JP2006141712A - Bioinformation measuring method and apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To remove noise superimposed on spectrum data as direct-current components. <P>SOLUTION: This bioinformation measuring method quantitatively determining bioinformation based on light absorption characteristics data obtained by entering a light to the interior of the body and detecting it outside the body is so formed as to obtain the light absorption characteristic data ranging over predetermined wavelength bands more than once (steps S1 and S4), preprocess the obtained data of the plurality of light absorption characteristics by a multiplicative scatter correction or a differential processing (step S3), compute an average of the preprocessed data of the plurality of light absorption characteristics by adding (step S5), and quantitatively determine bioinformation based on the obtained average light absorption characteristic data. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a biological information measuring method and a biological information measuring device.

Conventionally, a method of measuring biological information such as glucose concentration in a biological tissue using light absorption in the near-infrared region is known (see, for example, Patent Document 1).
This Patent Document 1 considers the influence of other biological components and disturbance factors such as temperature, scattering, and optical path length when quantifying the glucose concentration in biological components that change every moment such as biological tissue and body fluid. Thus, a quantitative method for quantitatively analyzing the glucose concentration with high accuracy is disclosed.
In this quantification method, in order to obtain one set of spectrum data, 128 times of spectrum measurement are performed, and the addition noise averaging is performed to remove random noise superimposed on the spectrum data.
JP-A-11-178813 (paragraph 0025)

  However, since the noise superimposed on the spectrum data is not limited to random noise, other various noises are also superimposed. Therefore, in the glucose concentration measurement method described in Patent Document 1, although random noise is removed from the obtained spectrum data, a state in which noise other than random noise is included simply by calculating the addition average. In this case, the averaged spectrum data is obtained, and the measurement accuracy cannot be improved.

  The present invention has been made in view of the above-described circumstances, and is calculated by removing noise superimposed as a direct current component on spectrum data and checking the quality of spectrum data used for calculating biological information. It is an object of the present invention to provide a biological information measuring method and apparatus capable of improving the accuracy of biological information.

In order to achieve the above object, the present invention provides the following means.
The present invention is a biological information measuring method for quantifying biological information based on light absorption characteristic data of light detected outside a living body by making light incident on the living body, and the light absorption characteristic data over a predetermined wavelength band is obtained a plurality of times. The pre-processing by multiplicative scattering correction or differential processing is applied to the obtained multiple light absorption characteristic data, and the arithmetic mean of the pre-processed light absorption characteristic data is calculated, and the resulting average light absorption is obtained. A biological information measuring method for quantifying biological information based on characteristic data is provided.

  According to the present invention, light absorption characteristic data over a predetermined wavelength band is acquired a plurality of times, and preprocessing is performed on the acquired light absorption characteristic data. The preprocessing is multiplicative scattering correction or differential processing, and both can remove DC component noise contained in each light absorption characteristic data. In this case, the differentiation process may be any differentiation process equal to or higher than the first-order differentiation. Further, by performing the second-order differentiation process, the peak position can be stored in the light absorption characteristic data before and after the pre-processing, and the subsequent data comparison can be facilitated.

  Then, random noise can be further removed by averaging the light absorption characteristic data. In this case, since the DC component noise is removed by the preprocessing from the light absorption characteristic data that is the basis of the addition average, the DC component noise may be included in the addition average absorption characteristic data obtained as a result of the addition average. The accuracy of the biological information that is prevented and quantified based on the addition average light absorption characteristic data can be improved.

In the above invention, it is preferable to smooth the acquired light absorption characteristic data prior to the pretreatment.
By smoothing, noise that fluctuates finely with a small amplitude can be smoothed, and such fine noise can be prevented from being revealed by preprocessing. Therefore, the effect of pre-processing can be exhibited efficiently.

  Further, in the above invention, the obtained plurality of light absorption characteristic data is grouped into a plurality of sets, and after adding and averaging each set of light absorption characteristic data, the obtained average light absorption characteristic data of each set is obtained. The standard deviation may be calculated, and the quality of the addition average light absorption characteristic data may be evaluated based on the obtained standard deviation value.

  Pre-processing removes DC component noise from the light absorption characteristics data, and random noise is removed by averaging. However, noise that cannot be removed by this will eventually be used as light absorption characteristics data to quantify biological information. It is possible to remain. According to the present invention, by calculating the standard deviation for the addition average absorption characteristic data calculated for each group of the absorption characteristic data grouped into a plurality of sets, the quality of the addition average absorption characteristic data based on the data variation Can be evaluated. As a result of evaluation, if the data varies from a predetermined value, the fact that the data is displayed or that re-measurement is necessary is displayed, so that the biological information is quantified based on the data having a large variation. This can be prevented beforehand.

  The present invention also provides a light source, a measurement head that receives light from the light source into a living body and receives light returning to the outside of the living body, and an absorption over a predetermined wavelength band based on the light received by the measuring head. A data acquisition unit that acquires characteristic data, a preprocessing unit that performs multiplicative scattering correction or differential processing on the plurality of acquired light absorption characteristic data, and an arithmetic average of the plurality of preprocessed light absorption characteristic data Provided is a biological information measuring device including an addition average unit and a biological information quantification unit that quantifies biological information based on the obtained addition average light absorption characteristic data.

  According to the present invention, the light emitted from the light source is incident on the living body via the measurement head, and the light returning to the outside of the living body is received. Absorption characteristic data over a predetermined wavelength band is acquired based on the received light by the operation of the data acquisition unit, and multiplicative scattering correction or differential processing is performed on the plurality of absorption characteristic data by the operation of the preprocessing unit. Is done. The preprocessed light absorption characteristic data is added and averaged by the operation of the addition average unit, and the biological information is quantified based on the addition average light absorption characteristic data by the operation of the biological information quantification unit.

  The quantified biological information is quantified with high accuracy because the quantified biological information is quantified based on the addition average light absorption characteristic data from which the DC component noise is removed by the preprocessing and the random noise is removed by the addition average.

In the said invention, it is good also as providing the smoothing part which performs smoothing with respect to the acquired light absorption characteristic data prior to the said pre-processing.
Since the light absorption characteristic data is smoothed prior to the preprocessing by the operation of the smoothing unit, the noise that fluctuates finely with a small amplitude is smoothed, and the effect of the preprocessing can be efficiently exhibited.

  Further, in the above invention, the addition averaging unit groups the acquired plurality of light absorption characteristic data into a plurality of sets, performs an addition average on the light absorption characteristic data of each set, and obtains the average of each set obtained. It is preferable to include a standard deviation calculation unit that calculates a standard deviation for the light absorption characteristic data and a data quality evaluation unit that evaluates the quality of the addition average light absorption characteristic data based on the obtained standard deviation value.

  The standard deviation of the addition average light absorption characteristic data obtained as a result of the addition averaging is calculated by the operation of the standard deviation calculation unit, and the quality is evaluated by the operation of the data quality evaluation unit. Therefore, it is possible to prevent the biological information from being quantified based on the addition average light absorption characteristic data that varies due to the presence of noise that could not be removed even by preprocessing and addition averaging.

  According to the present invention, it is possible to effectively remove direct current component noise that cannot be removed only by addition averaging, and to accurately quantify biological information. In addition, it is possible to prevent inconvenience that inaccurate biological information is quantified due to unexpected noise.

Hereinafter, a biological information measuring apparatus 1 and method according to a first embodiment of the present invention will be described with reference to FIGS.
As shown in FIG. 1, the biological information measuring apparatus 1 according to the present embodiment is a glucose concentration measuring apparatus that makes a light source 2 composed of a halogen lamp and light emitted from the light source 2 enter a living body A. A measurement head 3 that receives the return light from the living body A, a spectroscope 4 such as a grating that splits the light received by the measurement head 3, and a linear array sensor that detects the dispersed light as spectral data. Such a photodetector 5, a calculation unit 6 that quantifies biological information based on spectrum data detected by the photodetector 5, and a monitor 7 that displays the biological information quantified by the calculation unit 6. I have. In the figure, reference numerals 8 and 9 are condensing lenses, and reference numerals 10 and 11 are optical fibers.

The photodetector 5 is configured to detect, for example, 128 sets of spectral data in one measurement.
As shown in FIG. 2, the calculation unit 6 includes a smoothing unit 12, a preprocessing unit 13, an addition averaging unit 14, and a biological information quantification unit 15.
The smoothing unit 12 is configured to smooth all the spectral data by an arbitrary method such as a Savitzky Golay method or spline interpolation, for example.
For example, the pre-processing unit 13 is configured to perform second-order differentiation on all the spectrum data subjected to the smoothing process by wavelength.

The addition averaging unit 14 is configured to add all 128 sets of preprocessed spectrum data and calculate the average value.
The biological information quantification unit 15 is configured to perform multivariate analysis based on the addition average spectrum data calculated by the addition average unit 14 and to quantify the glucose concentration.

A biological information measuring method using the biological information measuring apparatus 1 according to the present embodiment configured as described above will be described below with reference to FIG.
According to the biological information measuring method according to the present embodiment, spectrum data is measured by the photodetector 5 (step S1), and a smoothing process is calculated on the obtained spectrum data (step S2). Next, the second-order differentiation process is performed on the spectrum data that has been subjected to the smoothing process (step S3). Then, by repeating the above steps S1 to S3 128 times, 128 sets of spectrum data subjected to differentiation processing are obtained (step S4).

  Thereafter, the averaging process is calculated based on the obtained 128 sets of differentiated spectrum data (step S5). Thereby, since addition average spectrum data is obtained, biological information is quantified based on the obtained addition average spectrum data (step S6).

  According to the biological information measuring apparatus 1 and the method according to the present embodiment, the spectral data detected by the photodetector 5 is added and averaged as it is and compared with the conventional method based on the quantitative determination of biological information, prior to the averaging. Since the pre-processing (step S3) for second-order differentiation of the spectrum data by wavelength is performed, the DC component noise inherent in each spectrum data can be removed. Therefore, since biological information can be quantified based on the addition average spectrum data from which direct current component noise has been removed, there is an advantage that more accurate biological information can be obtained.

  Furthermore, according to the biological information measuring apparatus 1 and the method according to the present embodiment, since the spectrum data is subjected to the smoothing process (step S2) prior to the preprocessing step (step S3), there is a slight variation from the spectrum data. Has been removed. Therefore, even if the differential process is performed in the preprocessing step (step S3), it is possible to prevent slight fluctuations from being manifested as noise. Therefore, there is an advantage that the effect of the differential process of removing the DC component noise can be maximized while suppressing the harmful effects of the differential process.

  Further, according to the biological information measuring apparatus 1 and the method according to the present embodiment, the second-order differentiation process by the wavelength is performed as the preprocessing (step S3). However, for the purpose of removing the DC component noise, 1 Differentiating above the first order is sufficient. By performing the second order differential processing as in the present invention, it is possible to remove DC component noise while maintaining the peak position in the detected raw spectrum data. As a result, there is an advantage that the waveforms of the spectrum data before and after the preprocessing (step S3) can be easily associated with each other using the peak position as a mark.

In the above embodiment, the second-order differentiation process is performed as the pre-process (step S3). Instead, multiplicative scattering correction (MSC: Multiplicative
Scattering Correction) may be adopted. According to the MSC, there is an advantage that the DC component can be removed while keeping the characteristics of the waveform of the spectrum data.
In the above embodiment, 128 sets of spectral data are detected in one measurement. However, the present invention is not limited to this, and any plural sets of spectral data may be used. Needless to say.

Next, a biological information measuring apparatus 1 ′ and method according to the second embodiment of the present invention will be described below with reference to FIGS. 4 and 5.
In the description of the present embodiment, the same reference numerals are given to portions having the same configuration as the biological information measuring apparatus 1 according to the first embodiment described above, and the description thereof is omitted.

  As shown in FIG. 4, the biological information measuring device 1 ′ according to the present embodiment is different from the biological information measuring device 1 according to the first embodiment in the addition averaging unit 14 ′, and the standard deviation calculating unit 16. And the quality evaluation unit 17 are also different from the biological information measuring apparatus 1 according to the first embodiment.

  The addition averaging unit 14 'in the present embodiment is configured to divide a plurality of sets of spectrum data preprocessed by the preprocessing unit 13 into a plurality of groups, and to perform an addition averaging process for each group. Has been. For example, as shown in FIG. 5, at the time of one measurement, 128 × 6 sets of spectrum data are detected, grouped into 6 groups of 128 pieces, and an averaging process is performed for every 128 pieces. 6 addition average spectrum data are obtained (step S10).

The standard deviation calculation unit 16 calculates a standard deviation value σ by a known method based on the six averaged spectrum data calculated above (step S11).
In addition, the quality evaluation unit 17 determines whether or not the standard deviation value σ obtained as described above is larger than a predetermined threshold value σ ₀ (step S12). When the obtained standard deviation value σ is smaller than the predetermined threshold value σ _0, it is assumed that the measurement of the noise with little variation in the addition average spectrum data obtained for each group is performed. The process proceeds to the step of quantifying information (step S6). On the other hand, when the obtained standard deviation value is larger than a predetermined threshold value, it can be determined that the variation of the addition average spectrum data is large for each group, so that it is informed to perform re-measurement. (Step S13).

According to the biological information measuring apparatus 1 ′ according to the present embodiment configured as described above, when it is determined by the operation of the quality evaluation unit 17 that the addition average spectrum data obtained for each group has a variation, as it is. It is possible to prevent the biological information from being quantified.
That is, a lot of noise is removed by the smoothing process (step S2), the pre-process (step S3), and the addition averaging process (step S5). When variation occurs, it is possible to prevent inaccurate biometric information from being quantified and to quantify the biometric information with high accuracy.
In the present embodiment, the number of detected spectral data is 128 × 6 sets, but the present invention is not limited to this.

It is the schematic which shows the whole structure of the biological information measuring device which concerns on the 1st Embodiment of this invention. It is a block diagram which shows the calculating part of the biological information measuring device of FIG. It is a flowchart explaining the biometric information measuring method using the biometric information measuring apparatus of FIG. It is a block diagram which shows the calculating part of the biometric information measuring apparatus which concerns on the 2nd Embodiment of this invention. It is a flowchart explaining the biometric information measuring method using the biometric information measuring apparatus of FIG.

Explanation of symbols

A living body 1, 1 'living body information measuring device 2 light source 3 measuring head 5 photodetector (data acquisition part)
12 Smoothing part (smoothing processing part)
13 Pre-processing unit 14, 14 'addition averaging unit 15 biological information quantification unit 16 standard deviation calculation unit 17 data quality evaluation unit

Claims (6)

A biological information measurement method for quantifying biological information on the basis of light absorption characteristic data of light detected outside the living body by making light incident on the living body,
Absorbance characteristics data over a predetermined wavelength band is acquired multiple times,
Pre-processing by multiplicative scattering correction or differential processing is applied to the obtained multiple light absorption characteristics data,
Calculate the average of multiple pre-processed light absorption characteristics data,
A biological information measuring method for quantifying biological information based on the obtained addition average light absorption characteristic data.   The biological information measuring method according to claim 1, wherein smoothing is performed on the acquired light absorption characteristic data prior to the preprocessing. Group multiple acquired absorption characteristics data into multiple sets,
After performing the average of the light absorption characteristics of each set,
Calculate the standard deviation for each set of averaged absorption characteristics data,
The biological information measuring method according to claim 1 or 2, wherein the quality of the addition average light absorption characteristic data is evaluated based on the obtained standard deviation value. A light source;
A measurement head that receives light from the light source into the living body and receives light returning from the living body; and
A data acquisition unit that acquires light absorption characteristic data over a predetermined wavelength band based on light received by the measurement head;
A pre-processing unit that performs multiplicative scattering correction or differential processing on the acquired plurality of light absorption characteristic data,
An averaging means for calculating an average of a plurality of pre-processed light absorption characteristics data;
A biological information measuring device comprising a biological information quantification unit that quantifies biological information based on the obtained addition average light absorption characteristic data.   The biological information measuring device according to claim 4, further comprising a smoothing processing unit that performs smoothing on the acquired light absorption characteristic data prior to the preprocessing. The addition averaging unit groups a plurality of acquired light absorption characteristic data into a plurality of sets, and performs an addition average for each set of light absorption characteristic data,
A standard deviation calculation unit for calculating a standard deviation for each set of averaged absorption characteristics data obtained, and
The biological information measuring device according to claim 4, further comprising a data quality evaluation unit that evaluates the quality of the addition average light absorption characteristic data based on the obtained standard deviation value.

Images