which is individually constant. We will show that it is possible to acquire good concentration predictions for physiological glucose concentrations with such a long optical path. A chemometric study of NIR glucose spectra with concentrations of glucose in water of 10–350mg/dL (0.56–1.94mmol/L) resulted in a calibration model which was able to predict physiological glucose concentrations with a root mean square error of prediction RMSEPTest=15.41mg/dL. The Clarke error grid diagram shows that the model performs well according to medical impact. Using a first in vivo set-up, the precision is not sufficient for a reliable prediction of glucose concentration, especially due to the flickering of the patient's eye and the low reflectivity of the eye lens. Therefore, we have designed a new in vivo set-up: a prototype for a self-monitoring device with controlled geometry and laser radiation at several distinct wavelengths instead of the halogen lamp as light source. This allows a far higher signal/noise ratio under much better reproducible geometrical conditions and at the same time a much smaller necessary light flux.
Article
The frequent monitoring of glucose is an essential part of diabetes management. Despite the fact that almost all the commercially successful blood glucose monitoring devices are invasive, there is an immense need to develop non-invasive glucose monitoring (NGM) devices that will alleviate the pain and suffering of diabetics associated with the frequent pricking of skin for taking the blood sample for glucose testing. There have been numerous developments in the field of NGM during the last decade, which stress the need for a critical review. This manuscript aims to review the various NGM techniques and devices. The challenges and future trends in NGM are also discussed.
Article
The theory of Mixtures of Experts (MOE) was applied to the signal from a noninvasive glucose monitor for the purpose of converting raw signal data into blood glucose values. The MOE algorithm can be described as a generalized predictive method of data analysis. This method uses a superposition of multiple linear regressions, along with a switching algorithm, to predict outcomes. Any number of input/output variables are possible. The unknown coefficients in this method are determined by an optimization technique called the Expectation Maximization (EM) algorithm. The noninvasive GlucoWatch biographer operation has been described. Briefly, a small electrical current results in the transport of glucose beneath the skin to a hydrogel placed on the skin surface. Within the hydrogel, the glucose reacts with the enzyme glucose-oxidase to produce hydrogen peroxide. This hydrogen peroxide then diffuses to a platinum-based electrode, where it reacts to produce a current. The integral of this current (charge) over the sensing time is the signal used to measure extracted glucose. This process is repeated, yielding up to three measurements per hour. The data used for this analysis were obtained from diabetic subjects wearing the biographer over a 15-h period. The MOE inputs consisted of elapsed time, integrated current, blood glucose value at the calibration point, and a calibrated signal. The output was the value of blood glucose at each measurement. These training data were used to determine the unknown parameters in the MOE by the EM algorithm. Using a 3-h time point for calibrating the biographer, the mean absolute error (MAE) between the actual blood glucose and the blood glucose predicted with the MOE, was 14.4%.
Article
To estimate the global health expenditure on diabetes among people aged 20-79 years for the years 2010 and 2030. Country-by-country expenditures for 193 countries, expressed in United States Dollars (USD) and in International Dollars (ID), were estimated based on the country's age-sex specific diabetes prevalence and population estimates, per capita health expenditures, and health expenditure ratios per person with and without diabetes. Diabetes prevalence was estimated from studies in 91 countries. Population estimates and health expenditures were from the United Nations and the World Health Organization. The health expenditure ratios were estimated based on utilization and cost data of a large health plan in the U.S. Diabetes expenditures for the year 2030 were projected by considering future changes in demographics and urbanization. The global health expenditure on diabetes is expected to total at least USD 376 billion or ID 418 billion in 2010 and USD 490 billion or ID 561 billion in 2030. Globally, 12% of the health expenditures and USD 1330 (ID 1478) per person are anticipated to be spent on diabetes in 2010. The expenditure varies by region, age group, gender, and country's income level. Diabetes imposes an increasing economic burden on national health care systems worldwide. More prevention efforts are needed to reduce this burden. Meanwhile, the very low expenditures per capita in poor countries indicate that more resources are required to provide basic diabetes care in such settings.
Article
Accurate monitoring of the blood glucose level in diabetics is essential in preventing complications. Generally, conventional over-the-counter glucose meters require frequent painful finger punctures to obtain samples, which makes a noninvasive method preferable. The purpose of this study was to demonstrate that glucose levels can be measured transdermally with the combination of the low-profile cymbal array and an electrochemical glucose sensor consisting of amperometric electrodes and a novel glucose oxidase hydrogel. Interstitial fluid glucose concentrations can be determined with the electrochemical glucose sensor after the skin is made permeable to glucose by ultrasound (US) (20 kHz) with the thin (< 7 mm) and light (< 22 g) cymbal array. Using this array to deliver insulin into hyperglycemic rats, our previous experiments demonstrated that blood glucose levels would decrease 233.3 mg/dl with 5 min of US exposure. With the sensor and array, our goal was to determine the glucose levels of hyperglycemic rats noninvasively and evaluate the possible bioeffects. A total of 12 anesthetized rats were placed into two groups (US exposure group and control group) and the array (I(SPTP) = 100 mW/cm(2)) with a saline reservoir operating for 20 min was affixed to the abdomen. The array was removed and an electrochemical glucose sensor was placed on the exposed area to determine the glucose concentrations through the skin. Comparison was made using a commercial glucose meter with the blood collected from a jugular vein. The average blood glucose level determined by the sensor was 356.0 +/- 116.6 mg/dl, and the glucose level measured by the commercial glucose meter was 424.8 +/- 59.1 mg/dl. These results supported the use of this novel system consisting of the electrochemical glucose sensor and the cymbal array for glucose monitoring.
Article
We demonstrate the use of Raman spectroscopy to measure the concentration of many important constituents (analytes) in serum and whole blood samples at physiological concentration in vitro across a multipatient data set. A near-infrared (830-nm) diode laser generates Raman spectra that contain superpositions of Raman signals from different analytes. Calibrations for glucose, cholesterol, urea, and other analytes are developed by use of partial least-squares cross validation. We predict six analytes in serum with significant accuracy in a 66-patient data set, using 60-s spectra. The calibrations are shown to be fairly robust against system drift over the span of seven weeks. In whole blood, a preliminary analysis yields accurate predictions of some of the same analytes and also hematocrit. The results hold promise for potential medical applications.