Near-infrared quantitative analysis instruments (170) non-invasively measure blood glucose by analyzing near-infrared energy following interactance with venous or arterial blood, or transmission through a blood containing body part. A replaceable cartridge (195) stores data unique to an individual patient/user. The analysis instruments (170) are accurate and readily lend themselves to at-home testidiabetics. A method is also disclosed for easily and reliably custom calibrating near-infrared quantitative analysis instruments including obtaining a plurality of blood samples and blood glucose level measurements thereof at a predetermined time interval. These values are entered into the analysis instrument. Near-infrared optical measurements of the individual are simultaneously taken using the analysis instrument at another predetermined time interval. Calibration regression analysis is performed on the data to custom calibrate the analysis instrument which involves linearly interpolating the blood sample glucose measurements with the near-infrared optical recordings.

Non-invasive measurement of blood glucose

Disclosed is a packaging material for forming an armor body for a battery, the armor body being adapted for use in such a manner that a battery body is inserted into the armor body and the peripheral edge of the armor body is then heat sealed for hermetic sealing, the packaging material being a laminate comprising at least a substrate layer, an adhesive layer, a barrier layer, a dry laminate layer, and a sealant layer, characterized in that at least the sealant layer comprises a low-fluidity polypropylene layer having low susceptibility to collapse upon exposure to heat and pressure at the time of heat sealing and a high-fluidity polypropylene layer having high susceptibility to collapse upon exposure to heat and pressure at the time of heat sealing, and the innermost layer is the high-fluidity polypropylene layer. This battery packaging material can impart moisture barrier properties and resistance to contents of the battery.

Packaging material for battery

The present invention relates to spectral analysis systems and methods for determining physical and chemical properties of a sample by measuring the optical characteristics of light emitted from the sample. In one embodiment, a probe head (100) for use with a spectrometer includes a reflector (118) for illuminating a sample volume (150) disposed circumferentially about the light source (140) of the probe head. In another embodiment, a probe head includes an optical blocking element for forcing the optical path between the light source and an optical pick-up optically connected to the spectrometer into the sample. The probe head also includes a reference shutter for selectively blocking light emitted from the sample from reaching the optical pick-up facilitate calibration of the spectrometer.

Optical probes and methods for spectral analysis

A method of non-invasive determination of the concentration of at least one analyte in the blood of a mammal, includes the steps of projecting near-infrared radiation on a portion of the body of the mammal, the radiation including a plurality of wavelengths; sensing the resulting radiation emitted from the portion of the body; deriving from the sensed resulting radiation emitted from the portion of the body a first expression for the magnitude of the sensed radiation as a function of wavelength of the sensed radiation; pretreating the first expression to minimize the influence of offset and drift to obtain a second expression for the magnitude of the sensed radiation as a function of wavelength; and performing multivariate analysis of the second expression to obtain a value for the concentration of the analyte.

Non-invasive determination of analyte concentration in body of mammals

Percentage concentrations of constituents of a sample of cereal grain or other agricultural product in a flowing stream are determined while harvesting or processing using a short wave near infrared analyzer. The analyzer irradiates the sample, picks up diffuse reflectance of individual wavelengths from the sample and spatially separates the diffuse reflectance into a response at individual wavelengths. The result is to simultaneously detect the intensities of the individual wavelengths in parallel from the same portion of the product being analyzed. Percentage constituents of the composite substance may then be compared with known percentage constituents to determine the constituents. The analyzer is optically stabilized and is therefore suitable for attaching to the agricultural combine to perform real time measurements in the field.

Grain quality monitor

An improved grain quality analyzer, for analyzing the percentage concentration of various constituents, e.g. protein and water, in a grain sample, photo-optically measures the change in the optical density of the sample, Δ OD, in a range of characteristic wavelengths and, for protein, in a range of neutral wavelengths and uses these measured values to compute the percentage concentration of the constituents. A grain sample is irradiated with light, the wavelength of which sweeps across the infrared light-spectrum including a range of wavelengths termed characteristic wavelengths, and, for protein, a range of wavelengths termed neutral wavelengths. The characteristic wavelengths are those wavelengths at which the optical characteristics of the irradiated sample, that is, the optical density, reflectivity, transmissivity, and/or absorption, are known to vary as a function of the concentration of the measured constituent, and, for the protein measurement, the neutral wavelengths are those wavelengths at which the optical characteristics are substantially independent of the concentration of the protein. In the preferred embodiment, photo-optical sensors measure the light reflected from the sample and provide output signals indicative of the optical density of the sample as a function of the wavelength of the irradiating light. Control and computing means sample the sensor output at spaced apart points in the range of characteristic wavelengths and in the range of neutral wavelengths to provide signals representative of the optical characteristics of the sample. Computing circuitry computes the percentage concentration of water and, for protein, computes the percentage concentration using an algorithm in accordance with the present invention which eliminates the inaccuracies introduced by light scattering from the sample.

Grain quality analyzer

In an instrument adapted particularly for analyzing agricultural products, such as grain, interference filters are rotated successively through an infrared wide wavelength band beam of light, which irradiates a sample of the product. Each of the interference filters operates to transmit a narrow wavelength band of light, the center frequency of which is scanned through a range of values as the angle of incidence of the beam of light to the interference filter varies as the filter rotates through the beam of light. An optical track is provided which rotates with the assembly of interference filters to generate pulses from index markings in the optical track. The markings in the optical track are spaced so that one pulse is generated for each angstrom of variation of the center frequency transmitted through an interference filter. The pulses are counted in the counter and the counts in the counter are used to control the selection of intensities reflected from the sample to be used in the analysis of the sample.

Optical analyzing instrument with equal wavelength increment indexing

In a spectrophotometer, a light source is provided in the form of a multiplicity of light emitting diodes which transmit light through an entrance slit to irradiate an oscillating grating. The light is dispersed by the grating toward an exit slit which transmits a narrow bandwidth of light to irradiate a sample. As the grating oscillates, the wavelength transmitted through the exit slit is scanned through a selected spectrum. The diodes in the array each emit light in a different wavelength band so that the diodes as a group emit light throughout the selected spectrum. As the grating oscillates, the diodes are energized and extinguished in sequence so that no more than two diodes will be energized at any given instant of time and the energized diode will be emitting light at the wavelength dispersed by the grating to the exit slit.

Spectrophotometer with light source in the form of a light emitting diode array

In the last issue of NIR news the Analytical Mode was defined as the recommended calibration method for analysing future samples that are outside of the calibration set. Fundamental to this approach is explaining cause and effect, i.e. identifying a specific absorption or other spectral property that is caused by the analyte. Once cause and effect are fully explained and accounted for in the calibration equation, then there is a high probability that this equation can be used to analyse future samples accurately. Explaining cause and effect does not require that a calibration sample set contain 100% of the total spectral variation that will be encountered in the future. Therefore, using the Analytical Mode has the advantage that far fewer calibration samples are required when compared to strictly numerical calibration methods (Enumerative Mode). In the latter close to 100% of the spectral variation must be present in the calibration set in order to safely predict future samples outside the original population. Typical analytical calibrations require 10 to 50 calibration samples in order to derive a calibration that can predict useful information about a process. Typically, enumerative calibrations require hundreds of samples to gain a high confidence level in the calibrations' predictability. It is important that a decision be made as to which calibration method is going to be used before the test method is developed since each one requires very different strategies. Typically it is recommended that the analytical method be used whenever possible and that the enumerative method be used by exception, i.e. whenever cause and effect cannot be explained. Examples of the latter abound in agricultural or natural product analysis-such as starch damage, loaf volume, fibre analysis (ADF/NDF), taste panel results, amino acids, total energy etc. In these cases, it is not fully understood what the reference method is measuring so that a direct NIR spectral cause and effect determination cannot usually be accomplished with today's knowledge. On the other hand, the organic constituent make-up of products can usually be related to the behaviour of the NIR spectrum. Therefore measurements of constituents such as protein, fat or moisture in food products or OH value in polyols are examples where the analytical mode works best. It is our experience that the analytical mode is the proper calibration method for virtually all organic analyses (excluding trace analysis) of manufactured products. Since most manufacturing processes are not widely replicated, the analytical mode fits well with the limited availability of calibration samples. The strategy for developing an analytical calibration is divided into three phases; the feasibility study, the calibration phase and the prediction phase. Figure 1 shows a block diagram of the first phase-the feasibility study. Starting with the first step, the NIR instrument must be configured to produce precise spectra on the samples that are going to be encountered in routine analysis. The type of sample material will determine instrumental parameters such as measurement mode (transmittance or reflectance), sample measurement size, path length etc. A simple test to determine whether the instrument is producing precise spectra on these samples is to compare the spectra from multiple reloads of the same sample. The second step in the applications study is the development of a preliminary calibration. It is this step that determines cause and effect and establishes feasibility. Starting with a representative sample set, the following analytical method, in its simplest form, is recommended for quantitative calibrations. Note that spectral identification, cluster analysis, calculation of distances, etc. are outside the scope of this paper since they all serve a purpose other than quantification and require different calibration strategies. 1. Scan and store the spectra from four reloads of each sample. 2. Average the groups of four spectra to obtain an averaged spectrum for each sample. 3. Mathematically pre-treat the spectra to eliminate or reduce the spectral variation due to physical effects that are unrelated to the constituent make up of the samples. It is important that the pretreatment be derived theoretically: it should not be unique to the calibration sample set. Current recommended pretreatments are second derivative to eliminate or reduce varying baseline shifts, as well as a multiplicative scatter correction for optically diffuse samples. 4. Use multilinear regression (MLR) on the pretreated data to develop a preliminary prediction equation. The first term in the equation should explain 80% or more of the measurement variation and must be spectrally related to the constituent being analysed. The use of a second correction term should not be done on the basis of correlation alone but rather on establishing a cause and effect relationship. Additional terms can be used but again they must be selected an the basis of cause and effect. Other regression techniques such as PLS should be avoided except in some very specific situations. In general, these other techniques fall into the enumerative calibration mode which requires a different strategy. Use the prediction equation and each of the original unaveraged spectra to determine sampling precision. Since sampling precision is a function of the instrumental sample presentation method and the sensitivity of the prediction equation, iterating the calibration process and improving the instrumental

NIR Calibration. Applying the Analytical Mode—When Should it Be Used and How:

A previous issue of NIR news had an extremely good paper, authored by Ernie H. Baughman et al., about the problems of using a laboratory developed NIR spectroscopy calibration (feasibility study) in the real world environment. The problem describedusing a few samples in the laboratory to calibrate an NIR instrument and then finding that this calibration does not work in routine operation on future samples-has plagued this technology from its inception in the early 1970s. Since NIR can easily produce good results in the short term, many researchers have concentrated on reporting these results using ever more sophisticated mathematical procedures. Solving the long term problem is left to instrument producers and industry users. I will be the first to admit that instrument manufacturers have used short term results as a selling tool. The implication and belief that the instrument would produce good long term results is often misguided. Many of us realised years ago that NIR technology would not progress unless the technology better met primary user needs. Those needs involve an instrument that is easy to calibrate and which will produce valuable information in the future. This paper gives an overview of recommended procedures for developing an NIR calibration. Two more detailed papers will follow. Each will add depth to the topic. Since NIR calibration is inherently statistically based, these recommendations trace their origin to a paper by a pre-eminent statistician-the late W. Edwards Deming. 1 The paper is entitled "On Probability As a Basis for Action". In his paper, Dr Deming describes two types of statistical studies, namely Enumerative and Analytical. Dr Deming further states in the paper that a statistical study in the enumerative mode can only be used if 100% of the total population variation that will ever be encountered is contained in the frame studied. On the other hand, if this is not the case, then an Analytical study (which can be used in all cases) is necessary to gain information to be used in the future.

Donald R. Webster

Papers

Grain quality analyzer

Optical analyzing instrument with equal wavelength increment indexing

Spectrophotometer with light source in the form of a light emitting diode array

NIR Calibration. Applying the Analytical Mode—When Should it Be Used and How:

NIR Calibration: Going from a Feasibility Study to Gaining Information That Can Be Used to Improve a Process