Noninvasive Monitoring of Glucose Using Near-Infrared Reflection Spectroscopy of Skin-Constraints and Effective Novel Strategy in Multivariate Calibration.

For many years, successful noninvasive blood glucose monitoring assays have been announced, among which near-infrared (NIR) spectroscopy of skin is a promising analytical method. Owing to the tiny absorption bands of the glucose buried among a dominating variable spectral background, multivariate calibration is required to achieve applicability for blood glucose self-monitoring. The most useful spectral range with important analyte fingerprint signatures is the NIR spectral interval containing combination and overtone vibration band regions. A strategy called science-based calibration (SBC) has been developed that relies on a priori information of the glucose signal ("response spectrum") and the spectral noise, i.e., estimates of the variance of a sample population with negligible glucose dynamics. For the SBC method using transcutaneous reflection skin spectra, the response spectrum requires scaling due to the wavelength-dependent photon penetration depth, as obtained by Monte Carlo simulations of photon migration based on estimates of optical tissue constants. Results for tissue glucose concentrations are presented using lip NIR-spectra of a type-1 diabetic subject recorded under modified oral glucose tolerance test (OGTT) conditions. The results from the SBC method are extremely promising, as statistical calibrations show limitations under the conditions of ill-posed equation systems as experienced for tissue measurements. The temporal profile differences between the glucose concentration in blood and skin tissue were discussed in detail but needed to be further evaluated.

FT-IR-cPAS--new photoacoustic measurement technique for analysis of hot gases: a case study on VOCs.

This article describes a new photoacoustic FT-IR system capable of operating at elevated temperatures. The key hardware component is an optical-readout cantilever microphone that can work up to 200 °C. All parts in contact with the sample gas were put into a heated oven, incl. the photoacoustic cell. The sensitivity of the built photoacoustic system was tested by measuring 18 different VOCs. At 100 ppm gas concentration, the univariate signal to noise ratios (1σ, measurement time 25.5 min, at highest peak, optical resolution 8 cm(-1)) of the spectra varied from minimally 19 for o-xylene up to 329 for butyl acetate. The sensitivity can be improved by multivariate analyses over broad wavelength ranges, which effectively co-adds the univariate sensitivities achievable at individual wavelengths. The multivariate limit of detection (3σ, 8.5 min, full useful wavelength range), i.e., the best possible inverse analytical sensitivity achievable at optimum calibration, was calculated using the SBC method and varied from 2.60 ppm for dichloromethane to 0.33 ppm for butyl acetate. Depending on the shape of the spectra, which often only contain a few sharp peaks, the multivariate analysis improved the analytical sensitivity by 2.2 to 9.2 times compared to the univariate case. Selectivity and multi component ability were tested by a SBC calibration including 5 VOCs and water. The average cross selectivities turned out to be less than 2% and the resulting inverse analytical sensitivities of the 5 interfering VOCs was increased by maximum factor of 2.2 compared to the single component sensitivities. Water subtraction using SBC gave the true analyte concentration with a variation coefficient of 3%, although the sample spectra (methyl ethyl ketone, 200 ppm) contained water from 1,400 to 100k ppm and for subtraction only one water spectra (10k ppm) was used. The developed device shows significant improvement to the current state-of-the-art measurement methods used in industrial VOC measurements.

On Wiener filtering and the physics behind statistical modeling.

The closed-form solution of the so-called statistical multivariate calibration model is given in terms of the pure component spectral signal, the spectral noise, and the signal and noise of the reference method. The "statistical" calibration model is shown to be as much grounded on the physics of the pure component spectra as any of the "physical" models. There are no fundamental differences between the two approaches since both are merely different attempts to realize the same basic idea, viz., the spectrometric Wiener filter. The concept of the application-specific signal-to-noise ratio (SNR) is introduced, which is a combination of the two SNRs from the reference and the spectral data. Both are defined and the central importance of the latter for the assessment and development of spectroscopic instruments and methods is explained. Other statistics like the correlation coefficient, prediction error, slope deficiency, etc., are functions of the SNR. Spurious correlations and other practically important issues are discussed in quantitative terms. Most important, it is shown how to use a priori information about the pure component spectra and the spectral noise in an optimal way, thereby making the distinction between statistical and physical calibrations obsolete and combining the best of both worlds. Companies and research groups can use this article to realize significant savings in cost and time for development efforts.