Non-destructive measurement device for light scattering materials

Abstract

A non-destructive measuring device for a light scattering medium is provided that can obtain good measurement accuracy even if the positional relationship between the light receiving opening of a light receiving unit and the object changes over time due to the movement of the object. [Solution] A nondestructive measuring device 1 for light scattering bodies includes a light source unit 8 that generates light of multiple wavelengths, an optical fiber cable 4 that has a light exit port and irradiates the light of multiple wavelengths, two optical fiber cables 5 and 6 that are located on the same line as the center of the light exit port and have light receiving ports at different distances from the light exit port and simultaneously receive reflected light of the irradiated light, two light detecting units 50 that simultaneously detect the light intensity of the light received by each light receiving unit, a signal processing unit 40 that extracts light intensity for each wavelength from each of the two light intensities, and an arithmetic processing unit that calculates reflectance, which is the ratio of light intensity of reflected light at the same wavelength, for each wavelength and calculates a property characteristic value of the inside of the subject based on the reflectance. The light of multiple wavelengths includes light of three or more wavelengths.

Description

[Technical Field]

[0001] The present invention relates to a non-destructive measuring device for a light scattering medium, which measures the property characteristic values ​​of a light scattering medium as a specimen. [Background technology]

[0002] BACKGROUND ART Conventionally, various non-destructive measuring devices for measuring light scattering media are known, which optically measure characteristic values ​​relating to the internal properties of a test object made of a light scattering media, such as for measuring the sugar content of fruits and vegetables or measuring the components of human blood. For example, the device described in Patent Document 1 sequentially illuminates a test subject with light of multiple wavelengths and irradiates the test subject from a light exit port, sequentially detects the total amount of reflected light at multiple wavelengths at two or more positions on two concentric circles with different radii centered on the light exit port, and sequentially calculates the ratio of the total amount of reflected light for each wavelength at positions separated by the difference in radius of the concentric circles as reflectance for each wavelength. Furthermore, by using a calibration curve expressed as a first-order polynomial of multiple relative absorbance ratios, the effects of scattering by the specimen and absorption by components other than the target component (impurities) can be reduced, allowing the target component to be measured with good accuracy.

[0003] Furthermore, Non-Patent Document 1 reports highly accurate measurement of biological composition such as total hemoglobin concentration in a living body. Non-Patent Document 1 reports that by calculating the ratio of total reflected light intensity for each wavelength as reflectance using the same procedure as in Patent Document 1, and further using a first-order polynomial of multiple relative absorbance ratios, total hemoglobin concentration and tissue oxygen saturation can be measured with high accuracy while reducing the effects of scattering by the subject, absorption by impurities such as water, protein, and fat, and the temperature of the subject. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2007-271575 [Non-patent literature]

[0005] [Non-Patent Document 1] Report of the 404th Workshop of the Laser Society of Japan, Laser Applications, No. RTM-10-27, pp. 1-5, 2010. Summary of the Invention [Problem to be solved by the invention]

[0006] However, the above-described conventional nondestructive measuring device for a light scattering medium has the following problems. Specifically, there is a problem that measurement accuracy deteriorates when the scattering coefficient of the subject or the composition of impurities changes over time, or when the positional relationship between the light receiving port of the light receiving unit and the subject changes over time due to the subject's movement.In particular, when measuring trace amounts of blood components such as blood glucose levels, the reflectance fluctuates over time due to the influence of the subject's slight movement or heartbeat during measurement, resulting in large measurement errors. Furthermore, as will be described in detail later, the inventors' investigations have revealed that the above measurement errors cannot be easily removed by simple processing such as filtering. Based on this knowledge, the inventors have conducted further investigations and completed the present invention.

[0007] An object of the present invention is to provide a non-destructive measurement device for light scattering bodies that can obtain good measurement accuracy even if the scattering coefficient of the subject or the composition of impurities changes over time, or even if the positional relationship between the light receiving opening of the light receiving unit and the subject changes over time due to movement of the subject. [Means for solving the problem]

[0008] The present invention is a non-destructive measurement device for light scattering bodies, comprising: a light source unit that generates light of multiple wavelengths; a light irradiating unit that has a light outlet that emits light of multiple wavelengths and irradiates the light of the multiple wavelengths from the light outlet toward a single irradiation area on a test body made of a light scattering body; two light receiving units that are located on the same line as the center of the light outlet and have light receiving units that are at different distances from the light outlet and that simultaneously receive reflected light after the light of the multiple wavelengths irradiated from the light outlet passes through the inside of the test body; two light detecting units that simultaneously detect the light intensity of the light received by each of the light receiving units; a signal processing unit that extracts light intensity for each of the multiple wavelengths from each of the two light intensities detected by the two light detecting units; and an arithmetic processing unit that calculates reflectance, which is the ratio of the light intensities of the two light receiving units at the same wavelength, for each of the multiple wavelengths and calculates a property characteristic value of the inside of the test body based on the reflectance. The multiple wavelength light has multiple wavelengths λ i Includes light of (i=1,2,…,n; n≧3). The light intensities of the reflected light received at the same time t by two light receiving units with distances of ρ1 and ρ2 (where ρ1<ρ2) are J1(t) and J2(t), respectively. The signal processing unit extracts the wavelength λ from J1(t) and J2(t). i The light intensity for each 1i (t) and J 2i When (t) is used, The calculation processing unit calculates the wavelength λ at time t, which is expressed by the following formula (1): i Reflectance R i (t) is calculated, and the relative absorbance ratio γ j (t) (j=1, ..., m; m=n-2) at time t The property characteristic value c(t) The relative absorbance ratio γ j It is calculated using the calibration formula expressed by the following formula (3) with (t) as an explanatory variable. value c(t) Time average of<c(t)> or by calculating the relative absorbance ratio γ j Time average of (t) < γ j (t)> is an explanatory variable, and the calibration formula expressed as the following formula (4) The value C is the property characteristic value Calculate. R i (t)=J 2i (t) / J1i (t)…(1) gamma j (t)=ln(R j+2 (t) / R1(t)) / ln(R2(t) / R1(t))…(2) c(t)=β0+β1×γ1(t)+β2×γ2(t)+…+β m ×γ m (t)…(3) C=β0+β1×<γ1(t)>+β2×<γ2(t)>+…+β m ×<γ m (t)>…(4) [Effects of the Invention]

[0009] According to the present invention, it is possible to provide a non-destructive measurement device for light scattering bodies that can obtain good measurement accuracy even if the scattering coefficient of the subject or the composition of impurities changes over time, or even if the positional relationship between the light receiving opening of the light receiving unit and the subject changes over time due to movement of the subject. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a schematic diagram showing a general configuration of a nondestructive measuring device for a light scattering medium according to one embodiment of the present invention. [Figure 2] FIG. 1(a) is a diagram showing a sensor probe of the nondestructive measuring device, and FIG. 1(b) is a cross-sectional view taken along line BB in FIG. [Figure 3] FIG. 2 is a functional block diagram showing the functional configuration of a control system of the nondestructive measuring device. [Figure 4] 1A is a diagram showing an example of the relationship between the subject and the probe of the nondestructive measuring device, and FIG. 1B is a partially enlarged view of FIG. [Figure 5] FIG. 5 is a further enlarged view of part of (b) in FIG. [Figure 6] 1 is a graph showing the absorption coefficient and temperature dependence for each wavelength of components used in the theoretical calculation of reflectance. [Figure 7] 1 is a graph showing the absorption coefficients of oxygenated hemoglobin, deoxygenated hemoglobin, and albumin at different wavelengths. [Figure 8]1 is a graph showing theoretically calculated values ​​of absorbance obtained for specimens with different biological compositions. [Figure 9] 9A is a graph showing the refractive index of the object used in the calculation of FIG. 8, and FIG. 9B is a graph showing the boundary conditions. [Figure 10] Graph (a) shows the estimated glucose concentration in subcutaneous interstitial fluid, and graph (b) shows the relationship between the estimation error of the glucose concentration and the absorbance fluctuation caused by electrical noise. [Figure 11] (a) is a graph showing the change in absorbance measured at a fingertip over time, (b) is the absorbance of the fingertip itself, (c) is the absorbance fluctuation due to the contact state between the specimen and the measurement probe, and (d) is the absorbance fluctuation due to electrical noise. [Figure 12] 1 is a graph showing the amount of change in absorbance due to a change in total hemoglobin concentration. [Figure 13] 1 is a graph showing the change in absorbance over time due to pulsation and the amount of change converted into the amount of change in total hemoglobin concentration and albumin concentration. [Figure 14] 1 is a graph showing the change in albumin concentration over time in a fingertip. [Figure 15] 10 is a graph showing the calculation results of absorbance obtained by irradiating a fingertip with light of eight different wavelengths used in the study. [Figure 16] 4 is a graph showing estimated values ​​of glucose concentration obtained by the nondestructive measuring device of the present embodiment. [Figure 17] 1 is a graph showing estimated values ​​of glucose concentration obtained by the nondestructive measuring device of Patent Document 1. [Figure 18] 10 is a graph showing the relationship between the estimation error of the glucose concentration obtained by each measuring device and the time width of the moving average. [Figure 19] (a) is a graph showing the estimated glucose concentration in subcutaneous interstitial fluid, and (b) is a graph showing the relationship between the estimation error of glucose concentration and the absorbance fluctuation due to electrical noise, when the glucose concentration is estimated using relative absorbance. [Figure 20]10 is a graph showing an estimated value of glucose concentration using relative absorbance obtained by the nondestructive measuring device of the present embodiment. [Figure 21] 1 is a graph showing an estimated value of glucose concentration using relative absorbance obtained by the non-destructive measuring device of Patent Document 1. [Figure 22] 10 is a graph showing the relationship between the estimation error of the glucose concentration using the relative absorbance and the relative absorbance ratio and the moving average time width, respectively. DETAILED DESCRIPTION OF THE INVENTION

[0011] An embodiment of the present invention will be described with reference to FIGS. FIG. 1 is a schematic diagram showing the general configuration of a nondestructive measuring device for a light scattering medium according to this embodiment (hereinafter simply referred to as "nondestructive measuring device").

[0012] As shown in FIG. 1, the nondestructive measuring device 1 is a device that optically measures property characteristic values ​​inside a test object 2 made of a light scattering medium that scatters light incident from the outside and emits it to the outside. Fig. 3 is a functional block diagram of the nondestructive measuring device 1. As shown in Fig. 1 and Fig. 3, the nondestructive measuring device 1 includes a light source unit 8, a light source control unit 30, a sensor probe 3 (fixed holding member), a light detection unit 50, a signal processing unit 40, a central control unit 20, and a display unit 23.

[0013] There are various types of specimens 2 to be measured, including fruits and vegetables, living animals and plants, food, beverages, soil samples, cell culture media, specimens for various quantitative analyses, and samples in solid, powder, jelly, liquid, etc. The property characteristic value to be measured is a characteristic value that represents a property correlated with the degree of absorption of incident light by the subject, and any value may be used as long as a calibration formula can be set. Examples of property characteristic values ​​include blood glucose level, moisture, fat, tissue oxygen saturation, and hemoglobin concentration of a living body, as well as sugar content and acidity of fruits and vegetables, moisture content and protein content of wheat flour, starch concentration of potatoes, and the amount of fertilizer components such as nitrogen in soil.

[0014] The nondestructive measuring device 1 may be a specialized device that measures one of the above-mentioned property characteristic values, or may be a general-purpose device that can switch between and measure multiple property characteristic values. In the latter case, the measurer may manually set the property characteristic value to be measured and measurement conditions for each measurement using an operation unit 22 (described later) that performs operation input, or the device may be configured to automatically and continuously measure multiple property characteristic values ​​for the same specimen. In the following description, the case where the device is used as a non-invasive measuring device for measuring blood glucose levels will be mainly described as an example.

[0015] As shown in FIG. 1, the light source unit 8 of this embodiment includes a plurality of light sources 8A1, 8A2, . . . , 8A n (n≧3) and emits light of three or more wavelengths. The wavelength emitted by light source unit 8 can be selected and set within the range of 1000 nm to 2000 nm, for example, in the case of non-invasive measurement of blood glucose levels.

[0016] Light source 8A1, 8A2,…, 8A n are the central wavelengths λ1, λ2, ..., λ n A light source that oscillates (n≧3), such as a semiconductor laser, can be used. The light source control unit 30 includes a power supply (not shown) and outputs light having frequencies f1, f2, . . . , f3, which are different for each wavelength, in response to a control signal from the central control unit 20. n The pulse-modulated current is applied to the light sources 8A1, 8A2, ..., 8A n , 10A1, 10A2, ..., 10A n occur simultaneously. Coupling lenses 9A1, 9A2, ..., 9A n are light sources 8A1, 8A2, ..., 8A n Irradiation light 10A1, 10A2, ..., 10A n (n≧3) are optical fibers 4b1, 4b2, ..., 4b n These are optical components for optically coupling the The fiber bundler 7 bundles the optical fibers 4b1, 4b2, ..., 4bn This is an optical component for optically coupling the optical fiber 4b to the optical fiber 4b described later.

[0017] Fig. 2(a) is a side view of the sensor probe 3 as viewed in the direction A in Fig. 1. Fig. 2(b) is a cross-sectional view taken along line BB in Fig. 2(a). As shown in (a) and (b) of FIG. 2, the sensor probe 3 receives irradiation light 10A1, 10A2, . . . , 10A2 emitted from the light source unit 8. n The optical fiber 4 is configured to guide the light from the object surface 2a to the light detection unit 50, and has a structure in which the ends of optical fiber cables 4, 5, and 6 are fixed.

[0018] 2, the optical fiber cable 4 (light irradiation unit) is composed of an optical fiber 4b having an emission end face 4a (light emission port) at one end. The emission end face 4a is fixed in a state aligned on the same plane as the probe end face 3a facing the object surface 2a. The optical fiber cables 5, 6 (light receiving sections) are made up of optical fibers 5b, 6b each having an incident end face 5a, 6a (light receiving port) at one end and an exit end face 5c, 6c at the other end, and are attached to the sensor probe 3.

[0019] The incident end faces 5a and 6a are arranged on two concentric circles with radii ρ1 and ρ2, respectively, in the same radial direction from the center position of the output end face 4a, and are aligned on the same plane as the probe end face 3a. In other words, the radii ρ1 and ρ2 correspond to the distances between the incident end faces 5a and 6a and the center position of the output end face 4a, and ρ2 > ρ1. Therefore, the exit end face 4a and the entrance end faces 5a and 6a are arranged side by side on the same straight line, and the distances between the exit end face 4a and the entrance end faces 5a and 6a are different.

[0020] The structure and material of the optical fibers 4b, 5b, and 6b preferably have low transmission loss, but are not particularly limited as long as the characteristic values ​​for calibration, such as the NA of the light receiving port and the transmission efficiency of the optical fiber for each wavelength, are known. For example, they may be multimode or single mode, and the refractive index distribution may be an appropriate distribution. They may also be glass fiber or plastic fiber. Furthermore, each of the optical fibers 4b, 5b, and 6b may be made up of one or more core wires each having a predetermined core diameter.

[0021] The sensor probe 3 is preferably provided with light absorption at least on the probe end surface 3a so that reflected light from the specimen 2 does not become measurement noise. The degree of light absorption depends on the required measurement accuracy, but for example, in measuring the composition of a living body or the sugar content of fruits and vegetables, a reflectance of 10% or less is preferable, and 5% or less is more preferable. In this embodiment, a black grade polyacetal material is assumed as a material having such good light absorption properties in the near-infrared region.

[0022] The light detection unit 50 includes light receiving elements 13 and 14 that detect outgoing light beams 11 and 12, which are reflected from the object surface 2a and enter the respective incident end faces 5a and 6a, propagate through the optical fibers 5b and 6b, and are emitted together from the exit end faces 5c and 6c, and photoelectrically convert the detected light beams into electrical signals, and amplifier circuits 15 and 16 that amplify the photoelectrically converted electrical signals by a predetermined magnification. The light receiving elements 13 and 14 receive the irradiated light 10A1, 10A2, . . . , 10A n A photodiode or the like having sufficient sensitivity to light of wavelengths of 1000 nm or less can be used.

[0023] The signal processing unit 40 is 11 , L 12 , …, L 1n , and L 21 , L 22 , …, L 2n It is equipped with 2×n lock-in amplifiers. The detection signal E1 output from the amplifier circuit 15 is fed to n lock-in amplifiers L 11 , L12 , …, L 1n The signal is simultaneously input to the lock-in amplifier L 11 , L 12 , …, L 1n Pulsed irradiation light 10A1, 10A2, ..., 10A n The same repetition frequencies f1, f2, ..., f n The reference signals of the lock-in amplifier L 11 , L 12 , …, L 1n Now, let's consider frequencies f1, f2, ..., f from the detection signal E1. n The detection signal E 11 , E 12 , …, E 1n Only the following are extracted: Similarly, the detection signal E2 output from the amplifier circuit 16 is input to n lock-in amplifiers L 21 , L 22 , …, L 2n The signal is simultaneously input to the lock-in amplifier L 21 , L 22 , …, L 2n Pulsed irradiation light 10A1, 10A2, ..., 10A n The same repetition frequencies f1, f2, ..., f n The reference signals of the lock-in amplifier L 21 , L 22 , …, L 2n Now, let us consider frequencies f1, f2, ..., f from the detection signal E2. n The detection signal E 21 , E 22 , …, E 2n Only the following are extracted: In this way, the signal processing unit 40 extracts 2×n detection signals E from the detection signals E1 and E2. 11 , E 12 , …, E 1n , and E 21 , E 22 , …, E 2n is sent to the central control unit 20.

[0024] As shown in Figures 1 and 3, the central control unit 20 is electrically connected to the light source control unit 30, the signal processing unit 40, the operation unit 22, and the display unit 23, and controls the operation of each unit by communicating various control signals and data with each unit. The central control unit 20 according to this embodiment is made up of a device control unit 21 and an arithmetic processing unit 60. The central control unit 20 may be made up of multiple pieces of hardware that perform the operations of the respective functional blocks, or may be made up of a computer equipped with a CPU, memory, an input / output interface, an appropriate storage unit, etc., and programs corresponding to the operations of the respective functional blocks may be executed by this computer.

[0025] The device control unit 21 coordinates and controls the operation of the light source control unit 30, signal processing unit 40, and calculation processing unit 60 in response to operation input from the operation unit 22 operated by the measurer, and performs measurement start, end, measurement operations, etc. The device control unit 21 sends control signals to the light source control unit 30 and the signal processing unit 40 to control the emission and detection of monochromatic light of multiple wavelengths, and also sends a clock signal to synchronize the operations of each unit. The device control section 21 notifies the calculation processing section 60 of the property characteristic value measured in response to the operation input of the operation section 22, and initializes the processing operation of the calculation processing section 60. The device control unit 21 controls operation inputs from the operation unit 22 and displays information such as measurement results sent from the calculation processing unit 60 on the display unit 23 .

[0026] The calculation processing unit 60 of this embodiment has an AD conversion unit 61 , a reflectance calculation unit 62 , a relative absorbance ratio calculation unit 63 , a property characteristic value calculation unit 64 , and a data storage unit 65 . The AD conversion unit 61 converts 2×n analog detection signals, E 11 , E 12 , …, E 1n , and E 21 , E 22 , …, E 2n is converted into a digital signal at a predetermined sampling period. The reflectance calculation unit 62 calculates the 2×n detection signals in time series digitized by the AD conversion unit 61, E 11 (t), E 12 (t), …, E 1n (t), and E 21 (t), E 22 (t), …, E 2n From (t), the wavelength λ defined by the following equation (1) based on the calibration value stored in advance in the data storage unit 65 is obtained. i Time series reflectance R i (t) (i=1, 2, …, n ) is calculated. Here, t in parentheses () represents the time digitized at a predetermined sampling period. R i (t)=J 2i (t) / J 1i (t)…(1)

[0027] The relative absorbance ratio calculation unit 63 calculates the n types of wavelengths λ at time t calculated by the reflectance calculation unit 62. i Reflectance R i From (t), the relative absorbance ratio γ j (t) and sends it to the property characteristic value calculation unit 64. gamma j (t)=ln(R j+2 (t) / R1(t)) / ln(R2(t) / R1(t))…(2) That is, the relative absorbance ratios γ of m (=n-2) time series are calculated using the following equations (2a), (2b), ..., (2m). j Calculate (t) (j=1, 2, ..., m). γ1(t)=ln(R3(t) / R1(t)) / ln(R2(t) / R1(t))…(2a) γ2(t)=ln(R4(t) / R1(t)) / ln(R2(t) / R1(t))…(2b) γ3(t)=ln(R5(t) / R1(t)) / ln(R2(t) / R1(t))…(2c) gamma m (t)=ln(R n (t) / R1(t)) / ln(R2(t) / R1(t))…(2m)

[0028] In the above explanation, R i The term "reflectance" is used for (t) because it is the ratio of the amount of light reflected from the light incident on the surface 2a of the subject, passing through the inside of the subject, and returning to the surface 2a. However, the ratio of the amount of light reflected here can also be considered as "transmittance" because the amount of light transmitted through the inside of the subject 2 and detected at a position distant from the incident position is the amount of light.

[0029] The property characteristic value calculation unit 64 calculates m+1 coefficients β k (k=0, 1, ..., m), and calculate the property characteristic value c(t) at time t using multiple relative absorbance ratios γ j (t) is calculated using a first-order polynomial (equation (3) described later) with the explanatory variable c(t) as the time average of the property characteristic value c(t) at time t.<c(t)> , or the relative absorbance ratio γ j Time average of (t) < γ j The property characteristic value C calculated by a first-order polynomial (Equation (4) described later) with (t)> as an explanatory variable is sent to the device control unit 21.

[0030] The data storage unit 65 stores the coefficients of the calibration formula and the calibration coefficients that are calculated in advance according to the property characteristic values ​​of the object to be measured. For example, when the calibration formula is the above formula (3) or formula (4), β0, β1, ..., β m (m≧1) is stored. The data storage unit 65 is configured from a storage unit such as a ROM, an external storage medium, or an external storage unit.

[0031] The operation of the nondestructive measuring device 1 according to this embodiment, which is configured as described above, during use will be described. First, as shown in FIG. 4(a), the user brings the surface 2a of the subject at the measurement site of the finger 2, which is the subject, into contact or nearly into contact with the probe end surface 3a on the fixed sensor probe 3. The object surface 2a varies depending on the shape of the object, and when the finger 2 shown here is the measurement site, it is natural that there will be unevenness or curvature. For this reason, as shown in the enlarged view of Figure 4(b), there will actually be a gap between the probe end face 3a and the object surface 2a. In this embodiment, it is preferable to bring the output end face 4a into contact with the object 2 with as little gap as possible, but there is no major problem even if there is a small gap between the object surface 2a and the input end faces 5a and 6a.

[0032] When the user inputs an instruction to start measurement via the operation unit 22, the central control unit 20 outputs n kinds of central wavelengths λ1, λ2, . . . , λ3 from the light source unit 8 via the light source control unit 30. n Monochromatic light of different frequencies f1, f2, ..., f n Pulsed irradiation light 10A1, 10A2, ..., 10A n and simultaneously generate the same. Irradiation light 10A1, 10A2,..., 10A n are optical fibers 4b1, 4b2, ..., 4b n and then propagates through one optical fiber 4b via a fiber bundler 7, and is irradiated from an emission end face 4a arranged on the probe end face 3a of the sensor probe 3 toward the object surface 2a.

[0033] The light irradiated onto the surface 2a of the subject is divided into reflected light that is reflected by the surface 2a of the subject, and transmitted light that enters the interior of the subject 2 and passes through the interior. The reflected light is attenuated by repeated reflections between the object surface 2a and the probe end face 3a. The transmitted light is scattered depending on the internal properties of the object 2, travels through various optical paths, and the light that reaches the object surface 2a again is emitted to the outside of the object 2 and enters the optical fibers 5b and 6b at the positions of the incident end faces 5a and 6a, respectively. For example, as shown in FIG. 4(b), internal light 25A and 26A, which are part of the light that has entered the inside of the subject 2, are incident on the incident end face 5a and the incident end face 6a, respectively. The internal light 25A and 26A are transmitted along their respective optical paths while undergoing scattering and absorption depending on the internal properties of the subject 2 between the exit end face 4a and the incident end face 5a or 6a and the distance between the exit end face 4a and the incident end face 5a or 6a. Therefore, the amount of light emitted (amount of reflected light) from each light source contains information on the properties of the subject 2 between the exit end face 4a and the incident end face 5a or 6a.

[0034] In this embodiment, light containing information about the internal properties of the subject 2 according to the arrangement positions of the incident end faces 5a, 6a is transmitted through the optical fibers 5b, 6b and is output together from the output end faces 5c, 6c as output light 11, 12, respectively. These output light beams are then received by the light-receiving elements 13, 14, respectively, and converted into electrical signals, which are then amplified by a predetermined magnification in the amplifier circuits 15, 16. The two amplified detection signals E1, E2 are sent to the signal processing unit 40.

[0035] As described above, the signal processing unit 40 11 , L 12 , …, L 1n , and L 21 , L 22 , …, L 2n It is equipped with 2×n lock-in amplifiers. The detection signal E1 output from the amplifier circuit 15 is input to the lock-in amplifier L 11 , L 12 , …, L 1n The signals are input simultaneously to each lock-in amplifier L 1i Pulsed irradiation light 10A i The same repetition frequency f i The reference signals of frequencies f1, f2, ..., f n The detection signal E is synchronized with 11 , E 12 , …, E 1n are extracted respectively. Similarly, the detection signal E2 output from the amplifier circuit 16 is input to the lock-in amplifier L 21 , L 22 , …, L 2n The signals are input simultaneously to each lock-in amplifier L 2iPulsed irradiation light 10A i The same repetition frequency f i The reference signals f1, f2, ..., f n The detection signal E is synchronized with 21 , E 22 , …, E 2n are extracted respectively. The signal processing unit 40 extracts 2×n detection signals E from the detection signals E1 and E2. 11 , E 12 , …, E 1n , and E 21 , E 22 , …, E 2n is sent to the AD conversion unit 61 of the central control unit 20.

[0036] The AD conversion unit 61 converts 2×n detection signals in time series, E 11 , E 12 , …, E 1n , and E 21 , E 22 , …, E 2n is converted into a digital signal at a predetermined sampling period and sent to the reflectance calculation unit 62.

[0037] The time-series detection signal E digitized by the AD conversion unit 61 1i (t), E 2i (t) is the wavelength λ incident on the end faces 5a and 6a i The reflected light amount J 1i (t), J 2i Using (t), it is expressed by the following equations (5) and (6), where t in parentheses represents the time digitized at a predetermined sampling period. E 1i (t)=J 1i (t)×η 1i ×ε 1i ×GP1×GL 1i …(5) E 2i (t)=J 2i (t)×η 2i ×ε 2i ×GP2×GL 2i …(6) where η 1i, η 2i is the wavelength λ i The transmission efficiency of the optical fibers 5b and 6b at ε 1i , ε 2i is the wavelength λ i GP1 and GP2 represent the amplification factors of the amplifier circuits 15 and 16, respectively. 1i , G.L. 2i is the lock-in amplifier L 1i , L 2i and the amplification factor of

[0038] From equations (5) and (6), the wavelength λ expressed by equation (1) at time t is i Reflectance R i (t) is the detection signal E 1i (t), E 2i (t) is expressed by the following equation (7). R i (t)=J 2i (t) / J 1i (t) =E 2i (t) / E 1i (t) ×(η 1i ×ε 1i ×GP1×GL 1i ) / (η 2i ×ε 2i ×GP2×GL 2i ) =(E 2i (t) / E 1i (t))×κ i …(7) where κ i =(η 1i ×ε 1i ×GP1×GL 1i ) / (η 2i ×ε 2i ×GP2×GL 2i )…(8) The reflectance calculation unit 62 calculates the digitized detection signal ratio E as shown in equation (7). 2i (t) / E 1i (t) is the calibration coefficient κ expressed by equation (8). i Multiplying by the reflectance R i Calculate the calibration coefficient κ iis the time-invariant wavelength λ i The constants can be calculated in advance as constants for each of the plurality of pulses and stored in the data storage unit 65.

[0039] On the other hand, the wavelength λ incident on the end faces 5a and 6a i The reflected light amount J 1i (t), J 2i (t) is calculated using the following formulas (9) and (10), respectively. J 1i (t)=B 1i (t)×Ω1(t)×ΔS1×cosθ1(t)…(9) J 2i (t)=B 2i (t)×Ω2(t)×ΔS2×cosθ2(t)…(10) Here, the symbols appearing in equations (9) and (10) will be explained using FIG. 5. FIG. 5 is an enlarged view of the dotted-line frame C in FIG. 4(b). The internal light 25A and 26A are emitted from minute regions with areas ΔS1 and ΔS2 on the surface 2a of the object. B 1i (t), B 2i (t) represents the radiance from the minute area, that is, the radiant flux per unit area and unit solid angle. Also, the radiant intensity per unit area from the minute area on the surface 2a of the object is expressed as I 1i (t), I 2i (t), then the radiance B 1i (k), B 2i (k) is expressed by the following equations (11) and (12). B 1i (k)=I 1i (t) / 2π…(11) B 2i (k)=I 2i (t) / 2π…(12) In addition, the radiance B 1i (t), B 2i θ1(t) and θ2(t) in Fig. 5 are the angles between the normal direction of the incident end faces 5a and 6a and the normal directions n1 and n2 of the object's microscopic region at the emission positions of the internal light 25A and 26A, respectively, and change over time due to the movement of the object. Ω1(t) and Ω2(t) represent the solid angles of the optical fiber end faces 5a and 6a viewed from the respective emission positions of the internal lights 25A and 26A, and are expressed by the following formula (13). Ω1(t)=A1 / δ1(t) 2 , Ω2(t)=A2 / δ2(t) 2 …(13) However, each solid angle is limited by Ω1(t)≦2×π×(1−cos(NA1)) and Ω2(t)≦2×π×(1−cos(NA2)) and the numerical apertures NA1 and NA2 of the receiving fibers 5b and 6b. In equation (13), A1 and A2 represent the light receiving areas of the end faces 5a and 6a, respectively. i (t) and δ2(t) represent the distance between the object surface and the end faces 5a and 6a from which the internal light 25A and 26A are emitted, respectively, that is, the gap between the probe end face 3a and the object surface 2a at each light receiving position, and change depending on the movement of the object.

[0040] Using equations (9) to (13), the wavelength λ expressed by equation (1) at time t is i Reflectance R i (t) is expressed by the following equation (14). R i (t)=J 2i (t) / J 1i (t) =(B 2i (t)×Ω2(t)×ΔS2×cosθ2(t)) / (B 1i (t)×Ω1(t)×ΔS1×cosθ1(t)) =(I 2i (t) / 2π×A2 / δ2(t) 2 ×ΔS2×cosθ2(t)) / (I 1i (t) / 2π×A1 / δ1(t) 2 ×ΔS1×cosθ1(t)) =(I 2i (t) / I 1i (t)) × G(t)…(14) where: G(t)=((A2 / A1)×(ΔS2 / ΔS1)×(δ1(t) 2 / δ2(t) 2 )×(cosθ2(t) / cosθ1(t))…(15) In equation (15), "((A2 / A1) × (ΔS2 / ΔS1)" represents the ratio of the light-receiving area of ​​the receiving fiber to the radiation area of ​​the object surface, and can be considered constant without changing over time. On the other hand, the remaining part is determined only by the geometric positional relationship between the probe end face 3a and the object surface 2a shown in Figure 5, and therefore becomes a term that changes over time with slight movements of the object due to breathing, body movement, etc., and is not wavelength-dependent. The wavelength λ calculated by the reflectance calculation unit 62 as described above i Reflectance R i (t) is sent to the relative absorbance ratio calculation unit 63.

[0041] The relative absorbance ratio calculation unit 63 calculates the relative absorbance ratio γ j (t) (j=1, ..., m, where m=n-2) is calculated and sent to the property characteristic value calculation unit 64. Here, the reflectance ratio R of two different wavelengths appearing in equation (2) is j+2 R2 / R1 and R2 / R1 are expressed by the following formulas (16) and (17), respectively, based on the definition of formula (1). R2(t2) / R1(t1)=(J 11 (t1)×J 22 (t2)) / (J 12 (t2)×J 21 (t1))…(16) R j+2 (t j+2 ) / R1(t1)=(J 11 (t1)×J 2j+2 (t j+2 )) / (J 1j+2 (t j+2 )×J 21 (t1))…(17) Here, the reflectances R1, R2, and R j+2 The measurement times are t1, t2, and t j+2 It states that:

[0042] Using equations (14) and (15), equation (16) can be expressed as the following equation (16a). R2(t2) / R1(t1) =(I 22 (t2) / I 12 (t2))×(I11 (t1) / I 21 (t1)) × G(t2) / G(t1)…(16a) Similarly, equation (17) can be expressed as equation (17a) below. R j+2 (t j+2 ) / R1(t1) =(I 2j+2 (t j+2 ) / I 1j+2 (t j+2 ))×(I 11 (t1) / I 21 (t1))×G( tj+2 ) / G(t1)…(17a)

[0043] Here, the wavelengths λ1 and λ2, and λ j+2 The reflectances of R1(t1), R2(t2), and R j+2 (t j+2 ) are measured at the same time, that is, t1=t2=t j+2 Then, G(t1)=G(t2)=G(t j+2 ) and equations (16a) and (17a) can be expressed as the following equations (16b) and (17b), respectively. R2(t1) / R1(t1)=(I 22 (t1) / I 12 (t1))×(I 11 (t1) / I 21 (t1))…(16b) R j+2 (t1) / R1(t1)=(I 2j+2 (t1) / I 1j+2 (t1 ))×(I 11 (t1) / I 21 (t1))…(17b) From equations (16b) and (17b), the different wavelengths λ i Reflectance R i If these are measured at the same time, the reflectance ratio is calculated as the radiation intensity I per unit area from the minute area on the surface 2a of the object. 1i (t), I 2i (t), and I 1j (t), I 2jIt can be seen that it is expressed only by (t) and does not depend on the light receiving area ratio of the receiving fiber or the radiating area ratio of the object surface. Furthermore, it is also found to be independent of changes in the geometric positional relationship between the object surface and the probe due to slight object movements caused by breathing, body movements, etc.

[0044] The relative absorbance ratio γ expressed by Equation (2) j (t) is expressed by the following equation (18) using equations (16b) and (17b). gamma j (t)=ln(R j+2 (t) / R1(t)) / ln(R2(t) / R1(t)) ={ln(I 2j+2 (t) / I 1j+2 (t))-ln(I 21 (t) / I 11 (t))} / {ln(I 22 (t) / I 12 (t))-ln(I 21 (t) / I 11 (t))}…(18) From equation (18), the reflectances R1(t), R2(t), and R j+2 (t), the relative absorbance ratio γ j (t) is the radiation intensity I per unit area from a small area on the surface 2a of the object. 11 (t), I 21 (t), I 12 (t), I 22 (t), and I 1j+2 (t), I 2j+2 It can be seen that it is expressed only by (t) and does not depend on the light receiving area ratio of the receiving fiber or the emitting area ratio of the object surface. It also becomes independent of changes in the geometric positional relationship between the object surface and the probe due to slight object movements caused by breathing, body movements, etc. In this way, m relative absorbance ratios γ j (t) (j=1, . . . , m, where m=n−2) is sent to the property characteristic value calculation unit 64.

[0045] The property characteristic value calculation unit 64 calculates the coefficient βk (k=0, 1, . . . , m) is called, and the relative absorbance ratio γ j The characteristic value c(t) of blood glucose level or the like at time t is calculated using the calibration formula expressed by the following first-order polynomial (3) with (t) (j=1, 2, ..., m, where m=n-2) as explanatory variables, and the time average of this characteristic value c(t) is calculated.<c(t)> , or the relative absorbance ratio γ j Time average of (t) < γ j The property characteristic value C is calculated using the calibration formula expressed by the following first-order polynomial (4) with (t)> as an explanatory variable. c(t)=β0+β1×γ1(t)+β2×γ2(t)+…+β m ×γ m (t)…(3) C=β0+β1×<γ1(t)>+β2×<γ2(t)>+…+β m ×<γ m (t)>…(4)

[0046] According to the above-mentioned Non-Patent Document 1, the total hemoglobin concentration (C tHb ) is 0.05 mM to 0.1 mM, estimating the volume fraction of blood in the fingertip to be 2 to 5%. This makes it difficult to directly measure the minute blood glucose levels contained in such a small volume of blood. On the other hand, the glucose concentration in subcutaneous interstitial fluid is delayed by several minutes in response to sudden changes in blood glucose levels, but is consistent at steady state (SN Thennadil et al., Diabetes Technol. Ther., 3 (2001) 357.). Therefore, continuous glucose monitors (hereafter referred to as CGMs) have been commercialized, which can continuously measure glucose concentrations in subcutaneous interstitial fluid by inserting thin needle-shaped sensor electrodes under the skin. Below, we will explain the measurement accuracy using Equation (4) using the example of measuring glucose concentrations in subcutaneous interstitial fluid as blood glucose levels, similar to CGMs.

[0047] The object 2 shown in FIG. s ' and absorption coefficient μ aAssuming that the light scattering medium has spatially uniformly distributed property characteristic values ​​such as those described above, the radiant light intensity I1 per unit area of ​​reflected light 25A that passes through the medium and is received by light-receiving surface 5a can theoretically be expressed by the following equation (19) (TJ Farrel et al., Med. Phys. 19(1992)879.):

[0048]

number

[0049] In equation (19), P0 represents the intensity of the light irradiated from the exit end face 4a onto the specimen 2. The absorption coefficient μ of the specimen 2 a and the reduced scattering coefficient μ S ', then z0=1 / (μ a +μ S ') and α and D are expressed as α=[3μ a (μ a +μ S ')] 0.5 , D=1 / 3 / (μ a +μ S ') and r1 is expressed as r1 2 =z0 2 +ρ1 2 and r2 is expressed as r2 using the boundary condition Ω described later. 2 =(z0+4Ω×D) 2 +ρ1 2 where ρ1 represents the distance between the centers of the output end face 4a and the input end face 5a.

[0050] In addition, the reduced scattering coefficient μ of the object required for the calculation of Equation (19) S As ', the following equation (20) is used, which is expressed as a function of wavelength λ with P1 and P2 as parameters.

[0051]

number

[0052] The boundary condition Ω is expressed by the following equation (21).

[0053]

number

[0054] In equation (21), n rel is the refractive index ratio between the refractive index n1 of the air before the light is incident, and the refractive index n2 of the object after the light is incident, with the object surface as the boundary, and n rel = n2 / n1, and since the refractive index of air is n1 = 1.00, n rel =n2.

[0055] On the other hand, the theoretical value I2 of the light intensity per unit area of ​​the reflected light 26A that passes through the interior and is received by the light-receiving surface 6a is given by the formula (19) expressing the theoretical value I1 of the light intensity described above, where the separation distance ρ1 is changed to the separation distance ρ2.

[0056] The reflectance R is expressed by the following formula (22) using the radiant light intensities I1 and I2 per unit area obtained by formula (19). R=I2 / I1…(22)

[0057] In addition, the relative absorbance ratio γ j (j=1, ..., m, where m=n-2) is the number of distinct n wavelengths λ i Reflectance R at (i=1, ..., n) i It is calculated using the following formula (23). gamma j =ln(R j+2 / R1) / ln(R2 / R1) =(A1-A j+2 ) / (A1-A2)…(23) where A i is the absorbance, A i =-ln(R i ) is defined as

[0058] Next, assuming a fingertip as the object 2, the absorption coefficient μ aIn addition to glucose, which is the target of measurement, the fingertip contains many other absorbing factors such as water, lipids, proteins, and even hemoglobin pigments. Therefore, we assume that the absorbing factors, such as water excluding lipids, are an aqueous solution containing glucose, albumin, and hemoglobin pigments. The volume fraction of lipids is k f Then, the volume fraction of the aqueous solution containing glucose and other substances other than lipids, k as is 1.0-k f The absorption coefficient of lipids is expressed as μ f , the absorption coefficient of the aqueous solution containing glucose, etc. is μ as Then, the absorption coefficient of the fingertip μ a is expressed by the following equation (24). μ a =k f ×μ f +(1-k f )×μ as …(twenty four)

[0059] The concentrations of glucose, albumin, oxygenated hemoglobin (HbO), and deoxygenated hemoglobin (Hb) in the aqueous solution are C (g / dl), C a (g / dl), C HbO (mM), C Hb The absorption coefficient μ of an aqueous solution is expressed in mM. as is expressed as a first-order polynomial (25) for each concentration. μ as =a0+a1×C+a2×C a +a3×T m +a4×C HbO +a5×C Hb …(twenty five)

[0060] Here, a0(mm -1 ) represents the absorption coefficient of water at 0°C. -1 / (g / dl)) and a2(mm -1 / (g / dl)) is the glucose and albumin concentration per 1g / dl, a3(mm -1 / ℃) is per 1℃ of temperature, a4(mm -1 / mM) and a5(mm -1 / mM) is the absorption coefficient μ of HbO and Hb in aqueous solution per 1 mM of each concentration.as These are coefficients that represent the amount of change in each of the above, and will be referred to as absorption coefficients hereinafter.

[0061] Figure 6 shows the absorption coefficient μ of lipids. f The results of measuring the wavelength dependence of the absorption coefficients a0, a1, a2, and a3 are shown. Here, lard oil at 40°C was used as the lipid. Figure 7 also shows the absorption coefficients a4 and a5 of HbO and Hb. The concentration unit of the albumin absorption coefficient a2 is also converted to the same unit (mM) as a4 and a5.

[0062] As shown in Figure 7, the absorption coefficient of hemoglobin dye is significantly different from that of albumin in the short wavelength region of 1300 nm or less, but in the long wavelength region of 1400 nm or more used for measuring glucose concentration, the absorption coefficient of hemoglobin dye shows wavelength dependence similar to that of albumin. Therefore, taking into account the wavelength used for measurement, if the absorption coefficients a4 and a5 of hemoglobin dye are considered to be equal to the absorption coefficient a2 of albumin, then the absorption coefficient μ expressed previously as Equation (25) can be calculated as follows: as can be expressed by the following equation (25a). μ as =a0+a1×C+a2×(C a +C HbO +C Hb )+a3×T m …(25a)

[0063] Figure 8 shows the absorbance A (=-ln(R)) obtained for 250 types of specimens with different biological parameters. Here, the detection distances ρ1 and ρ2 were set to 0.5 mm and 1.5 mm, respectively. In addition, random numbers were used to vary the biological parameters within the following ranges to eliminate correlation between the 250 types of specimens. The changes in biological parameters here are assumed to be differences between subjects, or changes within the same subject over months or years, and do not change over time during measurement. It is also assumed that the positional relationship between the probe end and the surface of the subject, which may occur due to slight movements of the subject, does not change over time. On the other hand, the fluctuations in absorbance δA due to the dark current of the photodiode used to detect reflected light and electrical noise in the amplifier, etc. n is the standard deviation ΔAn =1.0×10 -4 The normal distribution was assumed. Glucose concentration C ranges from 0 to 500 mg / dl, albumin concentration C a was varied in the range of 32.0 g / dL to 37.4 g / dL. According to Non-Patent Document 1, the total hemoglobin concentration C tHb The value ranges from 0.05mM to 0.10mM, but when converted to albumin concentration, it is 0.33g / dL to 0.66g / dL, which is a slight change. Therefore, here, we will use the above albumin concentration C a The change in temperature T m was varied in the range of 23.7℃ to 33.4℃, assuming the temperature of a fingertip. f was changed in the range of 0% to 2.8%. The reduced scattering coefficient μ S ', P1=0.87mm -1 ~1.16mm -1 , P2 was varied in the range of 0.87 to 1.12.

[0064] The boundary condition Ω expressed by equation (21) is expressed as a function of the refractive index n2 of the specimen, as described above. The refractive index of the specimen was calculated using the following equation (26) (TL Troy et al., J Biomed Opt 6 (2001) 167.): n2=k f ×n oil +(1-k f )×n as …(26) where n oil represents the refractive index of lard oil. as represents the refractive index of the aqueous solution, taking into account the albumin concentration of each analyte, temperature dependence, and wavelength dependence. The refractive index of the analyte, n2 (= n ref The average values ​​and standard deviations of the refractive index n2 and the boundary condition Ω are shown in Figure 9. The refractive index n2 decreases as the wavelength increases, with a difference of approximately 0.002 (standard deviation) between specimens. The boundary condition Ω also decreases as the wavelength increases, with a difference of approximately 0.01 (standard deviation) between specimens.

[0065] The glucose concentration of each of the above-mentioned specimens was estimated from the absorbance A shown in Figure 8 and the results are shown in Figure 10(a). Here, n = 8 wavelengths were selected from the wavelength range of 1400 nm to 2000 nm: λ1 = 1400 nm, λ2 = 1420 nm, λ3 = 1530 nm, λ4 = 1580 nm, λ5 = 1690 nm, λ6 = 1770 nm, λ7 = 1860 nm, and λ8 = 1880 nm. Using the absorbances A1(N), A2(N), ..., A8(N) corresponding to each wavelength (N is the specimen number, N = 1, ..., 250), m = 6 relative absorbance ratios γ1(N), γ2(N), ..., γ6(N) were calculated according to Equation (23), and the glucose concentration C(N) (mg / dL) was estimated according to the following first-order polynomial (27). C(N)=β0+β1×γ1(N)+β2×γ2(N)+…+β6×γ6(N)…(27) The coefficients in equation (27) are regression coefficients obtained by multiple regression analysis, and are β0 = 6.7287 × 10 4 , β1=-3.4610×10 4 , β2=7.3092×10 4 , β3= -7.7896×10 4 , β4=3.6663×10 4 , β5=3.4123×10 4 , β6=-4.8272×10 4 The values ​​are used.

[0066] The estimation error of the glucose concentration (defined as the root mean square error, hereafter referred to as "RMSE") was 9.7 mg / dl. Figure 10 (b) shows the relationship between the estimation error RMSE of the glucose concentration and the absorbance fluctuation ΔA due to electrical noise. n Standard deviation of ΔA n The results of the figure show the relationship between RMSE and ΔA n The relationship is expressed by the following approximate formula (28).

[0067]

number

[0068] From equation (28), the limit of measurement accuracy is ΔA nOn the other hand, to achieve the accuracy (2σ) of ±15 mg / dl (international standard ISO15197) required for a self-monitoring blood glucose (SMBG) system using this measurement method, 2 × RMSE must be 15 mg / dl or less. n ≦8×10 -5 It is necessary to keep it to. When the absorbance fluctuation ΔA is a normal distribution caused only by electrical noise, it is easy to achieve ΔA≦8×10 by averaging the absorbance over time. -5 Therefore, by using the technology described in Patent Document 1 to sequentially irradiate light of eight different wavelengths and measure the time average of reflectance or absorbance for each wavelength in turn, it is theoretically possible to measure the glucose concentration with high accuracy using stable absorbance. However, the inventors' investigations revealed that it is difficult to achieve high measurement accuracy by simply averaging absorbance over time, because the absorbance actually measured fluctuates greatly not only due to electrical noise, but also due to the expansion and contraction of blood vessels caused by heartbeats, and even slight body movements of the subject. Therefore, the following describes the accuracy of estimating the glucose concentration obtained by the device configuration according to this embodiment in the case where there is a fluctuation in absorbance due to pulsation or slight movement of the subject.

[0069] As shown in (a) of FIG. 4, the fingertip 2 was placed on the surface of the sensor probe 3, and the absorbance A(t) of the subject was measured by irradiating it with laser light. The results are shown in (a) of FIG. 11. Here, the detection distances in FIG. 2 were set to ρ1 = 3 mm and ρ2 = 6 mm. Furthermore, only the laser light with a wavelength λ0 = 900 nm was irradiated onto the subject 2 by oscillating pulsed light with a repetition frequency of 5 kHz. Two detection signals E from the lock-in amplifier of the signal processing unit 40 shown in FIG. 10 , E 20 was converted into a digital signal at time intervals of 0.2 ms by the AD conversion unit 61. According to equations (7) and (8), the absorbance A(t) (=-ln(R(t))) was calculated from the reflectance R(t) multiplied by the calibration coefficient κ(900 nm) at wavelength λ0=900 nm.

[0070] The absorbance A(t) shown in (a) of FIG. 11 is the absorbance of the fingertip itself shown in (b) as A0, and the absorbance fluctuation due to the contact state between the test object and the measurement probe shown in (c) as δA S , (d) Absorbance fluctuation due to electrical noise is δA n (t), it is expressed by the following equation (29). A(t)=A0(t)+δA S (t)+δA n (t)…(29) A0(t) represents the absorbance of the specimen itself, and its temporal fluctuation is the total hemoglobin concentration C tHb The maximum change δA0 is 0.025. S (t) is caused by a change in the geometric positional relationship between the surface of the object and the end face of the probe due to a slight movement of the object, and is expressed as δA using G(t) defined by Equation (14). s (t) = -ln(G(t)), and the maximum change is 0.013. n (t) is due to the dark current of the photodiode used to detect the reflected light and the electrical noise of the amplifier, etc., and the actual measured value in this example has a standard deviation ΔA n =3.0×10 -4 The change in absorbance due to pulsation or slight movement of the measurement site is determined by the target absorbance stability (standard deviation of fluctuation) ΔA = 4 × 10 -5 δA0 is up to 313 times larger than δA S The influence of each fluctuation is large, with the maximum being 163 times, and it is difficult to reduce the influence of the fluctuation by only averaging it over time.

[0071] On the other hand, in the device configuration according to this embodiment, as shown in equations (16b) and (17b), the above-mentioned eight wavelengths of light λ used for measuring glucose concentration are i Reflectance R i By measuring all (t) at the same time, the fluctuations in absorbance δA0 and δA S , and δA n can offset all of the effects. In the following, the fluctuations in absorbance δA0 and δA S , and δAn The accuracy of measuring the glucose concentration in this embodiment, taking into consideration all of the above influences, will be described below.

[0072] The time change ΔA0 of the absorbance A0 of the test object itself shown in (b) of FIG. 11 is the total hemoglobin concentration C tHb (mM) changes. tHb The change in absorbance A0 due to the change in δA0 / δC tHb The calculation results are shown in Figure 12. At a wavelength of 900 nm, ΔA0 / ΔC tHb =3.64 (1 / mM), the change in hemoglobin concentration can be calculated from the change in absorbance using the following formula (30). δC tHb =δA0 / 3.64(mM)…(30)

[0073] The change in absorbance ΔA0 shown in FIG. 11(b) and the change in total hemoglobin concentration ΔC from ΔA0 according to Equation (30) tHb The results of conversion into δC are shown in Figure 13. As shown in Figure 7, in the long wavelength region of 1400 nm or more used for measuring glucose concentration, the absorption coefficient of hemoglobin pigment shows the same wavelength dependency as albumin, so the change in total hemoglobin concentration can be considered as the change in albumin concentration. Therefore, the change in total hemoglobin concentration δC tHb is converted to the change in albumin concentration δC a The calculated change in albumin concentration, δC (g / dl), is also shown. a The peak value of 0.042 g / dl is observed on average. Therefore, the albumin concentration C a The albumin concentration change δC a The change in albumin concentration obtained by adding three times (t) is expressed by the following formula (31), and the results are shown in FIG. C a (t) = 35.0 + 3.0 × δC a (t)…(31) The peak value of the change in albumin concentration expressed by equation (31) is 0.13 g / dl on average.

[0074] The absorbances A1(t), A2(t), ..., A8(t) corresponding to the N=8 wavelengths λ1=1400 nm, λ2=1420 nm, λ3=1530 nm, λ4=1580 nm, λ5=1690 nm, λ6=1770 nm, λ7=1860 nm, and λ8=1880 nm used in the estimation of the glucose concentration shown in Figure 10(a) were calculated according to equation (29).The results are shown in Figure 15. To calculate the absorbance A0(t) of the test sample itself, the albumin concentration C expressed by Equation (31) a (t) was used. Albumin concentration C a The biological parameters other than (t) do not change over time. The glucose concentration C = 100 mg / dl, the fat volume fraction k f =1.5%, fingertip temperature T m = 29.0℃. The reduced scattering coefficient μ S ' are the two parameters P1 and P2, each 1.0 mm -1 , 1.0 and calculated using equation (20). The detection distances ρ1 and ρ2 were set to 0.5 mm and 1.5 mm, respectively, as in the analysis in Figure 10. The sampling time interval was set to 0.2 ms, and the absorbance was calculated for 10 seconds. Furthermore, the absorbance fluctuation δA due to electrical noise n (t) is the measured value (standard deviation) of 3.0×10 -4 Approximately three times the fluctuation ΔA n =1.0×10 -3 The absorbance fluctuation δA due to the contact state between the sample and the measurement probe was s The measured value shown in FIG. 11(c) was used as (t). Next, the relative absorbance ratio γ j (t) (j=1, ..., 6) is the absorbance A measured simultaneously at all wavelengths. i (t) (i=1, . . . , 8), and the glucose concentration c(t) was calculated using the following equation (3b). c(t)=β0+β1×γ1(t)+β2×γ2(t)+…+β6×γ6(t)…(3b) Here, the regression coefficient β k The same values ​​as the regression coefficients used in equation (27) are used for (k=0, 1, ..., 6).

[0075] The glucose concentration c(t) estimated using equation (3b) is shown in Figure 16. Here, it is displayed with an offset (translation along the vertical axis) for easier comparison. Without a temporal moving average, the estimated glucose concentration has a large error of standard deviation Δc = 90 mg / dl. However, when the time width AT of the moving average is increased to 10 ms, 100 ms, and 300 ms, the standard deviation Δc of the estimated glucose concentration is significantly reduced to 12.7 mg / dl, 3.9 mg / dl, and 2.4 mg / dl. This is because the glucose concentration estimated using equation (3b) does not include estimation errors due to pulsation and slight movements of the subject, which are offset. In other words, the glucose concentration calculated using equation (3b) does not include the estimation error Δc due to electrical noise. n Since only the σ is left, the estimation error can be easily reduced by taking a moving average over time.

[0076] For comparison with the device configuration according to this embodiment, FIG. 17 shows the results of calculating glucose concentrations by measuring eight types of wavelength light at different times. This assumes that the eight types of wavelength light are not irradiated simultaneously, but are irradiated sequentially as in the technology according to Patent Document 1, and the absorbance is measured sequentially. Here, the absorbance of the eight types of wavelength light shown in FIG. 15 is used, and the relative absorbance ratio γ is calculated using the absorbances A1(t), A2(t+Δt), A3(t+2Δt), ... A8(t+7Δt) at intervals of Δt=0.1 seconds. j (t) (j=1, ..., 6) was calculated using the following formula (32), and the glucose concentration was calculated using formula (3b). gamma j ={A j+2 (t+(j+1)×Δt)-A1(t)} / {A2(t+Δt)-A1(t)}…(32)

[0077] As shown in Figure 17, without moving average, the standard deviation Δc of the estimated glucose concentration is 114 mg / dL, which is a large error. Furthermore, by increasing the time interval AT of the moving average to 10 ms, 100 ms, and 300 ms, the standard deviation Δc of the estimated glucose concentration is improved. However, unlike the device configuration of this embodiment, no matter how much AT is increased, noise associated with subject movement cannot be completely eliminated, and the numerical value is unstable. In other words, even if the technology disclosed in Patent Document 1 is applied as is, it is difficult to achieve practical accuracy in applications such as glucose concentration measurement, where periodic and non-periodic noise associated with subject movement occurs. This demonstrates the superiority of the device configuration of this embodiment, which measures the absorbance of light at multiple wavelengths at all wavelengths simultaneously.

[0078] The RMSE of the estimation accuracy of the final glucose concentration C was calculated using the standard deviation Δc of the estimated glucose concentration and the measurement limit value of 3.2 mg / dl shown in equation (28) using the following equation (33). The results are shown in Figure 18.

[0079]

number

[0080] When glucose concentration is estimated based on the embodiment in which the absorbance of eight wavelengths of light is measured at the same time for all wavelengths of light (■ in Figure 18), a moving average over a time width of just 0.05 seconds achieves 2 x RMSE = 15 mg / dl or less (RMSE = 7.5 mg / dl or less), which shows that it is quite realistic to meet the required accuracy of ±15 mg / dl for SMBG (international standard ISO15197). On the other hand, when glucose concentration is estimated based on measuring the absorbance of eight wavelengths of light at different times (□ in Figure 18), it was shown that the RMSE is difficult to meet the above international standard even if the time width of the moving average is increased.

[0081] In this illustrative example, the absorbance fluctuation ΔA n (standard deviation) 3.0 x 10 -4However, by applying the nondestructive measurement device for light scattering materials according to this embodiment, it is possible to increase the measurement value to 1.0 × 10, which is more than three times that value. -3 Large absorbance fluctuations ΔA n Even with a measuring device that has a high level of accuracy, it is possible to estimate glucose concentrations with a high level of accuracy that satisfies international standards.

[0082] Although the embodiment using the relative absorbance ratio has been described so far, similar measurements can also be performed using the relative absorbance itself. This embodiment is slightly less accurate than the embodiment using the relative absorbance ratio, but has the advantage of requiring one less light source, making it possible to construct a simpler and less expensive device. The following describes the differences between the modified example using relative absorbance and the above embodiment.

[0083] The configuration of the nondestructive measuring device differs in that, in addition to the number of required light sources as described above, in the arithmetic processing unit 60, the relative absorbance ratio calculation unit 63 is replaced with a relative absorbance calculation unit 163. Since the only difference between the two is the content of the calculation, the relative absorbance calculation unit 163 can be the same as the relative absorbance ratio calculation unit 63 in terms of hardware. The relative absorbance calculation unit 163 calculates the reflectance at n wavelengths λ at time t calculated by the reflectance calculation unit 62. i Reflectance R i From (t), the relative absorbance Λ is calculated according to the following formula (2'): j (t) and sends it to the property characteristic value calculation unit 64. Λ j (t)=ln(R j+1 (t) / R1(t))…(2′) Furthermore, the relative absorbance Λ of m (=n-1) time series is calculated using the following equations (2a′), (2b′), ..., (2m′). j Calculate (t) (j=1, 2, ..., m). Λ1(t)=ln(R2(t) / R1(t))…(2a′) Λ2(t)=ln(R3(t) / R1(t))…(2b′) Λ3(t)=ln(R4(t) / R1(t))…(2c′) Λ m (t)=ln(Rn (t) / R1(t))…(2m′)

[0084] The property characteristic value calculation unit 64 calculates m+1 coefficients β′ stored in advance in the data storage unit 65. k (k=0,1,...,m) and calculate the property characteristic value c(t) at time t by multiple relative absorbances Λ j Calculated using a first-order polynomial (equation (3') below) with (t) as an explanatory variable, and then the relative absorbance Λ j (t) time average <Λ j The property characteristic value C calculated by a first-order polynomial (the following equation (4′)) with (t)> as an explanatory variable is sent to the device control unit 21. c(t)=β′0+β′1×Λ1(t)+β′2×Λ2(t)+…+β′ m ×Λ m (t)…(3′) C=β′0+β′1×<Λ1(t)>+β′2×<Λ2(t)>+…+β′ m ×<Λ m (t)>…(4′)

[0085] The data storage unit 65 stores the coefficients of the calibration formula and the calibration coefficients that are calculated in advance according to the property characteristic values ​​of the object to be measured. Specifically, since the calibration formula is the above formulas (3') and (4'), β'0, β'1, ..., β' m (m≧1) is stored.

[0086] The relative absorbance calculation unit 163 calculates the relative absorbance Λ according to the formula (2′). j (t) (j=1, ..., m, where m=n-1) is calculated and sent to the property characteristic value calculation unit 64. Here, the reflectance ratio R of two different wavelengths appearing in equation (2') is j+1 / R1 is expressed by the above formulas (17) and (17b) based on the definition of formula (1).

[0087] The relative absorbance Λ expressed by equation (2') j Using the above equation (17b), (t) is expressed by the following equation (18'). Λ j (t)=ln(R j+1 (t) / R1(t)) ={ln(I 2j+1 (t) / I 1j+1 (t))-ln(I 21 (t) / I 11 (t))}…(18′) From equation (18′), the reflectances R1(t) and R j+1 (t) is used to obtain the relative absorbance Λ j (t) is the radiation intensity I per unit area from a small area on the surface 2a of the object. 11 (t), I 21 (t), and I 1j+1 (t), I 2j+1 It can be seen that it is expressed only by (t) and does not depend on the light receiving area ratio of the receiving fiber or the emitting area ratio of the object surface. It also becomes independent of changes in the geometric positional relationship between the object surface and the probe due to slight object movements caused by breathing, body movements, etc. In this way, m relative absorbances Λ calculated at the same time according to equation (2′) j (t) (j=1, . . . , m, where m=n−1) is sent to the property characteristic value calculation unit 64.

[0088] The property characteristic value calculation unit 64 calculates the coefficient β′ stored in advance in the data storage unit 65. k (k=0, 1, . . . , m) is called, and the relative absorbance Λ calculated by the relative absorbance calculation unit 163 is j The characteristic value c(t) of blood glucose level at time t is calculated using a calibration formula expressed as a first-order polynomial (formula 3' above) with (t) (j=1, 2, ..., m, where m=n-1) as an explanatory variable, and the time average of this characteristic value c(t) is calculated.<c(t)> , or relative absorbance Λ j (t) time average <Λ j The property characteristic value C is calculated using a calibration formula expressed as a first-order polynomial (the above formula 4') with (t)> as an explanatory variable.

[0089] Next, the measurement accuracy of this mode will be explained using equation (4') by taking as an example the case where the glucose concentration in subcutaneous interstitial fluid is measured as the blood glucose level in the same way as with CGM. Relative absorbance Λ j(j=1, …, m, where m=n-1) are n distinct wavelengths λ i Reflectance R at (i=1, ..., n) i It is calculated using the following formula (23'). Λ j =ln(R j+1 / R1) =(A1-A j+1 )…(twenty three') Here, A1, A j+1 is the absorbance, A1=-ln(R1), A j+1 =-ln(R j+1 ) are defined respectively.

[0090] Next, the glucose concentration of each of the above-mentioned specimens was calculated from the absorbance A shown in FIG. 8 as a relative absorbance Λ j The results of estimation using this method are shown in Figure 19(a). Here, n = 7 wavelengths were selected: λ1 = 1400 nm, λ2 = 1420 nm, λ3 = 1530 nm, λ4 = 1580 nm, λ5 = 1690 nm, λ6 = 1770 nm, and λ7 = 1880 nm. Using seven absorbances corresponding to each wavelength, A1(N), A2(N), ..., A7(N) (N is the subject number, N = 1, ..., 250), m = 6 relative absorbances Λ1(N), Λ2(N), ..., Λ6(N) were calculated using equation (23'), and the glucose concentration C(N) (mg / dL) was estimated using the following first-order polynomial equation (27'). C(N)=β′0+β′1×Λ1(N)+β′2×Λ2(N)+…+β′6×Λ6(N)…(27′) The coefficients in equation (27′) are regression coefficients obtained by multiple regression analysis, and are respectively β′0=-9.9619×10 3 , β′1=-4.4260×10 4 , β′2=2.7632×10 4 , β′3=-4.6145×10 4 , β′4=3.2252×10 4 , β′5=-2.3759×10 4 , β′6=1.8567×10 4 The values ​​are used.

[0091] The RMSE of the estimation error of the glucose concentration was 9.8 mg / dl. j Estimation error RMSE of glucose concentration using and absorbance fluctuation δA due to electrical noise n Standard deviation of ΔA n The relationship between the absorbance and the relative absorbance ratio γ j For comparison, the estimation error RMSE when using is shown by the symbol ○ in the figure. n The relationship is expressed by the following approximate formula (28').

[0092]

number

[0093] From equation (28'), ΔA n The measurement limit at γ = 0 is 4.6 mg / dl, and the relative absorbance ratio γ j The figure shows that the relative absorbance Λ j To achieve the required accuracy (2 × RMSE ≦ 15 mg / dl) for a self-monitoring blood glucose (SMBG) system using a measurement method using n ≦7×10 -5 It is necessary to keep it to. On the other hand, if the absorbance fluctuation ΔA is a normal distribution caused only by electrical noise, then ΔA≦7×10 by averaging the absorbance over time. -5 Therefore, the above-mentioned relative absorbance Λ j In measurements using this method, seven different wavelengths of light are irradiated sequentially and the reflectance or time average of absorbance is measured for each wavelength in turn, which theoretically makes it possible to measure glucose concentration with high accuracy using stable absorbance.

[0094] Six types of relative absorbance Λ expressed by equation (23′) j (t) (j=1, ..., 6) Seven types of absorbance A measured simultaneously at all wavelengths iThe glucose concentration c(t) was calculated using equation (3b') below. The seven wavelengths used in the glucose concentration estimation shown in Figure 19(a) were λ1 = 1400 nm, λ2 = 1420 nm, λ3 = 1530 nm, λ4 = 1580 nm, λ5 = 1690 nm, λ6 = 1770 nm, and λ7 = 1880 nm. The absorbance corresponding to each wavelength was calculated using the results shown in Figure 15. c(t)=β′0+β′1×Λ1(t)+β′2×Λ2(t)+…+β′6×Λ6(t)…(3b′) Here, the regression coefficient β′ k The same values ​​as the regression coefficients used in equation (27′) are used for (k=0, 1, ..., 6).

[0095] The glucose concentration c(t) estimated by equation (3b') is shown in Figure 20. As with Figure 16, the results are offset for easier comparison. When there is no temporal moving average, the estimated glucose concentration has a large standard deviation Δc = 89.6 mg / dl, but as the time width AT of the moving average is increased to 10 ms, 100 ms, and 300 ms, the standard deviation Δc of the estimated glucose concentration decreases significantly to 12.7 mg / dl, 4.0 mg / dl, and 2.4 mg / dl. This is because the glucose concentration estimated by equation (3b') does not include estimation errors due to pulsation or slight movements of the subject, as they are offset. In other words, the glucose concentration calculated by equation (3b') does not include the estimation error Δc due to electrical noise. n Only the σ is left, and the estimation error can be easily reduced by taking a moving average over time.

[0096] For comparison with the device configuration according to this embodiment, the results of calculating glucose concentrations by measuring seven types of wavelength light at different times are shown in FIG. 21. This assumes that the seven types of wavelength light are not irradiated simultaneously, but are irradiated sequentially as in the technology according to Patent Document 1, and the absorbance is measured sequentially. Here, the absorbance of the seven types of wavelength light from the absorbance shown in FIG. 15 is used to calculate the relative absorbance Λ using the absorbances A1(t), A2(t+Δt), A3(t+2Δt), ... A7(t+6Δt) at intervals of Δt=0.1 seconds. j(t) (j=1, ..., 6) was calculated using the following formula (32'), and the glucose concentration was calculated using formula (3b'). Λ j =A1(t)-A j+1 (t+j×Δt)…(32′)

[0097] As shown in Figure 21, without moving average, the standard deviation Δc of the estimated glucose concentration is 113 mg / dL, which is a large error. Furthermore, increasing the time interval AT of the moving average to 10 ms, 100 ms, and 300 ms improves the standard deviation Δc of the estimated glucose concentration. However, unlike the device configuration of this embodiment, no matter how much the time interval AT is increased, noise associated with subject movement cannot be completely eliminated, and the standard deviation Δc cannot be reduced. In other words, even if the technology disclosed in Patent Document 1 is applied as is, it is difficult to achieve practical accuracy in applications such as glucose concentration measurement, which generates periodic and non-periodic noise associated with pulsation and subject movement. This demonstrates the advantages of the device configuration of this embodiment, which simultaneously measures the absorbance of light at multiple wavelengths, and this advantage is also fully maintained in this embodiment, which uses relative absorbance.

[0098] The RMSE (symbol ○) of the estimation accuracy of the final glucose concentration C was calculated using the standard deviation Δc of the estimated glucose concentration and the measurement limit value 4.6 mg / dl shown in equation (28′) using the following equation (33′). j The estimation accuracy RMSE is shown by the symbol ■.

[0099]

number

[0100] Based on the embodiment in which the absorbance of seven wavelengths of light is measured at the same time for all wavelengths, the glucose concentration is calculated as the relative absorbance Λ j The accuracy RMSE (symbol ○) estimated by the relative absorbance ratio γ jAlthough this is slightly larger than the RMSE (symbol ■) of the embodiment estimated in (a), by simply performing a moving average over a time width of 0.05 seconds, an RMSE of 7.3 mg / dl or less can be achieved, fully satisfying the required accuracy for SMBG of 2 x RMSE ≦ 15 mg / dl. In this case, too, 1.0 × 10 -3 Absorbance fluctuation ΔA due to relatively large electrical noise n Using a measuring device with this function, it is possible to estimate glucose concentrations with a high degree of accuracy that meets international standards.

[0101] The above describes one embodiment of the present invention and its variations, but the technical scope of the present invention is not limited to the above embodiment, and it is possible to change the combination of components, make various modifications to each component, or delete components, without departing from the spirit of the present invention.

[0102] For example, in the embodiment, the light intensity for each wavelength is simultaneously extracted by a signal processing unit having a plurality of lock-in amplifiers. However, instead of this, the time-series signals E1(t) and E2(t) detected by the light detection unit are first digitized, and then the same signal processing as that of the lock-in amplifier is performed digitally to extract the light intensity E for each wavelength. 1i (t), E 2i In other words, in the present invention, it is not essential that the light intensities of a plurality of wavelengths are extracted simultaneously. [Explanation of symbols]

[0103] 1. Non-destructive measurement device for light scattering materials 2. Subject 4. Optical fiber cable (light irradiation part) 4a Output end face (light exit) 5, 6 Optical fiber cable (receiving part) 5a, 6a Input end face (light receiving aperture) 8 Light source unit 40 Signal Processing Section 50 Light detection unit 60 Processing unit L 11, L 12 , L 1n , L 21 , L 22 , L 2n Lock-in amplifier

Claims

1. a light source unit that generates light of multiple wavelengths; a light irradiation unit having a light exit port for emitting the light beams of the plurality of wavelengths, and irradiating the light beams of the plurality of wavelengths from the light exit port toward one irradiation area on an object made of a light scattering medium; two light receiving units which are located on the same straight line as the center of the light exit port, each having a light receiving port at a different distance from the light exit port, and which simultaneously receive reflected light of the light of the plurality of wavelengths irradiated from the light exit port and transmitted through the inside of the subject; two light detecting units that simultaneously detect the light intensity of the light received by each of the light receiving units; a signal processing unit that extracts light intensities for each of the plurality of wavelengths from the two light intensities detected by the two light detection units; a processor that calculates a reflectance, which is a ratio of the light intensities of the two light receiving units at the same wavelength, for each of the plurality of wavelengths, and calculates a property characteristic value of the inside of the subject based on the reflectance; Equipped with The multiple wavelength lights include multiple wavelengths λ ｉ (i=1, 2, ..., n; n≧3) light, The distance is ρ １ and ρ ２ (However, ρ １ <ρ ２ ) and the light intensity of the reflected light received at the two light receiving units at the same time t is expressed as J １ (t), J ２ (t), and the signal processing unit １ (t) and J ２ The wavelength λ extracted from (t) ｉ The light intensity for each １ｉ (t) and J ２ｉ When (t) is used, The arithmetic processing unit The wavelength λ at the time t is expressed by the following formula (1): ｉ Reflectance R ｉ (t) is calculated, A plurality of relative absorbance ratios γ at the time t, which are represented by the following formula (2): ｊ (t) (j=1, ..., m; m=n-2) is calculated, The property characteristic value c(t) at the time t is expressed as the relative absorbance ratio γ ｊ Calculation is performed using the calibration formula represented by the following formula (3) with (t) as an explanatory variable: by calculating the time average <c(t)> of the values ​​c(t), or by calculating the relative absorbance ratio γ ｊ Time average of (t) < γ ｊ The value C, which is the property characteristic value, is calculated using a calibration formula represented by the following formula (4) in which (t)> is an explanatory variable. Non-destructive measurement device for light scattering materials. R ｉ (t)=J ２ｉ (t) / J １ｉ (t)…(1) γ ｊ (t)=ln(R ｊ＋２ (t) / R １ (t)) / ln(R ２ (t) / R １ (t))…(2) c(t)=β ０ +b １ ×c １ (t)+β ２ ×c ２ (t)+…+b Ｍ ×c Ｍ (t)…(3) C=β ０ +b １ ×<c １ (t)>+b ２ ×<c ２ (t)>+…+b Ｍ ×<c Ｍ (t)>…(4)

2. a light source unit that generates light of multiple wavelengths; a light irradiation unit having a light exit port for emitting the light beams of the plurality of wavelengths, and irradiating the light beams of the plurality of wavelengths from the light exit port toward one irradiation area on an object made of a light scattering medium; two light receiving units which are located on the same straight line as the center of the light exit port, each having a light receiving port at a different distance from the light exit port, and which simultaneously receive reflected light of the light of the plurality of wavelengths irradiated from the light exit port and transmitted through the inside of the subject; two light detecting units that simultaneously detect the light intensity of the light received by each of the light receiving units; a signal processing unit that extracts light intensities for each of the plurality of wavelengths from the two light intensities detected by the two light detection units; a processor that calculates a reflectance, which is a ratio of the light intensities of the two light receiving units at the same wavelength, for each of the plurality of wavelengths, and calculates a property characteristic value of the inside of the subject based on the reflectance; Equipped with The multiple wavelength lights include multiple wavelengths λ ｉ (i=1, 2, ..., n; n≧2) light, The distance is ρ １ and ρ ２ (However, ρ １ <ρ ２ ) and the light intensity of the reflected light received at the two light receiving units at the same time t is expressed as J １ (t), J ２ (t), and the signal processing unit １ (t) and J ２ The wavelength λ extracted from (t) ｉ The light intensity for each １ｉ (t) and J ２ｉ When (t) is used, The arithmetic processing unit The wavelength λ at the time t is expressed by the following formula (1): ｉ Reflectance R ｉ (t) is calculated, A plurality of relative absorbances Λ at the time t, represented by the following formula (2'): ｊ (t) (j=1, ..., m; m=n-1) is calculated, The value c(t) which is the property characteristic value at the time t is expressed as the relative absorbance Λ ｊ Calculation is performed using the calibration formula represented by the following formula (3') with (t) as an explanatory variable: by calculating the time average <c(t)> of the value c(t), or by calculating the relative absorbance Λ ｊ Time average of (t) < Λ ｊ The value C, which is the property characteristic value, is calculated using the calibration formula represented by the following formula (4') with (t)> as an explanatory variable. Non-destructive measurement device for light scattering materials. R ｉ (t)=J ２ｉ (t) / J １ｉ (t)…(1) Λ ｊ (t)=ln(R ｊ＋１ (t) / R １ (t))…(2′) c(t)=β′ ０ +b′ １ ×L １ (t)+β′ ２ ×L ２ (t)+…+b′ ｍ ×L ｍ (t)…(3′) C=β′ ０ +b′ １ ×<Λ １ (t)>+b′ ２ ×<Λ ２ (t)>+…+b′ ｍ ×<Λ ｍ (t)>…(4′)

3. The light irradiating unit may emit light having three or more wavelengths λ ｉ (i=1, 2, ..., n; n≧3) monochromatic light is generated at a repetition frequency f ｉ and irradiating the subject with periodic pulsed light of all of the above simultaneously, the signal processing unit has a plurality of lock-in amplifiers, The two light detecting units respectively calculate the light intensity of the reflected light received by the two light receiving units at time t as J １ (t), J ２ When detected as (t), The lock-in amplifier detects the frequency f ｉ is used as a reference signal, and the light intensity J of the reflected light detected by the two light detection units is １ (t), J ２ (t) to wavelength λ ｉ The two received light amounts for each １ｉ (t) and J ２ｉ (t) is extracted as 3. The non-destructive measuring device for a light scattering medium according to claim 1.

Images