JP3524976B2 - Concentration measuring device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a concentration measuring device for non-invasively measuring the concentration of a component existing in a living body by light.

[0002]

2. Description of the Related Art Conventionally, the basic structure of this type of concentration measuring device is described in detail in, for example, US Pat. No. 4,281,645. In this publication, a change in light absorption of light incident on the head of a human body is detected, and due to this change in light absorption, oxidized hemoglobin (HbO ₂ ), reduced hemoglobin (Hb), HbT (= HbO ₂ + H
b), a technique for measuring the concentration change of components such as cytochrome oxidase (C _y O _x ).

However, since this concentration measuring device cannot detect only the required concentration information of the human body, a device for improving this is US Pat. No. 4,223,680.
The one shown in the publication is proposed. In the same publication,
A technique is disclosed in which detection units are provided at two different distances from the light irradiation point and the difference between the lights detected by these detection units is obtained to obtain only the information inside the human body. A density measuring device using a technique similar to this is disclosed in Japanese Patent Publication No. 2-5.
No. 0,733, JP-A-2-163,634,
It is disclosed in International Application No. WO92 / 21283. These concentration measuring devices have great clinical significance in that they can measure a relatively deep part of human body tissue non-invasively.

The basis of the algorithm of these concentration measuring devices is the Beer-Lambert law shown in FIG. The hatched portion in the figure represents a human body tissue having a thickness of d [cm], and the human body tissue contains a predetermined component of concentration C [Mol]. The light absorption coefficient of human tissue is ε [1 / Mol / cm]
And part of the incident light I _i is absorbed by human tissue,
It appears as emitted light I _o . The Beer-Lambert law is used when there is no light scattering, and the light absorption amount OD is the absorption coefficient ε,
It means a law proportional to the product εCd of the concentration C and the distance d, and is represented by the following equation.

OD = -log (I _o / I _i ) = εCd When there is light scattering, as shown in FIG.
A modified Lambert's law is used.

OD = εCL + X where L is the average optical path length [cm] from the irradiation point to the detection point, and X is the contribution of light that is lost due to scattering without entering the detector. As it is, since X is unknown, it cannot be applied to the concentration measurement. Therefore, in the past, the time change of light absorption Δ
attention to _t OD measures the difference △ _d OD in OD between two points as shown in or using the following _{equation, △ t OD = ε · L} · △ t C 19, each components by the following relationship Was measured for the concentration change and hemoglobin oxygen saturation.

Δ _d OD = OD ₂ −OD ₁ = ε · C · Δ _d L where Δ _d L is the optical path length difference between two points.

[0008]

However, although these equations are used under the condition that the average optical path length L is constant, the average optical path length L is actually the optical absorption coefficient μ shown in the following equation.
_It changes depending on a (i: each component).

[0009]

[Equation 5]

For example, if the light absorption of human tissue is not large, of the light entering the photodetector, the light that detours from the light incident point to the detection point and enters the light detector is attenuated more than the light that enters the detour. For this reason, the proportion of light that is detected by making a relative short cut increases, and as a result, L becomes shorter. On the contrary, if absorption becomes small, L becomes long. Therefore, in reality, the light absorption amount O
There is no proportional relationship between D and concentration C.

[0011] That is, the improved the conventional respective concentration measuring device, basically, a delta _d OD (difference in log) ratio of detected light intensities between two points, and the light absorption coefficient mu _a, detection Based on the relationship between the point distance d and the following equation, the concentration change of hemoglobin and the hemoglobin oxygen saturation (%) are calculated.

[0012]

[Equation 6]

However, in reality, there is no proportional relationship between them as described above. This is because the average optical path length of light changes as the concentration of the component changes.

As described above, since the conventional algorithm directly applies the basic model OD = εCd of the Beer-Lambert rule to the light scattering system, it has such a contradiction / error. Therefore, in order to obtain an accurate algorithm for a light scattering system, it is necessary to perform it based on a method that directly describes the behavior of light in a scatterer.

It is an object of the present invention to provide a concentration measuring device which can process a signal with a novel algorithm and can accurately measure the concentration of a component.

[0016]

SUMMARY OF THE INVENTION The present invention has been made to solve such a problem, and a measuring probe for detecting incident light propagating in a living body and an in-living body in the living body by analyzing the detected light. In a concentration measuring device provided with a calculating means for calculating density information of a predetermined component, this calculating means changes the light absorption amount (optical density OD) obtained from the detected light amount in the distance direction (d) from the light incident point. rate

[0017]

[Equation 1]

Using the first means for computing and the linear relationship that is satisfied by the rate of change of the light absorption amount and the square root of the light absorption coefficient μ _a when the behavior of light in the living body is a diffusion phenomenon, Second means for calculating the concentration ratio of each measured component from the concentration information contained in the absorption coefficient μ _a is provided.

The calculating means is the first means for calculating the change rate of the amount of light absorption with respect to the distance direction from the light incident point, and the light absorption when the behavior of light in the living body is a diffusion phenomenon. Change rate of quantity over time

[0020]

[Equation 3]

[0021] and the time change △ mu _a light absorption coefficient mu _a square root using a linear relationship that satisfies is the density information included in the light absorption coefficient mu _a third for calculating the time variation of the concentration of the measured component And means.

The calculating means is the first means for calculating the change rate of the amount of light absorption with respect to the distance direction from the light incident point, and the light absorption when the behavior of light in the living body is a diffusion phenomenon. A second means for calculating the concentration ratio of each component to be measured from the concentration information contained in the light absorption coefficient μ _a , using the first-order relation that the change rate of the amount and the square root of the light absorption coefficient μ _a satisfy. Using the second linear relationship that the square root of the change in the rate of change of the light absorption amount and the change in the light absorption coefficient μ _a with time Δμ _a is used, the concentration information of the component to be measured is calculated from the concentration information included in the light absorption coefficient μ _a . And a third means for calculating the time change of the concentration.

[0023]

The square component of the rate of change of the light absorption amount with respect to the distance direction from the light incident point is proportional to the light absorption coefficient, and
The proportional constant that controls the proportional relationship is not affected by the light absorption coefficient.

Therefore, the concentration ratio of each component to be measured can be accurately obtained based on this proportional relationship.

The time change of the square of the rate of change of the light absorption amount is proportional to the time change of the light absorption coefficient, and
The proportional constant that controls this proportional relationship is not affected by the light absorption coefficient.

Therefore, based on this proportional relationship, the time change of the concentration of the measured component can be accurately obtained.

Further, if the concentration is measured based on these proportional relationships, the reliability of the obtained concentration information is extremely improved.

[0028]

FIG. 2 shows a probe 10 used in a concentration measuring apparatus according to an embodiment of the present invention. FIG. 2 (a) is a plan view of the probe and FIG. 2 (b) is a sectional view thereof. The probe comprises a light irradiation unit 1 and a light detection unit 2, both of which are substantially integrated with a flexible black silicon rubber holder 3 at an interval of about 5 cm. This interval is roughly 4c
It should be at least m. The light irradiator 1 is composed of an optical fiber bundle 4 and a prism 5, and has a structure for irradiating light of 5 wavelengths sequentially transmitted from the main body of the concentration measuring device to the skin layer of the human body almost vertically. The photodetector 2 is a one-dimensional photosensor arranged in the direction of distance from the light source, and is composed of photodiodes 6 in an array of n sections. The photodetector 2 also includes a preamplifier 7 that integrates and amplifies the photocurrent from the photodiode 6. As a result, even a weak signal can be detected with high sensitivity and transmitted to the measuring apparatus main body via the cable 8. The photodetector 2 may be a two-dimensional photosensor or may be a charge coupled device (CCD) instead of a photodiode.

FIG. 3 shows an example of using this probe. The main subject of the concentration measurement in this example is the head, and clinical monitoring of oxygen is the most important part. The probe 10 is usually fixed to the forehead part without hair with an adhesive tape or an elastic band. The state shown in the figure is fixed by the adhesive tape 9.

FIG. 4 schematically shows the overall structure of the concentration measuring device according to one embodiment of the present invention.

The laser unit 11 is composed of a semiconductor laser, and the laser unit 11 sequentially radiates pulse lights λ _{1 to} λ ₅ having five kinds of wavelengths shown in FIG. 5A. The pulsed lights λ _{1 to} λ ₅ are transmitted through the optical fiber bundle 4 to the probe 1
It is guided to 0, and the living tissue (human body part) is irradiated from the light irradiation unit 1. The light source need not be a semiconductor laser as long as it can sequentially output light of a plurality of wavelengths in the near infrared region.

This pulsed light applied to the forehead propagates inside the head and is detected by the photodiode 6 provided in the photodetector 2 of the probe 10. The n signals detected by the photodiode 6 divided into n are transferred to the main body via the cable 8 and simultaneously sampled by the n sample and hold circuits 12. The sampled signals are simultaneously input to the A / D converter 13, and the A / D converter 1
In FIG. 3, A / D conversion (analog / digital conversion) is sequentially performed as shown in FIG. For example, pulsed light λ ₁
Is converted to a digital value of 11 to 1n, and the pulsed light λ ₂ is converted to a digital value of 21 to 2n. This n is usually 2 to 8, and the total number from the signal 11 at the head to the signal 5n at the end of the A / D-converted signal sequence of 1 cycle is 10 to 40. Further, the time required for the AD conversion of the entire one cycle and the predetermined operation of the CPU 14 is several 100 μsec. Therefore, 1000s per second
Data can be stored in cycles. Since the detection signal obtained by one light irradiation is weak, it is necessary to store data for 5 to 10 seconds in order to obtain sufficient accuracy, as shown in FIG. When this accumulation period ends, CPU1
Reference numeral 4 performs logarithmic conversion of data for each wavelength and each section, and calculates an OD signal according to an algorithm described later. Then, this calculation result is subjected to linear regression processing for each wavelength as shown in the graph of FIG. 1 for each of the pulsed light types λ _{1 to} λ ₅ .

[0033]

[Equation 7]

Calculate The horizontal axis of the graph is the distance d from the light irradiation point to the light detection point, and the vertical axis is the light absorption amount OD.
(Λ) is shown, and the distance d on the horizontal axis is graduated in sections 1 to n.

Based on this calculated value, the CPU 14 solves the equation to be described later, and graphically displays the concentration changes of hemoglobin oxygen saturation SO ₂ [%], HbO ₂ , Hb, and CyOx on the display unit 15. .

Next, the algorithm for measuring the concentration in the present invention will be described.

The state of light incident on the living body using the probe 10 according to this embodiment is shown in FIG. 6. For example, under the light incident conditions as shown in FIG. 6, the light entering the living body 19 has a diffusion coefficient D =
It can be described as a diffuser that moves based on {3 (μ _a + μ ′ _s )} ⁻¹ . That is, in general, the behavior of light in a living body can be approximated as a diffusion phenomenon. Where μ _a is the optical absorption coefficient [1
/ Cm], μ ′ _s is the effective light scattering coefficient [1 / cm]. The impulse light emitted from the laser unit 11 and incident via the optical fiber bundle 4 and the prism 5 is a living body 19
Inside, the output (impulse response) described by the following response function is obtained at the position of distance d from the incident point.

[0038] R (d, t) = ( 4πc / 3μ' s) -2/3 · 1 / μ s · t -5/2 · exp (-μ a Ct) · exp (-3μ' s d 2 / 4Ct) ... However, it is necessary to satisfy the condition of the following equation as the actual measurement condition.

[0039]

[Equation 8]

Since the detected light amount I (d) is a value obtained by temporally integrating the impulse response,

[0041]

[Equation 9]

It becomes Further, since the light absorption amount OD (d) is the logarithm of the reciprocal of the detected light amount I (d),

[0043]

[Equation 10]

It becomes Conventionally, this equation is OD = μ _a L +
It means that it was approximated by X (μ _a = εC).

Since the relationship between OD and μ _a is complicated with the equation as it is, the following rate of change with respect to the distance d

[0046]

[Equation 1]

Attention is paid to This rate of change is expressed by the following equation.

[0048]

[Equation 11]

The first term in this equation is constant with respect to d, and the second term decreases with increasing d. For example, under the condition of the following formula as a general value,

[0050]

[Equation 12]

The second term has a value that is one-tenth or less of the first term, and the rate of change can be approximated only by the first term. Therefore, the rate of change is

[0052]

[Equation 13]

Alternatively, it is expressed by the following equation.

[0054]

[Equation 14]

The algorithm used in the present invention is based on the formula or substantially the formula, and specifically, the following formula is obtained at a plurality of wavelengths λ _i by the measurement probe described above.

[0056]

[Equation 15]

The following formula is measured.

[0058]

[Equation 16]

The relational expression with (j: each component) is made simultaneous to obtain the density ratio of each component j. Here, when the proportional constant 3 μ ′ _s / ln ² 10 in the formula is compared with the conventional proportional constant L, the major difference is that μ _a is not included inside. Therefore μ _a

[0060]

[Equation 17]

Since a proportional relationship is maintained with and, accurate measurement is possible. μ ′ _s can be regarded as almost constant if the measurement wavelength range is appropriately selected, and can be treated as a constant. Specifically, the following is used as the component to be measured.

[0062] oxyhemoglobin (HbO ₂₎ deoxyhemoglobin (Hb) cytochrome Oki synthetase _{(C y O x) H 2} O O ff relative concentration KC _j of each component by solving (offset) the following formula (j) Ask for.

[0063]

[Equation 18]

[0064] Here, a ^{_{K = 3μ' s / ln 2 10}} .

From this, the hemoglobin oxygen saturation (HbO ₂₎ which is a parameter generally used in the clinical setting is obtained.
/ HbT; HbT = the _{HbO 2 + Hb), KC HbO2} /
It can be calculated using the formula (KC _HbO2 + KC _Hb ).

According to the above method, the relative concentrations of all the components are formally determined, but as a practical matter, the measurement accuracy of C _y O _{x with} a small concentration is not sufficient. As a formula to supplement this

[0067]

[Equation 17]

The calculation paying attention to the change with time is simultaneously performed.
This formula itself includes light absorption due to all the components existing in the measurement area, and when dealing with this, there is also water that was not previously measured and offset for correcting unknown absorption terms. It is necessary to solve the equation including it. On the other hand,

[0069]

[Formula 19]

In the case of, the component Hb that changes during measurement
It suffices to solve the equation by considering only O ₂ , Hb, and C _y O _x .
Since it is not necessary to consider the influence of offset or the like, the accuracy is greatly improved (although the measured value is the change in concentration K · ΔC). Specifically, the following formula is used.

[0071]

[Equation 20]

As the measured value of C _y O _x , the density change obtained by the formula is output. Regarding Hb and HbO ₂ ,
Changes in KC _HbO2 and KC _{Hb calculated} by the formula and K _calculated by the formula
△ _t C _HbO2, compared with the K △ _t C _Hb, if both are generally matched, a criterion the accuracy of measuring hemoglobin by the formula is ensured.

As described above,

[0074]

[Equation 17]

By measuring, the non-linear error that occurs when the OD is directly handled as in the conventional case is improved, and

[0076]

[Equation 21]

By simultaneously measuring the change in concentration by, the accuracy and reliability of the measurement are improved. Also,

[0078]

[Equation 1]

The measurement probe is the most important part in performing the measurement of (1) with high accuracy. That is, the measurement probe to be attached to the patient in an actual clinical setting is important for determining the measurement performance, but conventionally, no probe satisfying the wearability and the measurement accuracy has been obtained. However, according to the structure of the probe according to the present embodiment, excellent wearability,
Accurate distance change rate of light absorption

[0080]

[Equation 1]

A probe capable of measuring is obtained. In other words, a measurement probe that can be used at the clinical level is realized,
By processing the signal obtained from this probe with the novel algorithm described above, it becomes possible to accurately measure the concentration of the component.

Next, a confirmation experiment conducted using a phantom for confirming the effectiveness of the concentration measuring apparatus according to the above embodiment will be described.

This experiment was performed using the cooled CCD spectroscopy system shown in FIG. The phantom 21 is made of polystyrene microspheres and an aqueous solution containing an NIR dye, and has a light absorption coefficient μ _a = 0.04 mm ⁻¹ and an effective light scattering coefficient μ ′ _s.
= 2.0 to 1.6 mm ^-1 . Pulsed light enters the phantom 21 from the halogen lamp 22 through the optical fiber 23. This optical fiber 23 is used for the electric moving table 2
4, the optical fiber 23 is moved along with the movement of the electric movable table 24, so that the phantom 2
It is possible to adjust the light incident position on the light source 1.
The light propagating in the phantom 21 is detected by the fiber 25 and converted into image information by the spectroscope and the CCD element 26. This image information is input to the personal computer (PC) 27,
The predetermined arithmetic processing described above is performed. In this experiment, the distance between the optical fibers 23 and 25 was set to 1 mm by the electric moving table 24.
Each of them moves within a range of 20 to 30 mm, and the wavelength λ is 820.
The OD change in nm was measured.

The graph shown in FIG. 8 shows the results of this confirmation experiment, in which the horizontal axis is the Fiber spacing [mm] indicating the spacing d between the fibers, and the vertical axis is the attenuation rate Attenuation of the incident light. (OD). The slope (Slope) of the regression line in the graph is 0.236 OD / mm.
The ratio of the optical attenuation factor OD to the fiber spacing d is given by the following equation.

[0085]

[Equation 11]

When the theoretical value of this change rate is calculated with d = 25 mm in the above equation, the effective light scattering coefficient μ ′ _s of 2.0 to
In the range of 1.6:

[0087]

[Equation 22]

[0088] That is, considering the accuracy of Myu' _s, the theoretical value almost matches the experimental results shown in the graph, the accuracy of the concentration measurement by the concentration measuring apparatus of the present embodiment was confirmed.

Further, in the concentration measurement based on the algorithm of the above embodiment, noise of the device / measuring system causes non-linear distortion in the detected light amount. That is, according to the measurement principle of the concentration measuring apparatus according to the above-mentioned embodiment, the detected light amount (the logarithm of this reciprocal is the detected absorbance) is shown in the graph of FIG. 9A with respect to the actual light amount of the light incident on the detector light receiving portion. It is assumed to have the ideal linear characteristics shown. Here, the horizontal axis of the graph shows the amount of light incident on the detector that actually enters the photodetector, and the vertical axis shows the amount of light detected by this photodetector. Therefore, the detected absorbance OD with respect to the distance d from the light incident point to the light detection point is ideally represented by a regression line passing through the origin shown in the graph of FIG. 9B. Here, the horizontal axis of the graph is the distance d _i −d ₀ between the light incident point position d ₀ and the light detecting point position d _i, and the vertical axis is the absorbance OD ₀ and the light detecting point at the light incident point position d ₀ . The detected absorbance OD _i −OD ₀ corresponding to the difference from the absorbance OD _i at the position d _i is shown.

However, some noise is generally present in an actual device / measurement system. When the amount of light incident on the light receiving portion of the detector is small, the DC contribution of noise is large, and nonlinear distortion occurs in the detected light amount as shown in the graph of FIG. Here, the horizontal axis of the graph shows the detector incident light amount, and the vertical axis shows the detected light amount. That is, even if the actual amount of light incident on the detector becomes 0, there is a detected amount of light,
It happens that the detection signal does not become zero. A of noise
Even if there is a C contribution, the S / N can be improved by accumulating the signal.
However, the DC contribution remains even after accumulation, resulting in the dark subtraction error ± Δ shown in the graph. In the regression processing, this is logarithmically converted, so that the nonlinear distortion is further expanded. This distortion causes an error in the regression calculation, and as shown in the graph of FIG. 10B, a regression line indicated by a broken line should be obtained in the same graph, but a regression line indicated by a solid line is obtained. Change rate of detected light intensity with precise measurement accuracy

[0091]

[Equation 1]

Cannot be obtained. Here, the horizontal axis of the graph shows the distance from the light incident point to the light detection point, and the vertical axis shows the detected absorbance detected at this light detection point. Therefore, when the detected light amount becomes smaller than the detector incident light amount which is the amount of light actually incident on the detector, the regression error of the signal at the light detection point farthest from the light incident point becomes large. For example, the detected absorbance, which should normally be located on the circle in the same graph, is now located on the circle, which includes a large non-linear distortion.

As described above, it is conceivable that the light incident on the living body is extremely absorbed by the living body as a factor that the detected light amount deviates from the proper range. For example, it is conceivable that the incident light is extremely absorbed on the surface of the living tissue due to bleeding, moles, bruises, or the incident light is extremely absorbed inside the living tissue due to internal bleeding. In addition, the amount of light detected decreases even if dirt is present on the light receiving surface of the photodetector or the sensitivity of the detector itself deteriorates. It is also conceivable that the concentration of the measured portion becomes high during the concentration measurement due to bleeding or congestion and the light absorption becomes large and the detected light amount decreases.

Next, the means for correcting this non-linear distortion will be described. The signal distortion that occurs when the detected light amount is insufficient can be corrected by increasing the detected light amount. That is, when the amount of detected light is insufficient, the non-linear distortion is corrected by controlling the light source and increasing the amount of light incident on the living body. Further, as another means for correcting this non-linear distortion, the detection light amount is monitored at each light detection point, a predetermined weight is set for the detection absorbance at each light detection point, and the detection light amount is insufficient during data evaluation. It is possible to reduce the contribution of data at certain detection points. In the present invention, during the regression calculation of the rate of change of the detected absorbance with respect to the distance d, this weighting is set using the y-intercept of the regression line that is simultaneously guided to correct the nonlinear distortion.

FIG. 11 shows an apparatus for correcting the non-linear distortion by controlling the light source and increasing the irradiation light amount when the detected light amount is insufficient. In the figure, parts that are the same as or correspond to those in FIG. 4 are assigned the same reference numerals and explanations thereof are omitted.

The laser section 11 is controlled by a trigger signal for emitting laser light and a signal from the laser light quantity control circuit 31 for controlling the laser light quantity. The CPU 14 monitors the light amount obtained at each light detection point, and when the detected light amount is insufficient, controls the laser light amount control circuit 31 to increase the laser light amount emitted from the laser unit 11 to the biological tissue. By this light amount control, the range of the detected light amount signal becomes a range in which the noise Δ (dark subtraction error) of the detected light amount signal can be ignored, and data with less nonlinear error can be obtained.

FIG. 12 shows an apparatus for weighting the detection data using the y-intercept of the regression line and correcting the non-linear distortion. In the figure, parts that are the same as or correspond to those in FIG. 4 are assigned the same reference numerals and explanations thereof are omitted. Further, FIG. 13 is a flowchart showing an algorithm of data correction processing performed by using this apparatus, and shows a data processing algorithm for each sampling period.

The laser beams of a plurality of wavelengths emitted from the laser section 11 are sequentially guided to the measuring probe 10 and the living tissue 33
Is irradiated. The living tissue 33 has the above-mentioned characteristics of μ _a and μ _s ′, and these incident lights propagating through the living tissue 33 are detected by the photodiode in the measurement probe 10,
The data collection unit 32 collects the detected light amount data.
The data collection unit 32 performs A / D conversion on the detection signal, and collects and stores data of the detected light amount PC during the sampling period (see step 101 in FIG. 13).

Next, the following processing is performed for the wavelength number (LD number) 0 to m of the laser light incident on the living tissue (see step 102). Here, m is the number of LDs + 1 and L
It corresponds to the number of dark lights added to the D number.

First, the detected light amount PC collected and accumulated in the data collecting section 32 is sent to the data processing section 36 by the I / F circuits 34 and 35, and is logarithmically converted into the absorbance OD based on the following equation (step 103). reference). This arithmetic processing and the arithmetic processing described below are performed by the C in the data processing unit 36.
This is performed using the PU 37 and the memory 38.

OD = −log ₁₀ PC Next, linear regression calculation of the detected absorbance OD and the irradiation point-detection point distance d is performed (see step 104). Now, assuming that the detected absorbance OD at the detection point x is y, the regression line is expressed by the following equation using these variables x and y.

Y = γx + β Here, γ is the slope of the regression line and β is the y-intercept.

Considering the weighting W at each detection point in the relation between x and y, γ and β can be expressed by the following equation in a general form by vector notation.

[0104]

[Equation 23]

If the variables x and y change as follows, x = d ₀ , d ₁ , ..., D _n-1 y = OD ₀ , OD ₁ , ..., OD _n-1

[0106]

[Equation 24]

It is

[0108]

[Equation 25]

Is a diagonal matrix, and

[0110]

[Equation 26]

Is

[0112]

[Equation 27]

The transposed matrix of is shown.

By the way, in the case of an ideal regression line, the y-intercept β is 0, and the regression line is shown by the following equation.

Y = γx Further, since OD = −log ₁₀ PC = γx, the function represented by y = γx, that is, when PC = 10 ^−γx is the true value of the detected light amount PC, this true value is obtained. The detected light amount PC ′ assuming the noise amount Δ is expressed by the following equation.

PC ′ = PC + Δ With respect to this detected light amount PC ′, a slope γ ′ including an error was obtained based on the above regression equation. Then, the ratio γ ′ / γ of γ ′ to the assumed slope γ (true value) was calculated for each noise component Δ. In the graph of FIG. 14, the characteristic (without correction) plotted by a white square mark is a graph showing the ratio of γ ′ / γ to the noise component Δ. The horizontal axis of the graph shows the noise amount Δ, and the vertical axis shows the ratio γ ′ / γ. Here, the number of detection points n is 4 (n = 4), and all the elements of the weighting W are equal (equal weighting). That is,

[0117]

[Equation 28]

[0118]

Next, this is weighted to reduce the influence of the noise component Δ (see step 105). The y-intercept β obtained by the above regression calculation is considered to be an amount related to the distortion of the detected light amount due to the noise component Δ. Therefore, the present invention pays attention to this point and performs weighting. FIG. 15 shows an empirically obtained weighting function w3.
(Β), where the horizontal axis represents β and the vertical axis represents this weighting function value w3. This weighting is selected such that the ratio γ '/ γ approaches 1. In this example, of the four detection points, only the weight w3 of the point farthest from the irradiation point is changed. Regression calculation is performed again based on this weighting (see step 106) to obtain γ ′ / γ. 14
The characteristic plotted by the black squares in the graph is the result of this calculation.

The above weighting correction processing is performed for the number of LDs as described above (see step 107). Next, concentration calculation using spectrum curve fitting is performed for each measurement component (see step 108).
Then, the calculation result is displayed in a graph (step 10).
9). The above is the nonlinear distortion correction processing shown along with the processing of one sample.

FIG. 16 is a graph showing the results of actually measuring the hemoglobin oxygen saturation using the apparatus shown in FIG. 12, in which the above-mentioned weighting correction processing using the y-intercept is performed. The figure (a) shows the result of measuring the oxygen saturation by completely bleeding the forearm, and the figure (b) shows the result of measuring the oxygen saturation by bleeding only the veins of the forearm. The horizontal axis of each of these graphs represents time (min.), And the vertical axis represents the hemoglobin oxygen saturation [%]. In this measurement, the concentration measurement using the time-resolved measurement method was performed at the same time as the concentration measurement according to the present example shown in FIG. The characteristic indicated by the solid line in each graph is the characteristic obtained by the apparatus according to the present embodiment, and the characteristic plotted by the open diamond mark indicates the characteristic obtained by using the time-resolved measurement method.
The wavelength of the laser light source used for measurement is 775 nm, 8
There are three wavelengths of 25 nm and 850 nm. The above-mentioned time-resolved measurement method is described in a paper entitled "Application of near-infrared time-resolved spectroscopy to in vivo hemoglobin concentration measurement" on pages 67 to 71 of the following document.

O plus E, No. 180, November 1994 Both measurement results showed a very good agreement as shown in each graph. The time-resolved measurement method requires an ultra-high-speed photodetector and a picosecond pulse laser, is a large-scale device, and has problems in cost and practicality, but its method and the reliability of measured values are widely accepted. ing. On the contrary, the apparatus according to the present embodiment provides a practical apparatus which is much smaller in size and lower in cost than the apparatus using the time-resolved measurement method.

[0123]

As described above, according to the present invention,

[0124]

[Equation 2]

By the calculation method based on the relation, the nonlinear distortion of the conventional algorithm with respect to the change of μ _a can be improved, and the concentration can be calculated with high accuracy. In addition, the reliability of the obtained data is improved by performing the calculation based on both μ _a itself and Δ _t μ _a at the same time.

Further, even when the amount of detected light is insufficient, the measurement error is small.

[0127]

[Equation 1]

Therefore, μ _a is obtained.

[Brief description of drawings]

FIG. 1 is a graph showing a rate of change of a light absorption amount OD with respect to a distance d measured using a concentration measuring device according to an embodiment of the present invention.

FIG. 2 is a plan view and a cross-sectional view showing a probe used in the concentration measuring device according to the present embodiment.

FIG. 3 is a front view showing a usage example of the probe shown in FIG.

FIG. 4 is a block diagram showing a schematic configuration of a system of the concentration measuring device according to the present embodiment.

5 is a waveform chart showing signals in various parts of the system shown in FIG.

FIG. 6 is a cross-sectional view showing a condition of light incident on a living body in the present embodiment.

FIG. 7 is a diagram showing the outline of an experimental device for confirming the effectiveness of the concentration measuring device according to the present embodiment.

FIG. 8 is a graph showing data measured by using the experimental apparatus shown in FIG.

FIG. 9 is a graph showing a relationship between a detector incident light amount and a detected light amount in an ideal case.

FIG. 10 is a graph showing the relationship between the detector incident light amount and the detected light amount when the detector incident light amount is too small.

FIG. 11 is a block diagram showing a schematic configuration of a concentration measuring device including a laser light amount control means.

FIG. 12 is a block diagram showing a schematic configuration of a concentration measuring device having a weighting correction processing function using y-intercepts.

FIG. 13 is a flowchart showing an algorithm of weighting correction processing using y-intercepts.

FIG. 14 is a graph showing a result of simulating the slope of the regression line depending on whether or not the distortion of the detected light amount is corrected.

FIG. 15 is a graph showing a function of a weighting curve used for correction, that is, a y-intercept.

FIG. 16 is a graph showing a measurement result of hemoglobin oxygen saturation obtained by performing a correction process.

FIG. 17: Vias that are the basis of the concentration measurement algorithm
It is sectional drawing explaining Lambert's law.

FIG. 18 is a sectional view for explaining a conventional algorithm.

FIG. 19 is a sectional view for explaining another conventional algorithm.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Light irradiation part, 2 ... Light detection part, 3 ... Silicon rubber holder (black), 4 ... Optical fiber bundle, 5 ... Prism, 6 ...
Photodiode, 7 ... Preamplifier section, 8 ... Cable, 9
... Adhesive fixing tape, 10 ... Probe, 11 ... Laser part,
12 ... Sample and hold circuit, 13 ... A / D converter,
14 ... CPU, 15 ... Display section.

Continuation of the front page (56) Reference JP-A-4-12736 (JP, A) JP-A-4-92646 (JP, A) JP-A-4-215742 (JP, A) JP-A-4-191642 (JP , A) JP-A-2-163634 (JP, A) JP-B 2-50733 (JP, B2) JP-A 7-500259 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB) (Name) A61B 5/145 A61B 10/00 G01N 21/35 G01N 21/49

Claims (8)

(57) [Claims] 1. A concentration measuring device comprising: a measuring probe for detecting incident light propagating through a living body; and a calculating means for analyzing the detected light to calculate concentration information of a predetermined component in the living body. The means is the rate of change of the light absorption amount (OD) obtained from the detected light amount with respect to the distance direction (d) from the light incident point. When the behavior of light in a living body is assumed to be a diffusion phenomenon, the first-order relation satisfied by the rate of change of the light absorption amount and the square root of the light absorption coefficient μ _a is used to calculate the light absorption coefficient μ _{and a} second means for calculating the concentration ratio of each component to be measured from the concentration information contained in a. 2. As the first-order relation, The concentration measuring device according to claim 1, wherein the relationship represented by 3. A concentration measuring device comprising a measuring probe for detecting incident light propagating through a living body, and a calculating means for analyzing the detected light to calculate concentration information of a predetermined component in the living body, The means is the rate of change of the light absorption amount (OD) obtained from the detected light amount with respect to the distance direction (d) from the light incident point. When the behavior of light in a living body is assumed to be a diffusion phenomenon, the change rate of the light absorption amount with time is expressed as Comprising and using a linear relationship root satisfies the time variation [Delta] [mu _a light absorption coefficient mu _a, and a third means for calculating a time change of concentration of the component to be measured from density information which is contained in the light absorption coefficient mu _a A concentration measuring device characterized in that 4. As the first-order relation, The concentration measuring device according to claim 3, wherein the relationship represented by 5. A concentration measuring device comprising a measuring probe for detecting incident light propagating through a living body and a calculating means for analyzing the detected light to calculate concentration information of a predetermined component in the living body, The means is the rate of change of the light absorption amount (OD) obtained from the detected light amount with respect to the distance direction (d) from the light incident point. Using the first means for computing and the first linear relationship that the change rate of the light absorption amount and the square root of the light absorption coefficient μ _a satisfy when the behavior of light in the living body is a diffusion phenomenon, Second means for calculating the concentration ratio of each component to be measured from the concentration information contained in the absorption coefficient μ _a , and time change of the rate of change of the light absorption amount And the temporal change of the optical absorption coefficient μ _a Δ2 _a The second linear relationship satisfied by the square root of the Δ _a is used to calculate the temporal change of the concentration of the measured component from the concentration information included in the optical absorption coefficient μ _a . And a concentration measuring device. 6. The first linear relationship is as follows: Using the relation expressed by, as the second-order relation described above, The concentration measuring device according to claim 5, wherein the relationship represented by 7. The concentration measuring device according to claim 1, further comprising a light irradiation amount control means for controlling an irradiation amount of light incident on a living body. 8. The first means for calculating the change rate of the light absorption amount uses a y-intercept of a regression line when performing regression processing for obtaining the change rate of the light absorption amount at three or more detection points. The concentration measuring device according to any one of claims 1 to 6, wherein a regression process is performed by weighting the light absorption amount with a predetermined weight.