JP2010069065A - Non-invasive blood sugar measuring apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a non-invasive blood sugar measuring apparatus efficiently detecting abnormal measurement and improving the accuracy in the measurement. <P>SOLUTION: The non-invasive blood sugar measuring apparatus comprises: a control means 10 for controlling a light source 20; the light source 20 for irradiating a living body 30 with the light of a specific wavelength; a photoacoustic detecting means 40 for detecting photoacoustic wave signals generated in the living body 30; a characteristic quantity estimating means 50 for estimating the blood sugar value from the output of the photoacoustic detecting means 40; and a characteristic quantity display means 60 for displaying the result of estimation of the blood sugar value. A plurality of waveforms of photoacoustic wave signals, obtained by a plurality of times of estimation of the blood sugar value in addition to the estimation operation of the blood sugar value by the characteristic quantity estimating means 50, are compared, and if the result of comparison is mismatch, it is determined that abnormalities occur during the measurement, so that a process such as re-measurement can be taken. Accordingly, the accuracy in the measurement can be improved. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a measuring apparatus, and more particularly to a noninvasive blood sugar measuring apparatus capable of detecting a measurement error based on a photoacoustic wave signal from a living body and improving measurement accuracy.

  The number of patients with diabetes, a typical lifestyle-related disease, is increasing worldwide. Diabetic patients require routine glycemic control to reduce complications from diabetes and improve the quality of life of patients. Therefore, patients must measure blood glucose regularly every day under the guidance of a doctor.

  As a typical example of measuring a blood glucose level, there is an invasive blood glucose measuring device that measures blood glucose level by collecting blood by inserting a patient's finger with a needle. However, this invasive blood glucose measuring device is troublesome and painful when blood is collected by puncturing a finger with a needle, and in addition, there is a risk of infectious diseases. Non-invasive blood glucose measurement devices have been proposed.

  As a non-invasive blood glucose measurement device, a “biological measurement system” using a photoacoustic effect is described. When light having a wavelength that is absorbed by glucose is irradiated from a biological measurement system onto a part of a living body such as a fingertip, the irradiated light is collected at a relatively small focal region in the living body. In general, light is absorbed by glucose and converted into kinetic energy in the tissue in the region adjacent to the focal region. Kinetic energy increases the temperature and pressure of the absorbing tissue region and generates sound waves. This sound wave is hereinafter referred to as “photoacoustic wave signal”. The photoacoustic wave signal is emitted from the absorbing tissue region and detected by an acoustic sensor included in the biological measurement system. The acoustic sensor is mounted so as to contact the surface of the living body. The intensity of the photoacoustic wave signal is a function of glucose in the absorbing tissue region, and the intensity measured by the acoustic sensor is used to examine the blood glucose level. At this time, in order to minimize the influence of noise in one blood glucose level measurement, light irradiation is performed a plurality of times, and photoacoustic wave signals corresponding to the number of times of irradiation are detected and averaged. And a blood glucose level is estimated based on the averaged signal (for example, patent document 1).

As another means for improving the measurement accuracy, the blood glucose level measured by the noninvasive blood glucose measuring device is used to calibrate the blood glucose level measured by the invasive blood glucose measuring device (hereinafter referred to as calibration). There is a device that performs. However, if a measurement abnormality occurs during calibration, the blood glucose level measurement result after that may be adversely affected. Therefore, a method is disclosed in which the measurement state during calibration is monitored and remeasurement is performed when a measurement abnormality is detected (for example, Patent Document 2).
Special table 2001-526557 gazette JP 2000-60826 A

  However, in the conventional non-invasive blood glucose measurement device as shown in Patent Document 1, after superposing a plurality of photoacoustic wave signals detected by multiple times of light irradiation and averaging the plurality of photoacoustic wave signals, Since blood glucose level is estimated, the following problems occur.

  FIG. 17 shows an averaged photoacoustic wave signal. (A) shows a photoacoustic wave signal when measurement is normal, and (b) shows a photoacoustic wave signal when measurement is abnormal. The horizontal axis represents the time axis direction, and the vertical axis represents the amplitude level.

  When the measurement is performed normally, the photoacoustic wave signal after averaging has a waveform as shown in FIG. 17A, and an improvement in measurement accuracy can be expected.

  However, when a measurement abnormality occurs, for example, when the blood glucose measurement device receives an impact from the outside during the superposition of a plurality of photoacoustic wave signals, the mounting location of the device shifts, and the photoacoustic The superposition position of the wave signal is shifted from the middle. As a result, the averaged photoacoustic wave signal has a waveform as shown in FIG. 17B, and the waveform quality of the photoacoustic wave signal deteriorates, so that the blood sugar level cannot be estimated normally.

  As described above, the conventional non-invasive blood glucose measurement device has a problem that the measurement accuracy decreases when a measurement abnormality occurs.

  Moreover, in the conventional non-invasive measuring apparatus as shown in Patent Document 2, the measurement accuracy is improved by performing calibration. However, a measurement abnormality occurs during measurement other than calibration, and measurement conditions differ from those during calibration. When the measurement is performed, there is a problem that the measurement accuracy is lowered.

  The present invention has been made to solve the above-described problems, and is a noninvasive blood glucose measurement device capable of accurately estimating blood glucose levels without degrading measurement accuracy even when a measurement abnormality occurs. The purpose is to provide.

  In order to solve the above-mentioned problem, a non-invasive blood glucose measurement device according to claim 1 of the present invention provides a photoacoustic wave signal generated by a specific substance in the living body absorbing light energy by light irradiated on the living body surface. In a non-invasive blood glucose measuring device that estimates blood glucose levels by detecting on the surface, an irradiation control signal for controlling irradiation of pulsed light that is repeated at least once is output for each blood glucose level estimation. Control means; at least one light source that irradiates the surface of the living body with the pulsed light according to the irradiation control signal; and photoacoustic detection means that detects the photoacoustic wave signal and outputs a detection signal sampled at a predetermined frequency. The feature amount estimating means for averaging the detection signals for the number of times of irradiation of the pulsed light repeated at least once and estimating the blood glucose level using the averaged detection signals And a feature amount display means for displaying the estimated blood glucose level, wherein the feature amount estimation means includes a plurality of the averaged detection signals obtained by estimating the blood glucose level at least twice. A reference detection signal serving as a comparison reference is created from at least one signal, and each of a plurality of averaged detection signals other than the signal used to create the reference detection signal is used as a comparison detection signal. It is characterized in that whether or not an abnormality has occurred during measurement is determined by comparing the waveform shapes of the detection signal for reference and the reference detection signal.

  A non-invasive blood glucose measurement device according to claim 2 of the present invention is the non-invasive blood glucose measurement device according to claim 1, wherein the control means estimates the blood glucose level at least once. And a second period for estimating blood glucose level at least once after the end of the first period, and a switching signal for switching between the first period and the second period is set as the feature quantity estimation One of the averaged detection signals obtained by the estimated number of times of the blood glucose level at least once when the switching signal indicates the first period. One signal or a plurality of averaged detection signals having the same waveform shape, or a signal obtained by further averaging any number of averaged detection signals among them, is used as a reference detection signal, and the switching signal is a second signal. Indicates the period of A comparison detection signal composed of an averaged detection signal obtained for each estimation of the blood glucose level at least once is compared with the detection signal for reference, and if both the detection signals match, the measurement is normal It is determined that an abnormality has occurred during the measurement if it does not match.

  According to a third aspect of the present invention, there is provided a noninvasive blood sugar measurement device according to the first aspect, wherein the control means estimates the blood sugar level at least once at a predetermined time interval. A third period to be performed, and a part of the third period within the predetermined interval, wherein a fourth period in which blood glucose levels are continuously estimated at least twice or more is set, and The feature quantity estimation means compares the waveform shapes of the averaged detection signals obtained for the estimated number of blood glucose levels at least two times in the fourth period, and the number of averaged detection signals whose waveform shapes match each other. Is equal to or greater than the first threshold, one signal of the plurality of averaged detection signals having the same waveform shape, or a plurality of averaged detection signals having the same waveform shape, or any of them The number average detection signal The reference detection signal is used as the reference detection signal, and the comparison detection signal including the averaged detection signal obtained every time the blood glucose level is estimated at least once in the third period. Compared with the signal, if the two detection signals match, it is determined that the measurement is performed normally, and if they do not match, it is determined that an abnormality has occurred during the measurement.

  According to a fourth aspect of the present invention, there is provided a non-invasive blood sugar measuring device according to the first aspect, wherein the feature quantity estimating means performs at least one time of estimating the blood sugar level once. The irradiation period of the pulse light repeated as described above is divided into n (n is a natural number of 2 or more), an averaged detection signal is obtained in each of the n-divided irradiation periods, and one of the n averaged detection signals is obtained. One reference detection signal, and each comparison detection signal made up of (1-n) averaged detection signals other than the reference detection signal is compared with the reference detection signal. If the waveform shapes match, it is determined that the measurement is performed normally. If they do not match, it is determined that an abnormality has occurred during the measurement.

  A non-invasive blood glucose measurement device according to claim 5 of the present invention is the non-invasive blood glucose measurement device according to claim 1, wherein the characteristic amount estimation means indicates at least one estimated blood glucose level. Using the estimated number of times setting value, the averaged detection signal obtained by estimating the blood sugar level as a reference detection signal using the averaged detection signal in the estimation of the blood sugar level before the number of times indicated by the reference estimated number of times setting value as a reference detection signal The comparison detection signal is compared with the reference detection signal.If the waveform shapes of the two detection signals match, it is determined that the measurement is performed normally. It is determined that an abnormality has occurred.

  A non-invasive blood sugar measuring device according to claim 6 of the present invention is the non-invasive blood sugar measuring device according to any one of claims 2 to 5, wherein the feature amount estimating means is configured to detect the comparison. When the waveform shape of the signal and the reference detection signal is compared, the measurement abnormality determination setting value indicating at least one natural number is used, and the number of consecutive mismatches becomes the number indicated by the measurement abnormality determination setting value. It is determined that the measurement is normally performed until the value reaches, and when the number of times indicated by the measurement abnormality determination setting value is reached, it is determined that an abnormality has occurred during the measurement.

  A non-invasive blood sugar measuring device according to claim 7 of the present invention is the non-invasive blood sugar measuring device according to claim 2, wherein the first period is a calibration mode and is estimated by the non-invasive blood sugar measuring device. This is a period for calculating a calibration formula for calibrating the pre-calibrated blood glucose level with the blood glucose level measured by the invasive blood glucose measuring device.

  Further, the noninvasive blood sugar measurement device according to claim 8 of the present invention is the noninvasive blood sugar measurement device according to claim 2, wherein the second period is a normal measurement mode and is estimated by the noninvasive blood sugar measurement device. In this period, the blood glucose level is estimated by applying a calibration formula to the blood glucose level before calibration.

  A non-invasive blood sugar measuring device according to claim 9 of the present invention is the non-invasive blood sugar measuring device according to claim 1, wherein the feature amount estimating means compares the detection signal for reference and the detection signal for comparison. The reference detection signal and the comparison detection signal are synchronized with the irradiation timing of the pulsed light by the irradiation control signal as a reference position, and the maximum peak position where the absolute value of the amplitude of the reference detection signal is maximized is The position where the polarity of the slope of the detection signal for reference changes at the position closest to the maximum peak position with respect to both positive and negative sampling timings from the maximum peak position is detected as the first comparison start position and the first comparison start position, respectively. In the comparison period from the first comparison start position to the first comparison end position, the number of change in polarity of the slope of the comparison detection signal is And und, it determines that when the count result is 1 measurement is performed normally, if the count result is not 1 abnormality judged to have occurred during the measurement, characterized in that.

  A non-invasive blood glucose measurement device according to claim 10 of the present invention is the non-invasive blood glucose measurement device according to claim 1, wherein the feature amount estimation means compares the detection signal for reference and the detection signal for comparison. The reference detection signal and the comparison detection signal are synchronized with the irradiation timing of the pulse light by the irradiation control signal as a reference position, and after the pulse light irradiation, the reference detection signal and the comparison detection signal In the comparison period in which the shapes of the reference detection signals are compared, normalization is performed such that the maximum amplitude value of the reference detection signal and the maximum amplitude value of the comparison detection signal are equal, and the amplitude value of the reference detection signal and the comparison detection signal The sum of square errors in the amplitude value of the detection signal is calculated. If the sum of square errors is equal to or smaller than the second threshold, it is determined that the measurement is normally performed, and is larger than the second threshold. If the abnormality is judged to have occurred during the measurement, characterized in that.

  A non-invasive blood glucose measurement device according to claim 11 of the present invention is the non-invasive blood glucose measurement device according to claim 1, wherein the feature amount estimation means compares the detection signal for reference and the detection signal for comparison. The reference detection signal and the comparison detection signal are synchronized with the irradiation timing of the pulse light by the irradiation control signal as a reference position, and after the pulse light irradiation, the reference detection signal and the comparison detection signal In a comparison period in which the shapes of the reference are compared, a correlation coefficient between the reference detection signal and the comparison detection signal is calculated, and when the correlation coefficient is greater than a third threshold, the measurement is performed normally. It is judged that, if it is below the third threshold, it is judged that an abnormality has occurred during the measurement.

  A non-invasive blood glucose measurement device according to claim 12 of the present invention is the non-invasive blood glucose measurement device according to claim 1, wherein the feature amount estimation means compares the detection signal for reference and the detection signal for comparison. The reference detection signal and the comparison detection signal are synchronized with the irradiation timing of the pulse light by the irradiation control signal as a reference position, and after the pulse light irradiation, the reference detection signal and the comparison detection signal In the comparison period in which the shapes of the reference detection signals are compared, normalization is performed so that the maximum amplitude value of the reference detection signal is equal to the maximum amplitude value of the comparison detection signal, and the reference detection signal and the comparison detection signal When the number of samplings whose amplitude level exceeds the fourth threshold is counted, and the difference between the count results is not less than the fifth threshold and not more than the sixth threshold, the measurement is normal. Have determined that performed, is larger than the fifth threshold value less than or sixth threshold anomaly is determined to have occurred during the measurement, characterized in that.

  A non-invasive blood glucose measurement device according to claim 13 of the present invention is the non-invasive blood glucose measurement device according to claim 1, wherein the feature amount estimation means compares the detection signal for reference and the detection signal for comparison. The reference detection signal and the comparison detection signal are synchronized with the irradiation timing of the pulse light by the irradiation control signal as a reference position, and after the pulse light irradiation, the reference detection signal and the comparison detection signal In the comparison period in which the shapes of the reference are compared, the number of change in polarity of the slope of the reference detection signal and the comparison detection signal is counted, and the difference between the respective count results is not less than the seventh threshold and not more than the eighth threshold Is characterized in that the measurement is performed normally, and if it is less than the seventh threshold value or greater than the eighth threshold value, it is determined that an abnormality has occurred during the measurement.

  According to a fourteenth aspect of the present invention, there is provided the noninvasive blood sugar measurement device according to the first aspect, wherein the feature amount estimation means compares the reference detection signal with the comparison detection signal. The reference detection signal and the comparison detection signal are synchronized with the irradiation timing of the pulse light by the irradiation control signal as a reference position, and after the pulse light irradiation, the reference detection signal and the comparison detection signal In the comparison period in which the shapes of the reference detection signals are compared, the inclinations of the reference detection signal and the comparison detection signal are respectively calculated, the position where the inclination of the comparison detection signal is equal to or less than a ninth threshold, and the reference detection signal If the slope is equal to or less than the ninth threshold, it is determined that the measurement is performed normally, and if it is different, it is determined that an abnormality has occurred during the measurement. .

  A non-invasive blood glucose measurement device according to claim 15 of the present invention is the non-invasive blood glucose measurement device according to claim 1, wherein the feature amount estimation means compares the detection signal for reference and the detection signal for comparison. The reference detection signal and the comparison detection signal are synchronized with the irradiation timing of the pulse light by the irradiation control signal as a reference position, and after the pulse light irradiation, the reference detection signal and the comparison detection signal The reference detection signal and the comparison detection signal are frequency-converted in a comparison period in which the shapes of the reference detection signal and the comparison comparison detection signal are subjected to frequency conversion. If there is no difference between the frequency components, it is determined that the measurement is performed normally, and if there is a difference in the frequency component, it is determined that an abnormality has occurred during the measurement. And wherein the door.

  A non-invasive blood glucose measurement device according to claim 16 of the present invention is the non-invasive blood glucose measurement device according to claim 5, wherein the feature amount estimation means includes a writable register, and the reference is made by the register. The estimated number of times setting value can be changed from the outside.

  The non-invasive blood glucose measurement device according to claim 17 of the present invention is the non-invasive blood glucose measurement device according to claim 6, wherein the feature amount estimation means includes a writable register, and the measurement is performed by the register. The abnormality determination set value can be changed from the outside.

  The non-invasive blood glucose measurement device according to claim 18 of the present invention is the non-invasive blood glucose measurement device according to claim 3, wherein the feature amount estimation means includes a writable register, and the register uses the register. One threshold value can be changed from the outside.

  The non-invasive blood glucose measurement device according to claim 19 of the present invention is the non-invasive blood glucose measurement device according to claim 10, wherein the feature amount estimation means includes a writable register, and the register uses the register. The second threshold value can be changed from the outside.

  A non-invasive blood glucose measurement device according to claim 20 of the present invention is the non-invasive blood glucose measurement device according to claim 11, wherein the feature amount estimation means includes a writable register, and the register uses the register. The third threshold value can be changed from the outside.

  The non-invasive blood glucose measurement device according to claim 21 of the present invention is the non-invasive blood glucose measurement device according to claim 12, wherein the feature amount estimation means includes a plurality of writable registers, and each of the registers The fourth threshold value, the fifth threshold value, and the sixth threshold value can be changed from the outside.

  The non-invasive blood glucose measurement device according to claim 22 of the present invention is the non-invasive blood glucose measurement device according to claim 13, wherein the feature amount estimation means includes a plurality of writable registers, and each of the registers The seventh threshold value and the eighth threshold value can be changed from the outside.

  The non-invasive blood glucose measurement device according to claim 23 of the present invention is the non-invasive blood glucose measurement device according to claim 14, wherein the feature amount estimation means includes a writable register, and the register uses the register. The ninth threshold can be changed from the outside.

  A non-invasive blood glucose measurement device according to claim 24 of the present invention is the non-invasive blood glucose measurement device according to claim 15, wherein the feature amount estimation means includes a writable register, and the register uses the register. The ten threshold values can be changed from the outside.

  A non-invasive blood glucose measurement device according to claim 25 of the present invention is the non-invasive blood glucose measurement device according to claim 9, wherein the feature quantity estimation means is specified for the comparison detection signal in the comparison period. After performing a filtering process that passes through only the frequency band, the number of change in polarity of the slope of the comparison detection signal is counted.

  A non-invasive blood glucose measurement device according to claim 26 of the present invention is the non-invasive blood glucose measurement device according to claim 13, wherein the characteristic amount estimation means is configured to use the reference detection signal and the comparison signal in the comparison period. The detection signal is subjected to filtering that passes only a specific frequency band, and then the number of change in polarity of the slope of the reference detection signal and the comparison detection signal is counted.

  A non-invasive blood glucose measurement device according to claim 27 of the present invention is the non-invasive blood glucose measurement device according to claim 14, wherein the characteristic amount estimation means is configured to use the reference detection signal and the comparison signal in the comparison period. It is characterized in that after the detection signal is filtered through only a specific frequency band, the slopes of the reference detection signal and the comparison detection signal are calculated.

  According to the non-invasive blood glucose measurement device of the present invention, a blood glucose level is estimated by detecting a photoacoustic wave signal emitted from a specific substance in the living body by absorbing light energy by light irradiated on the living body surface. In the non-invasive blood glucose measurement device, for one estimation of blood glucose level, a control means for outputting an irradiation control signal for controlling irradiation of pulsed light repeated at least once, and the pulse by the irradiation control signal At least one light source for irradiating light on the surface of the living body, photoacoustic detection means for detecting the photoacoustic wave signal and outputting a detection signal sampled at a predetermined frequency, and pulsed light repeated at least once A feature amount estimating means for averaging the detection signals for the number of times of irradiation and estimating the blood glucose level using the averaged detection signals; and a feature for displaying the estimated blood glucose level. A quantity display means, and the feature quantity estimation means generates an abnormality during measurement by comparing the waveform shapes of the plurality of averaged detection signals obtained by estimating blood glucose levels at least twice. It is possible to determine whether the measurement is normal or abnormal by checking the change in the shape of the averaged detection signal. If the measurement is abnormal, the user is notified or remeasured. Therefore, it is possible to measure with high accuracy.

In the following, embodiments of the non-invasive blood sugar measuring device of the present invention will be described in detail with reference to the drawings.
(Embodiment 1)
The noninvasive blood sugar measurement device according to Embodiment 1 of the present invention will be described below.
FIG. 2 is a block diagram showing a configuration of the noninvasive blood sugar measurement device 1 according to the first embodiment.

  The noninvasive blood glucose measurement device 1 shown in FIG. 2 includes a control unit 10, a light source 20, a photoacoustic detection unit 40, a feature amount estimation unit 50, and a feature amount display unit 60. Although the case where one light source 20 is provided is shown here, a plurality of light sources 20 may be provided.

  The control means 10 outputs an irradiation control signal 104 for controlling the irradiation of the pulsed light that is repeated at least once with respect to one estimation of the blood glucose level to the light source 20 and the characteristic amount estimation means 50. Further, the control means 10 repeatedly irradiates the pulsed light 201 in the measurement period signal 101 indicating the measurement period of the blood glucose level, the mode switching signal 102 for switching between the period A and the period B, and the measurement period. An irradiation period signal 103 indicating the period is output to the feature quantity estimation means 50. Period A is a period during which blood glucose level is estimated at least once, and period B is a period during which blood glucose level is estimated at least once after period A ends. In addition, the control unit 10 outputs the feature display control signal 105 to the feature display unit 60 based on the waveform comparison signal 502 output from the feature estimation unit 50.

  The light source 20 irradiates the surface of the living body 30 with the pulsed light 201 based on the irradiation control signal 104.

  The photoacoustic detection means 40 detects a photoacoustic wave signal 301 generated by absorbing a light energy by a specific substance in the living body 30 by using the pulsed light 201 irradiated on the surface of the living body 30, and samples it at a predetermined frequency. The detection signal 401 converted in this way is output to the feature quantity estimation means 50.

  The feature amount estimation means 50 estimates a blood glucose level using an averaged detection signal obtained by averaging the detection signals 401 corresponding to the number of times of irradiation of the pulsed light 201 repeated at least once, and estimates a feature amount that is an estimated blood glucose level. The signal 501 is output to the feature amount display means 60. Further, the feature quantity estimation means 50 compares the waveform shapes of a plurality of averaged detection signals obtained by performing blood glucose level estimation at least twice, and the waveform comparison signal 502 indicating the comparison result is compared with the control means 10. Output to.

  The feature amount display means 60 displays the blood glucose level using the feature amount estimation signal 501 when the measurement is correctly performed. When the measurement abnormality is detected, the feature amount display control signal 105 is input from the control means 10. The measurement error is displayed.

  FIG. 3 is a block diagram showing a detailed configuration of the feature quantity estimation means 50. As shown in FIG.

  3 includes an averaging unit 510, an estimation unit 520, a memory selection unit 540, a comparison detection signal memory 550, a reference detection signal memory 560, and a waveform comparison unit 570.

  The averaging means 510 averages the detection signal 401 using the irradiation period signal 103, the irradiation control signal 104, and the measurement cycle signal 101, and outputs the averaged detection signal 511 to the estimation means 520 and the memory selection means 540. Then, the detection signal averaging end pulse 512 is output to the memory selection means 540.

  The estimation unit 520 estimates a blood glucose level using the averaged detection signal 511 and outputs a feature amount estimation signal 501 as an estimation result to the feature amount display unit 60.

  Upon receiving the detection signal averaging end pulse 512, the memory selection unit 540 selects an output destination of the averaged detection signal 511 based on the mode switching signal 102. When the mode switching signal 102 indicates the period A, the reference detection signal memory 560 is selected as the output destination of the averaged detection signal 511, and the averaged detection signal 542 after selection is output. When the mode switching signal 102 indicates the period B, the comparison detection signal memory 550 is selected as the output destination of the averaged detection signal 511, and the averaged detection signal 541 after selection is output.

  The comparison detection signal memory 550 outputs a comparison detection signal storage completion signal 551 to the waveform comparison unit 570 when the data storage in the memory using the averaged detection signal 541 after selection is completed. When the comparison detection signal memory 550 receives the comparison memory read signal 571, the comparison detection signal memory 550 outputs the data stored in the memory to the waveform comparison unit 570 as the comparison detection signal 552 to be compared.

  The reference detection signal memory 560 outputs a reference detection signal storage completion signal 561 to the waveform comparison unit 570 when data storage in the memory using the average selection detection signal 542 after selection is completed. When the reference detection signal memory 560 receives the reference memory read signal 572, the reference detection signal memory 560 outputs the data stored in the memory to the waveform comparison unit 570 as a reference detection signal 562 that serves as a comparison reference.

  When the waveform comparison means 570 receives the comparison detection signal storage completion signal 551, the waveform comparison means 570 outputs the comparison memory read signal 571 to the comparison detection signal memory 550, and upon receiving the reference detection signal storage completion signal 561, the reference detection signal memory. A reference memory read signal 572 is output to 560. Then, the comparison detection signal 552 read from each of the memories 550 and 560 is compared with the reference detection signal 562 to determine whether the waveform shapes match, and a waveform comparison signal 502 indicating the comparison result is output to the control means 10. .

  Here, the usage example of the noninvasive blood glucose measuring device 1 of the present invention will be described.

  FIG. 1 is a diagram showing an example of use of the non-invasive blood sugar device 1 of the present invention.

  First, the noninvasive blood glucose measurement device 1 is mounted so as to be in contact with the surface of the living body 30 such as a patient's arm. Thereafter, the patient activates a blood sugar level measurement start switch (not shown) provided in the non-invasive blood sugar measuring device 1, and the light source 20, the photoacoustic detection means 40, and the feature amount estimation means 50 are stably operated by the control means 10. The light source 20 is controlled at the timing of entering.

  When the surface of the living body 30 is irradiated with pulsed light 201 from the light source 20, the pulsed light 201 propagates through the living body 30 and is absorbed by a specific substance in the blood vessel 31, for example, glucose, thereby generating a photoacoustic wave signal 301. Is done. The generated photoacoustic wave signal 301 is detected by the photoacoustic detection means 40 in the non-invasive blood glucose measurement device 1, and the blood glucose level is estimated from the detection result.

  In the first embodiment, one measurement interval is a measurement unit (0.1 second), and the interval until blood glucose level is estimated is a measurement basic cycle (10 seconds), and is repeated 25 times every measurement cycle (5 minutes). The blood glucose level is continuously measured.

  Hereinafter, the operation in the case where the noninvasive blood glucose measurement device 1 of the first embodiment performs continuous blood glucose level measurement will be described with reference to FIGS. 2, 3, and 4.

  FIG. 4 is a timing chart in which the operation of the noninvasive blood sugar measurement device 1 according to the first embodiment is plotted on the time axis. 4A shows a feature amount estimation signal 501, FIG. 4B shows a mode switching signal 102, and FIG. 4C shows a measurement cycle signal 101.

  Note that the initial state of the mode switching signal 102 shown in FIG. Further, a period when the mode switching signal 102 is Hi is a period A, and a period when the mode switching signal 102 is Lo is a period B.

  At time 0 minutes, a pulse of the measurement cycle signal 101 shown in FIG. 4C is generated, and the blood sugar level is estimated. Here, the blood glucose level is estimated once for one pulse of the signal in FIG.

  At time 5 minutes, a pulse of the measurement cycle signal 101 shown in FIG. 4C is generated, and the blood sugar level is estimated. Thereafter, the blood glucose level is repeatedly estimated at intervals of 5 minutes up to 35 minutes.

  In a period A from time 0 minutes to time 35 minutes, the blood sugar level is estimated eight times by the feature amount estimation means 50 and the reference detection signal 562 is generated. The creation of the reference detection signal 562 will be described later.

  At time 38 minutes, the mode switching signal 102 shown in FIG. 4B changes from Hi to Lo (T40), and the period B starts.

  At time 40 minutes, a pulse of the measurement cycle signal 101 shown in FIG. 4C is generated, the blood sugar level is estimated, and the estimation result is used as a comparison detection signal 552. 3 compares the comparison detection signal 552 with the shape of the reference detection signal 562 created in the period A to create the waveform comparison signal 502 and outputs the waveform comparison signal 502 to the control means 10.

  When the waveform comparison signal 502 generates a coincidence pulse, the control unit 10 in FIG. 2 determines that the measurement performed at time 40 minutes has been completed normally, and waits for the next measurement period of time 45 minutes. To do. At time 45 minutes, a pulse of the measurement cycle signal 101 shown in FIG. 4C is generated, the blood glucose level is estimated, and the comparison detection signal 552 as the estimation result and the reference detection created in the period A The waveform comparison signal 502 is generated by comparing the shapes of the signals 562 and output to the control means 10. Thereafter, a waveform comparison signal 502 is created for each measurement, and when a pulse indicating that the waveform comparison signal 502 is coincident is generated, measurement is repeated every 5 minutes, which is the measurement cycle.

  On the other hand, when the waveform comparison signal 502 indicates a mismatch, that is, when no pulse is generated, the pulse of the measurement cycle signal 101 is generated again in the same measurement cycle, and a series of comparison operations are performed again.

  Here, when determining the measurement error, it is desirable to set the number of consecutive times that the waveform results do not coincide again by a register or the like. In this case, the number of times the waveform results measured again are continuously mismatched is counted, and when the count value reaches the measurement abnormality determination setting value set in advance by the register, it is determined that there is a measurement error.

  Next, the detailed operation of the feature amount estimation means 50 for one measurement will be described with reference to FIGS.

  FIG. 5 is a timing chart showing in detail the operation of the feature quantity estimation means 50 for one measurement period. In FIG. 5, (c) is a measurement period signal 101, (d) is an irradiation period signal 103, (e) is an irradiation control signal 104, (f) is a detection signal averaging end pulse 512, and (g) is a reference detection. The signal storage completion signal 561 or the comparison detection signal storage completion signal 551 (h) is the waveform comparison signal 502. In addition, each signal of FIG.5 (c), (d), (e), (f), (g), (h) is cleared to Lo level at the start of a measurement period.

  First, when the pulse of the measurement period signal 101 shown in FIG. 5C is generated (T50), the irradiation period signal 103 shown in FIG. 5D is changed to Hi (T51), and the repetition irradiation period of the pulsed light 201 is changed. Start.

  While the irradiation period signal 103 shown in FIG. 5D is Hi (T51-T52), the irradiation control signal 104 shown in FIG. 5E repeatedly generates pulses at intervals of the measurement units. The pulsed light 201 is irradiated for each generated pulse, and detection signals 401 corresponding to the number of times of irradiation are obtained. In the first embodiment, 60 pulses of the irradiation control signal 104 are generated while the irradiation period signal 103 is Hi.

  When the irradiation period signal 103 shown in FIG. 5D changes to Lo and the repetition irradiation period of the pulsed light 201 ends (T52), the averaging means 510 causes the detection signal 401 corresponding to the number of irradiation times of the pulsed light 201 to be detected. Start the averaging operation.

  Here, the averaging means 510 includes an irradiation number counter, a detection signal averaging memory, an adder, and a divider (not shown). The irradiation number counter is reset to 0 when the irradiation period signal 103 rises, and is incremented by one every time the irradiation control signal 104 is generated. The detection signal averaging memory is reset to 0 when the irradiation period signal 103 is switched from Lo to Hi. The adder synchronizes each irradiation of the pulsed light 201 with the pulse of the irradiation control signal 104 as a trigger, and then stores the sum of the detection signals 401 in the detection signal averaging memory. When the irradiation period signal 103 is switched from Hi to Lo, the divider divides the sum of the detection signals 401 stored in the detection signal averaging memory by the value indicated by the irradiation number counter to calculate an average detection signal 511. To do.

  When the averaging operation of the detection signal 401 by the averaging means 510 is completed, a detection signal averaging end pulse 512 shown in FIG. 5 (f) is generated (T53), and the average detection signal 511 is stored in the memory. Is called. The average detection signal 511 is stored in the reference detection signal memory 560 when the mode switching signal 102 shown in FIG. 4B indicates the period A, and when the mode switching signal 102 indicates the period B. It is stored in the comparison detection signal memory 550.

  When the data storage to each memory is completed, the reference detection signal storage completion signal 561 or the comparison detection signal storage completion signal 551 shown in FIG. 5G is generated (T54). When both the reference detection signal storage completion signal 561 and the comparison detection signal storage completion signal 551 are notified to the waveform comparison unit 570, the reference detection signal 562 and the comparison detection signal 552 are read from the memories 560 and 550. Then, the waveform shapes of the two detection signals are compared, and a waveform comparison signal 502 indicating the comparison result is output to the control means 10.

  When the waveform comparison signal 502 generates a pulse as shown in FIG. 5 (h) (T55), it is determined that the shapes of the reference detection signal 562 and the comparison detection signal 552 match, and the estimation means 520 estimates the waveform. Displays blood glucose level. On the other hand, if no pulse is generated, it is determined that an abnormality has occurred during measurement, a measurement error is displayed, and the user is notified.

  The mode switching signal 102 is switched during the period from the end of the measurement basic cycle to the start of the next measurement cycle (T55 to T56).

  Here, the creation of the reference detection signal 562 will be described.

  When the number of blood sugar levels estimated during the period when the mode switching signal 102 is Hi is a plurality of times, a reference detection signal 562 is created by further averaging the plurality of averaged detection signals 511, and the mode After the switching signal 102 is switched from Hi to Lo, the reference detection signal storage completion signal 561 is output to the waveform comparison unit 570. The averaging of the plurality of averaged detection signals 511 may be performed in the same manner as the averaging of the detection signals 401 by the averaging means 510 described above.

  In addition, as another method for creating the reference detection signal 562, for example, the averaged detection signal 511 at the time of the last measurement in the period A is detected for reference without averaging the plurality of averaged detection signals 511 described above. An averaged detection signal 511 at a specific measurement, such as the signal 562, can be used as the reference detection signal 562. At this time, it is also possible to arbitrarily set which blood glucose level estimation averaged detection signal 511 is to be used by a register or the like.

  In addition, when the number of blood sugar levels estimated during the period when the mode switching signal 102 is Hi is 1, the obtained average detection signal 511 is used as the reference detection signal 562.

  Hereinafter, a method of comparing the reference detection signal 562 and the comparison detection signal 552 in the waveform comparison unit 570 will be described.

  In comparison between the reference detection signal 562 and the comparison detection signal 552, the reference detection signal 562 and the comparison detection signal 552 are synchronized with the irradiation timing of the pulsed light 201 by the irradiation control signal 104 as a reference position.

1) First Comparison Method Hereinafter, a first comparison method of the reference detection signal 562 and the comparison detection signal 552 will be described with reference to FIG.

  6A shows a reference detection signal 562, and FIG. 6B shows a comparison detection signal 552. In the figure, the horizontal axis indicates the sampling direction, and the vertical axis indicates the amplitude level.

  In the explanatory diagrams of the reference detection signal 562 or the comparison detection signal 552 after this description, the expression “specific” “position” is used, but this indicates a specific location in the horizontal axis direction in the explanatory waveform diagram. Show.

  (Step 1) The maximum peak position T61 at which the absolute value of the amplitude level of the reference detection signal 562 is maximized is detected.

  (Step 2) Compare the position at which the polarity of the slope of the reference detection signal 562 changes at a position closest to the maximum peak position T61 with respect to both the positive and negative directions (horizontal axis direction in the figure) from the maximum peak position T61. It is detected as a start position T60 and a comparison end position T62.

  (Step 3) In the comparison period from T60 to T62, the number of locations where the polarity of the slope of the comparison detection signal 552 changes (the number of polarity changes) is counted. In FIG. 6B, the places where the polarity changes are indicated by circles.

  (Step 4) When the number of polarity changes is 0 or 2 or more, the waveform comparison signal 502 becomes 0 indicating inconsistency. On the other hand, when the number of polarity changes is 1, the waveform comparison signal 502 is 1 indicating coincidence.

  As described above, it is possible to detect whether or not an abnormality has occurred during measurement by comparing the shapes of the reference detection signal 562 and the comparison detection signal 552 using the method of Step 1 to Step 4.

  It should be noted that between step 2 and step 3, the comparison detection signal 552 may be subjected to a low-pass filtering process for cutting the high frequency in the sampling direction. In this case, it is possible to prevent erroneous detection of the number of changes in the polarity of inclination due to spike-like noise.

2) Second Comparison Method Hereinafter, a second comparison method of the reference detection signal 562 and the comparison detection signal 552 will be described with reference to FIG.

  7A shows the reference detection signal 562, FIG. 7B shows the comparison detection signal 552, and FIG. 7C shows the reference detection signal 562 and the comparison detection signal 552 with normalized amplitude levels. FIG. In the figure, the horizontal axis indicates the sampling direction, and the vertical axis indicates the amplitude level.

  L7A and L7B indicate levels at which the absolute value of the amplitude in each signal is maximized.

  T70 is a comparison start position where the comparison between the reference detection signal 562 and the comparison detection signal 552 is started, and T71 is a comparison end position where the comparison between the reference detection signal 562 and the comparison detection signal 552 is ended. Specifically, after the pulse of the irradiation control signal 104 is generated, T70 is set at a sampling position before the timing at which the photoacoustic wave signal 301 propagates through the living body 30 and reaches the photoacoustic detection means 40 on the surface of the living body 30. And T71 may be set at the timing at which the change of the photoacoustic wave signal 301 reaching the photoacoustic detection means 40 converges after T70.

  T70, T71, and the second threshold value, which is a judgment reference for waveform shape comparison, are set in advance by a register (not shown).

  (Step 1) In the comparison period from T70 to T71, the maximum amplitude value L7B of the comparison detection signal 552 shown in FIG. 7B is the same as the maximum amplitude value L7A of the reference detection signal 562 shown in FIG. Normalization is performed in the amplitude level direction so as to be equal, and the waveform of FIG. 7C is calculated.

  (Step 2) During the comparison period from T70 to T71, the square error between the amplitude level of the reference detection signal 562 and the normalized amplitude level of the comparison detection signal 552 shown in FIG. 7C is calculated for each sampling. To do.

  (Step 3) The total amount of square errors in the comparison period from T70 to T71 is calculated.

  (Step 4) The total sum of the square errors is compared with the second threshold, and if the sum of the square errors is less than or equal to the second threshold, it is determined that they match. It is judged that.

  As described above, it is possible to detect whether or not an abnormality has occurred during measurement by comparing the shapes of the reference detection signal 562 and the comparison detection signal 552 using the method of Step 1 to Step 4.

  For example, the second threshold value may be set as a set value by searching for a level at which the measurement abnormality can be detected most efficiently in the adjustment process at the time of shipment of the noninvasive blood glucose measurement device 1.

  The second threshold value can also be set to an arbitrary value from the outside after shipment from the factory using a register.

3) Third Comparison Method Hereinafter, a third comparison method of the reference detection signal 562 and the comparison detection signal 552 will be described with reference to FIG.

  8A shows the reference detection signal 562, and FIG. 8B shows the comparison detection signal 552. In the figure, the horizontal axis indicates the sampling direction, and the vertical axis indicates the amplitude level.

  T80 is a comparison start position where the comparison between the reference detection signal 562 and the comparison detection signal 552 is started, and T81 is a comparison end position where the comparison between the reference detection signal 562 and the comparison detection signal 552 is ended. Setting is performed in the same manner as T70 and T71.

  T80, T81, and the third threshold value, which is a criterion for waveform shape comparison, are set in advance by a register (not shown).

  (Step 1) In the comparison period from T80 to T81, the average value (average value A) of the amplitude levels of the reference detection signal 562 and the average value (average value B) of the comparison detection signal 552 are calculated.

  (Step 2) The difference between the reference detection signal 562 and the average value A (deviation A) and the difference between the comparison detection signal 552 and the average value B (deviation B) are calculated.

  (Step 3) After calculating the product sum of the deviation A and the deviation B in the comparison period from T80 to T81, the calculated product-sum result is divided by the total number of samplings (covariance AB).

  (Step 4) In the comparison period from T80 to T81, the sum of the squares of the deviation A and the deviation B is obtained, and then the variance A and variance B are calculated by dividing each sum result by the total number of samplings.

  (Step 5) The square roots of the variance A and the variance B are obtained, and the standard deviation A and the standard deviation B are calculated.

  (Step 6) A correlation coefficient is calculated by a calculation formula of correlation coefficient = covariance AB / (standard deviation A × standard deviation B).

  (Step 7) The correlation coefficient is compared with the third threshold value, and if the correlation coefficient is greater than the third threshold value, it is determined that they match, and if the correlation coefficient is equal to or less than the third threshold value, it is determined that they do not match. To do.

  As described above, it is possible to detect whether or not an abnormality has occurred during measurement by comparing the shapes of the reference detection signal 562 and the comparison detection signal 552 using the method of Step 1 to Step 7.

  The correlation coefficient obtained in step 6 is a statistic that measures the similarity between two signals. The correlation coefficient is a number between −1 and 1, and the closer to 1, it can be determined that the two signals are more similar. Therefore, if the third threshold value can be set to a value between −1 and 1 and set to a value close to 1, the accuracy of the comparison result is improved.

  The third threshold value, for example, in the adjustment process at the time of factory shipment of the noninvasive blood glucose measurement device 1 according to the first embodiment, is searched for a level at which measurement abnormality can be detected most efficiently, and the value is set as a set value. do it.

  The third threshold value can be set to an arbitrary value from the outside after shipment from the factory using a register.

4) Fourth Comparison Method Hereinafter, a fourth comparison method of the reference detection signal 562 and the comparison detection signal 552 will be described with reference to FIG.

  FIG. 9A shows the reference detection signal 562, and FIG. 9B shows the comparison detection signal 552. In the figure, the horizontal axis indicates the sampling direction, and the vertical axis indicates the amplitude level.

  L9A indicates the amplitude level set by the fourth threshold value.

  T90 is a comparison start position where the comparison between the reference detection signal 562 and the comparison detection signal 552 is started, and T91 is a comparison end position where the comparison between the reference detection signal 562 and the comparison detection signal 552 is ended. Setting is performed in the same manner as T70 and T71.

  T90, T91, the fourth threshold value, the fifth threshold value, and the sixth threshold value, which are judgment criteria for waveform shape comparison, are set in advance by a register (not shown).

  (Step 1) In the comparison period from T90 to T91, normalization is performed so that the maximum amplitude value of the reference detection signal 562 is equal to the maximum amplitude value of the comparison detection signal 552. The waveforms of the signals 562 and 552 after normalization are as shown in FIGS. 9A and 9B, respectively.

  (Step 2) In the comparison period from T90 to T91, the number of samplings in which the amplitude level of the reference detection signal 562 exceeds the level L9A set by the fourth threshold is counted, and the count result is CNTA.

  (Step 3) In the comparison period from T90 to T91, the number of samplings in which the amplitude level of the comparison detection signal 552 exceeds the level L9A set by the fourth threshold is counted, and the count result is CNTB.

  (Step 4) The difference between CNTA and CNTB is calculated.

  (Step 5) If the result of Step 4 is greater than or equal to the fifth threshold and less than or equal to the sixth threshold, it is determined that they match, and if the result is less than the fifth threshold or greater than the sixth threshold, they do not match Judge.

  As described above, it is possible to detect whether or not an abnormality has occurred during measurement by comparing the shapes of the reference detection signal 562 and the comparison detection signal 552 using the method of Step 1 to Step 5.

  Here, the accuracy of the comparison result improves as the difference between the set values of the fifth threshold value and the sixth threshold value decreases.

  The fourth threshold value, the fifth threshold value, and the sixth threshold value are searched for the level at which the measurement abnormality can be detected most efficiently in the adjustment process at the time of factory shipment of the noninvasive blood glucose measurement device 1, for example. The value may be a set value.

  Further, the fourth threshold value, the fifth threshold value, and the sixth threshold value can be set to arbitrary values from the outside after shipment from the factory using a register.

5) Fifth Comparison Method Hereinafter, a fifth comparison method of the reference detection signal 562 and the comparison detection signal 552 will be described with reference to FIG.

  10A shows a reference detection signal 562, and FIG. 10B shows a comparison detection signal 552. In the figure, the horizontal axis indicates the sampling direction, and the vertical axis indicates the inclination.

  T100 is a comparison start position where the comparison between the reference detection signal 562 and the comparison detection signal 552 is started, and T101 is a comparison end position where the comparison between the reference detection signal 562 and the comparison detection signal 552 is ended. Setting is performed in the same manner as T70 and T71.

  T100, T101, the seventh threshold value and the eighth threshold value, which are judgment criteria for waveform shape comparison, are set in advance by a register (not shown).

  (Step 1) In the comparison period from T100 to T101, the number of points where the polarity of the inclination of the reference detection signal 562 changes (the position indicated by the arrow in FIG. 10A) is counted, and the count result is defined as CNTC. .

  (Step 3) In the comparison period from T100 to T101, the number of points where the polarity of the inclination of the comparison detection signal 552 changes (the position indicated by the arrow in FIG. 10B) is counted, and the count result is CNTD. .

  (Step 4) The difference between CNTC and CNTD is calculated.

  (Step 5) If the result of Step 4 is greater than or equal to the seventh threshold and less than or equal to the eighth threshold, it is determined that they match, and if the result is less than the seventh threshold or greater than the eighth threshold, they do not match to decide.

  As described above, it is possible to detect whether or not an abnormality has occurred during measurement by comparing the shapes of the reference detection signal 562 and the comparison detection signal 552 using the method of Step 1 to Step 5.

  Note that before step 1, a filtering process that passes only a specific frequency band may be performed on the comparison detection signal 552 and the reference detection signal 562. In this case, it is possible to prevent erroneous detection of the number of changes in the polarity of inclination due to spike-like noise.

  Here, the accuracy of the comparison result improves as the difference between the set values of the seventh threshold value and the eighth threshold value decreases.

  The seventh threshold value and the eighth threshold value are searched for the level at which the measurement abnormality can be detected most efficiently in the adjustment process at the time of factory shipment of the noninvasive blood glucose measurement device 1 of the first embodiment, for example. The value may be a set value.

  Further, the seventh threshold value and the eighth threshold value can be set to arbitrary values from the outside after shipment from the factory using a register.

6) Sixth Comparison Method A sixth comparison method for the reference detection signal 562 and the comparison detection signal 552 will be described below with reference to FIG.

  11A shows the inclination of the reference detection signal 562, FIG. 11B shows the reference inclination detection signal, FIG. 11C shows the reference inclination detection signal with a margin, and FIG. ) Shows the inclination of the comparison detection signal 552, FIG. 11 (e) shows a comparison inclination detection signal, FIG. 11 (f) shows a reference inclination detection signal having a margin shown in FIG. FIG. 6 is a diagram illustrating a timing at which the inclination of the comparison detection signal 552 becomes large at a timing at which the inclination of the reference detection signal 562 is small, using the comparison inclination detection signal shown in FIG.

  In FIG. 11C, the hatched portion indicates the margin period in the time axis direction with respect to the change point of the reference inclination detection signal in FIG. 11A and 11D, the horizontal axis indicates the sampling direction, and the vertical axis indicates the inclination. In FIGS. 11B, 11C, 11E, and 11F, the horizontal axis indicates the sampling direction and the vertical direction. The axis indicates the level obtained by binarizing the amount of inclination with L11A. Here, L11A is the amount of inclination set by the ninth threshold value.

  T110 is a comparison start position where the comparison between the reference detection signal 562 and the comparison detection signal 552 is started, and T111 is a comparison end position where the comparison between the reference detection signal 562 and the comparison detection signal 552 is ended. Setting is performed in the same manner as T70 and T71.

  The ninth threshold value for T110, T111, and L11A is set in advance by a register (not shown).

  (Step 1) The inclination of the reference detection signal 562 is detected for each sampling, and the signal shown in FIG.

  (Step 2) As shown in FIG. 11B, a reference inclination detection signal indicating Lo if the inclination of the reference detection signal 562 is within the range of ± L11A, and Hi if the inclination of the reference detection signal 562 exceeds the range of ± L11A. create.

  (Step 3) As shown in FIG. 11 (c), a signal having a margin (shaded portion) is created with respect to the change point of the reference inclination detection signal in FIG. 11 (b).

  (Step 4) The inclination of the comparison detection signal 552 is detected for each sampling, and the signal shown in FIG.

  (Step 5) As shown in FIG. 11E, a comparison inclination detection signal indicating Lo if the inclination of the comparison detection signal 552 is within the range of ± L11A, and Hi if the inclination of the comparison detection signal 552 exceeds the range of ± L11A. create.

  (Step 6) In the comparison period from T110 to T111, the reference inclination detection signal having the margin of FIG. 11C is compared with the comparison inclination detection signal of FIG. When the period in which Lo becomes Lo and the period in which the comparative inclination detection signal becomes Lo are the same, it is determined that they match, and as shown by a circle in FIG. 11 (f), In the period when the reference inclination detection signal having a margin is Lo, the period when the comparison inclination detection signal of FIG. 11E becomes Hi, that is, the change amount of the reference detection signal 562 is small. If there is a timing at which the amount of change in the comparison detection signal 552 becomes large at the timing, it is determined that there is a mismatch.

  As described above, it is possible to detect whether or not an abnormality has occurred during measurement by comparing the shapes of the reference detection signal 562 and the comparison detection signal 552 using the method of Step 1 to Step 6.

  In step 3, the margin amount indicated by the hatched portion is set to an arbitrary amount in advance by a register.

  Further, low-pass filtering processing for cutting the high frequency in the sampling direction may be performed on the signals obtained in Step 1 and Step 3 (FIGS. 11A and 11C). In this case, it is possible to prevent erroneous detection of the number of changes in the polarity of inclination due to spike-like noise.

  Further, the ninth threshold value is set, for example, by searching for a level at which measurement abnormality can be detected most efficiently in the adjustment process at the time of factory shipment of the noninvasive blood glucose measurement device 1 according to Embodiment 1 of the present invention. It can be a value.

  Further, the ninth threshold value can be set to an arbitrary value from the outside after shipment from the factory using a register.

7) Seventh Comparison Method A seventh comparison method of the reference detection signal 562 and the comparison detection signal 552 will be described below with reference to FIG.

  12A shows a reference detection signal 562 after frequency axis conversion, and FIG. 12B shows a comparison detection signal 552 after frequency axis conversion. In the figure, the horizontal axis represents frequency and the vertical axis represents signal power.

  F12A, F12B, F12C, F12D, and F12E indicate comparison target frequencies, and L12A indicates a threshold value of the signal power amount set by the tenth threshold value.

  The position where the comparison between the reference detection signal 562 and the comparison detection signal 552 is started is T120, the position where the comparison between the reference detection signal 562 and the comparison detection signal 552 is finished is T121, and the same as T70 and T71 described above. Set by the method. Note that since the diagram shown in FIG. 12 is a frequency-axis signal, T120 and T121, which are time parameters, are not shown.

  The tenth threshold value for T120, T121, and L12A is set in advance by a register (not shown).

  (Step 1) In the comparison period from T120 to T121, the reference detection signal 562 is Fourier-transformed to calculate the frequency axis converted reference detection signal 562 shown in FIG.

  (Step 2) In the comparison period from T120 to T121, the comparison detection signal 552 is Fourier-transformed to calculate the frequency axis converted comparison detection signal 552 shown in FIG.

  (Step 3) In the reference detection signal 562 after frequency axis conversion, a comparison target frequency exceeding the level of L12A is detected. Here, F12A is detected.

  (Step 4) In the comparison detection signal 552 after frequency axis conversion, a comparison target frequency exceeding the level of L12A is detected. Here, F12A and F12C are detected.

  (Step 5) The comparison target frequencies detected in Step 3 and Step 4 are compared, and if there is no difference in frequency components, it is determined that they match, and if there is a difference, it is determined that they do not match. For example, in FIG. 12, since F12C is detected only in the comparison signal 552 after frequency axis conversion, it is determined that there is a mismatch.

  As described above, it is possible to detect whether or not an abnormality has occurred during measurement by comparing the shapes of the reference detection signal 562 and the comparison detection signal 552 using the method of Step 1 to Step 5.

  The tenth threshold value, for example, in the adjustment process at the time of factory shipment of the noninvasive blood glucose measurement device 1 in the first embodiment, is searched for a level at which measurement abnormality can be detected most efficiently, and the value is set as a set value. do it.

  The tenth threshold value can also be set to an arbitrary value from the outside after shipment from the factory using a register.

  As described above, as the comparison method by the waveform comparison unit 570, any method of the first to seventh comparison methods described above may be used. In addition, it is possible to use any combination of the first to seventh comparison methods described above.

  Further, means for learning the second to tenth threshold values, which are the judgment criteria for waveform shape comparison, every time the blood sugar level is estimated is provided inside the non-invasive blood sugar measuring device 1 according to the first embodiment, and adaptively. It is also possible to change it.

  In the first embodiment, the waveform comparison means 570 does not compare the waveform shape in the period A, and in the period B, the reference detection signal 562 obtained in the period A and the comparison detection signal 552 are compared. However, the comparison timing is not necessarily limited to this. For example, the reference estimated number setting value is set to 3 times in advance in the register, and the averaged detection signal 511 measured 3 times before every measurement is reflected in the reference detection signal 562, and the comparison detection signal 552 May be compared.

  In the first embodiment, the mode switching signal 102 is described as a 1-bit signal. However, this mode may be composed of a multi-bit signal. When configured with a multi-bit signal, the period that can be switched by the mode switching signal 102 is not limited to two periods, period A and period B. For example, after period A, the period can be shifted to period C, and then can be shifted to period B. It is.

  In this case, for example, the period A in which the reference detection signal 562 is created, the period B in which the comparison detection signal 552 is created, and the period C in which neither the reference detection signal 562 nor the comparison detection signal 552 is created are switched. In addition, the order in which this period moves can be arbitrarily set.

  Further, the period A is set to, for example, the calibration mode, and in addition to the generation of the reference detection signal 562, the feature amount is calculated using the feature amount estimation signal 501 and the target value input from the outside of the noninvasive blood glucose measurement device 1. It is also possible to calculate a calibration condition for calibrating the estimated signal 501.

  It is also possible to set the period B to, for example, the normal measurement mode, and calibrate the feature quantity estimation signal 501 using the calibration conditions obtained in the period A in addition to the generation of the comparison detection signal 552. .

Thus, according to the noninvasive blood glucose measurement device of the first embodiment, the control means 10 that controls the irradiation of the pulsed light 201 repeated at least once from the light source 20 with respect to one blood glucose level estimation. Then, the photoacoustic wave signal 301 generated by glucose in the living body 30 absorbing light energy is detected by the pulsed light 201 irradiated on the surface of the living body 30, and a detection signal 401 sampled at a predetermined frequency is output. A photoacoustic detection means 40, a feature quantity estimation means 50 for averaging the detection signals 401 for the number of times of irradiation of the pulsed light 201 and estimating the blood glucose level using the averaged detection signal 511, and the estimated blood glucose level Characteristic amount display means 60 for displaying the averaged detection signal 511 obtained every time the blood glucose level is estimated at least once in the period A. One signal, or a plurality of averaged detection signals 511 having the same waveform shape, or a signal obtained by further averaging any number of averaged detection signals 511 among them, is referred to as a reference detection signal 562, and In B, the comparison detection signal 552 consisting of the averaged detection signal 511 obtained at least once every time the blood glucose level is estimated is compared with the reference detection signal 562, and when both the detection signals match, If it is determined that the measurement is performed normally and does not match, it is determined that an abnormality has occurred during the measurement, so it is possible to detect changes in the measurement environment for blood glucose levels, and feature quantities It is possible to prevent an accident such as an overdose of insulin by the patient against an erroneous blood glucose level displayed on the display means 60.
(Embodiment 2)

  The noninvasive blood sugar measurement device according to Embodiment 2 of the present invention will be described below.

  The non-invasive blood sugar measurement device according to the second embodiment includes a feature quantity estimation unit 51 instead of the feature quantity estimation unit 50 of FIG. 2, and the mode switching signal 102 input from the control unit 10 is used. do not do.

  FIG. 13 is a block diagram showing a configuration of the feature amount estimation means 51 in the noninvasive blood sugar measurement device according to the second embodiment. In the figure, the same components as those in FIG.

  13 includes a memory selection unit 543 and a reference detection signal memory 563 instead of the memory selection unit 540 and the reference detection signal 560 in the characteristic amount estimation unit 50 of FIG. It is.

  The memory selection unit 543 selects an output destination of the averaged detection signal 511 using the detection signal averaging end pulse 512. When the comparison detection signal memory 550 is selected as the output destination of the averaged detection signal 511, the averaged detection signal 541 after selection is output, and when the reference detection signal memory 563 is selected, the averaged detection signal 542 after selection is output. To do.

  Here, a method of determining the output destination of the average detection signal 511 in the memory selection unit 543 will be described.

  The memory selection unit 543 internally includes a counter that is reset to 0 when the pulse interval of the measurement cycle signal 101 is longer than the measurement basic cycle described later and shorter than the blood glucose level measurement interval, for example, 30 seconds. The counter counts up each time a pulse of the measurement period signal 101 is input. If the count result by the counter is 1, that is, the first measurement timing T130, the output destination of the averaged detection signal 511 is set in the reference detection signal memory 563. When the count result is 2 to 5, that is, the timing from T131 to T134, the output destination of the averaged detection signal 511 is set in the comparison detection signal memory 550.

  The reference detection signal memory 563 outputs the reference detection signal storage completion signal 561 to the waveform comparison unit 570 when the data storage in the memory using the average selection detection signal 542 after selection is completed. Further, when the reference detection signal memory 563 receives the reference memory read signal 572, the reference detection signal memory 563 outputs the data stored in the memory to the waveform comparison unit 570 as the reference detection signal 562.

  Hereinafter, the operation in the case where the noninvasive blood glucose measurement device 1 according to the second embodiment performs continuous blood glucose measurement will be described with reference to FIGS. 13 and 14.

  FIG. 14 is a timing chart in which the operation of the noninvasive blood glucose measurement device 1 according to the second embodiment is plotted on the time axis. 14A shows the feature amount estimation signal 501, and FIG. 14C shows the measurement period signal 101.

  In the second embodiment, in the repeated measurement, the blood glucose level is continuously measured repeatedly with the interval until the blood glucose level is estimated as the measurement basic cycle (10 seconds). The blood glucose level is measured for 120 minutes.

  At time 0 minutes, a pulse of the measurement cycle signal 101 shown in FIG. 14C is generated, and the blood glucose level is estimated (T130). Here, the blood sugar level is estimated once for one pulse of the signal in FIG. At this time, the averaged detection signal 511 is stored in the reference detection signal memory 563, and the reference detection signal 562 is created.

  At time 0 minutes 10 seconds, a pulse of the measurement period signal 101 shown in FIG. 14C is generated, and the blood sugar level is estimated (T131). At this time, the averaged detection signal 511 is stored in the comparison detection signal memory 550, and the comparison detection signal 552 is created. The waveform comparison means 570 compares the shapes of the comparison detection means 552 and the reference detection signal 562 and outputs a waveform comparison signal 502 indicating the comparison result to the control means 10.

  Between time 0 minute 20 seconds and time 0 minutes 40 seconds, a pulse of the measurement cycle signal 101 shown in FIG. 14C is generated every 10 seconds to estimate the blood glucose level. A comparison detection signal 552 is created from the averaged detection signal 511 obtained during this period, the comparison detection signal 552 is compared with the reference detection signal 562 for each measurement, and a waveform comparison signal 502 indicating the comparison result is controlled. 10 (T132-T134).

  At time 0 minute 40 second, a total of four waveform comparison signals 502 are input to the control means 10, but the control means 10 counts the number of times that the waveform comparison signals 502 are determined to match, When the count result is equal to or greater than the first threshold, a pulse of the measurement cycle signal 101 shown in FIG. 14C is generated when the next non-invasive blood glucose measurement device 1 has a basic measurement cycle of time 5 minutes, Value estimation is performed (T135). At this time, continuous blood glucose level estimation is not performed every 10 seconds as in T130 to T134. Then, the measured averaged detection signal 511 is compared as a comparison detection signal 552 with a reference detection signal 562 obtained at the time of measurement for 0 minute, and a match or mismatch is determined. In subsequent measurements, blood glucose levels are estimated periodically at 5-minute intervals.

  When the count result of the number of times of matching among the four waveform comparison signals 502 is less than the first threshold in the time 0 minute 40 second, the time from 0 minute 50 second to 1 minute 30 second During this period, continuous blood glucose level estimation is performed every 10 seconds, such as T130 to T134.

  When the count result of the number of times of matching among the four waveform comparison signals 502 at the time of 1 minute 30 is equal to or more than the first threshold, the next measurement is performed at the time of 5 minutes, and thereafter every 5 minutes. The blood sugar level is estimated.

  If the count result of the number of times of matching among the four waveform comparison signals 502 becomes less than the first threshold again at time 1 minute 30, the counter provided in the memory selection unit 543 is reset. After waiting for the time required for the operation (30 seconds), continuous blood glucose level estimation is performed again every 10 seconds, such as T130 to T134.

  Here, regarding the determination of the re-continuous measurement in the control means 10, an arbitrary number of times is set in the register, and the operation of the noninvasive blood glucose measuring device 1 is stopped when the continuous measurement is repeated a predetermined number of times. Then, processing such as notifying the user of a measurement error using the feature amount display means 60 is performed.

  As described above, the operation of the non-invasive blood sugar measurement device 1 according to the second embodiment has been described. Here, in the continuous blood sugar level estimation from T130 to T134, the timing for creating the reference detection signal 562 is not limited to T130. Absent. For example, the timing for creating the reference detection signal 562 using a register (not shown) is set as T132, the reference detection signal 562 is created at the timing of T132, and other timings (T130, T131, T133, T134). Thus, the comparison detection signal 552 may be generated.

  In addition, after the continuous blood glucose level estimation from T130 to T134 is performed, if the count result of the number of times determined to be equal is less than the first threshold, the continuous blood glucose level estimation is not performed immediately thereafter, and is set in the register. After waiting for an arbitrary time, continuous blood glucose level estimation may be performed again.

  Moreover, in this Embodiment 2, although the timing which starts continuous blood glucose level estimation was time 0 minutes, this start timing is not restricted to this, It can set to arbitrary timings.

  Moreover, in this Embodiment 2, after performing continuous blood glucose level estimation from T130 to T134, when the count result of the number of times determined to be coincident is less than the first threshold, remeasurement is performed. At this time, one reference detection signal 562 is created from the averaged detection signal 511 at the timing determined as coincidence from T130 to T134 without making a determination of remeasurement, and a waveform at the time of subsequent blood sugar level estimation It is also possible to make it a shape comparison target.

  Further, in the second embodiment, the case where the averaged detection signal 511 obtained at the timing of T130 is used as the reference detection signal 562 has been described, but a plurality of values obtained in continuous blood glucose level estimation from T130 to T134 are described. When the number of signals that match when the shapes of the averaged detection signals 511 are equal to or greater than the first threshold value, one averaged detection signal among a plurality of averaged detection signals that match the shapes, or A reference detection signal 562 may be a signal obtained by further averaging a plurality of averaged detection signals having the same waveform shape or an arbitrary number of averaged detection signals among them.

  In the second embodiment, the measurement interval is 10 seconds and the measurement count is 5 in continuous blood glucose level estimation. However, the measurement interval and the measurement count are not limited to this, and any interval and count Can be set.

  Thus, according to the non-invasive blood glucose measurement device of the second embodiment, a part of the period (T130) within the predetermined time interval in which the blood glucose level is estimated at least once at the predetermined time interval. To T134), a plurality of averaged detection signals 511 are created by performing continuous blood sugar level estimation every 10 seconds, and among the plurality of averaged detection signals 511, the number of signals having the same waveform shape is the first. Is equal to or greater than the threshold value, the averaged detection signal 511 at T130 is used as the reference detection signal 562, and after T135, the detection for comparison consisting of the averaged detection signal 511 obtained by estimating the blood glucose level at predetermined time intervals. The signal 552 is compared with the reference detection signal 562. If the waveform shapes of the two detection signals match, it is determined that the measurement is performed normally. If they do not match, there is an abnormality during the measurement. Since it is determined that the error has occurred, a measurement abnormality can be detected efficiently, and the measurement accuracy can be improved.

(Embodiment 3)
The noninvasive blood sugar measurement device according to Embodiment 3 of the present invention will be described below.

  The non-invasive blood sugar measurement device according to the third embodiment is provided with a feature quantity estimation means 52 instead of the feature quantity estimation means 50 of FIG. 2, and the mode switching signal 102 input from the control means 10 is used. do not do.

  FIG. 15 is a block diagram illustrating a configuration of the feature amount estimation unit 52 in the noninvasive blood sugar measurement device according to the third embodiment. In the figure, the same components as those in FIG. 13 are denoted by the same reference numerals.

  The feature quantity estimation means 52 shown in FIG. 15 includes an estimation means 521 and a memory selection means 544 in place of the estimation means 521 and the memory selection means 543 in the feature quantity estimation means 51 of FIG.

  The estimation means 521 estimates the blood glucose level using the averaged detection signal 511 and the waveform comparison signal 502, and outputs a feature quantity estimation signal 501 as an estimation result to the feature quantity display means 60.

  The memory selection unit 544 includes a counter that resets for each pulse of the measurement cycle signal 101, counts up each time the detection signal averaging end pulse 512 is input, and performs average detection based on the count value. The output destination of the signal 511 is selected. When the comparison detection signal memory 550 is selected as the output destination of the averaged detection signal 511, the averaged detection signal 541 after selection is output, and when the reference detection signal memory 563 is selected, the averaged detection signal 542 after selection is output. To do.

  Hereinafter, the operation in the case where the noninvasive blood glucose measurement device 1 according to the third embodiment performs continuous blood glucose measurement will be described with reference to FIGS. 15 and 16.

  FIG. 16 is a timing chart in which the operation of the noninvasive blood glucose measurement device 1 according to the third embodiment is plotted on the time axis. In FIG. 16, (a) is a feature amount estimation signal 501, (c) is a measurement period signal 101, (d) is an irradiation period signal 103, (i) is an averaging unit signal in the averaging means 510, (e). Indicates an irradiation control signal 104. Note that at the start of the measurement cycle, each signal in FIGS. 16D and 16E is cleared to the Lo level, and the signal in FIG. Also, the averaging unit signal in FIG. 16 (i) is a signal controlled by a count result by a counter (not shown) in the averaging means 510.

  In the repeated measurement, the blood glucose level is measured for 120 minutes every measurement cycle (5 minutes) with one measurement interval as the measurement unit (0.1 second).

  In the noninvasive blood glucose measurement device 1 according to the third embodiment, the comparison between the comparison detection signal 552 and the reference detection signal 562 is completed within one measurement cycle.

  In the noninvasive blood glucose measurement device 1 according to the third embodiment, a coincidence comparison threshold value for controlling the measurement cycle signal 101 is set in advance in a register (not shown) twice.

  For example, the following operation is performed at time 0 minutes.

  When the pulse of the measurement period signal 101 shown in FIG. 16C is generated (T150), the irradiation period signal 103 shown in FIG. 16D becomes Hi (T151). At this time, the averaging unit signal shown in FIG.

  While the irradiation period signal 103 is Hi (T151 to T155), the pulse of the irradiation control signal 104 shown in FIG. 16E is generated for each measurement unit. The pulse of the irradiation control signal 104 is generated 60 times in total.

  After the fifteenth irradiation control signal 104 pulse is generated, the averaging unit signal shown in FIG. 16 (i) becomes 2 (T152). Here, TA is a period from T151 to T152.

  Then, the averaging means 510 averages the detection signals 401 for 15 times in the TA period, and outputs the averaged detection signal 511 to the estimation means 521 and the memory selection means 544, and also detects the detection signal averaging end pulse 512. Is output to the memory selection means 544.

  When the first detection signal averaging end pulse 512 is input to the memory selection unit 544, the averaged detection signal 511 is stored in the reference detection signal memory 563 as the reference detection signal 562.

  After the TA period, the 16th to 30th pulses of the irradiation control signal 104 are generated (T152 to T153). After the 30th pulse of the irradiation control signal 104 is generated, the averaging unit signal shown in FIG. 16 (i) becomes 3 (T153). Here, it is assumed that the period from T152 to T153 is TB.

  Then, the averaging means 510 averages the detection signals 401 for 15 times in the TB period, and outputs the averaged detection signal 511 to the estimation means 521 and the memory selection means 544, and also detects the detection signal averaging end pulse 512. Is output to the memory selection means 544.

  When the second detection signal averaging end pulse 512 is input to the memory selection unit 544, the averaged detection signal 511 is stored in the comparison detection signal memory 550 as the comparison detection signal 552. Then, the waveform comparison means 570 compares the reference detection signal 562 obtained during the TA period and the comparison detection signal 552 obtained during the TB period.

  After the TB period, the 31st to 45th pulses of the irradiation control signal 104 are generated (T153 to T154). After the 45th pulse of the irradiation control signal 104 is generated, the averaging unit signal shown in FIG. 16 (i) becomes 4 (T154). Here, the period from T153 to T154 is TC.

  Then, the averaging means 510 averages the detection signals 401 for 15 times during the TC period, outputs an averaged detection signal 511 and outputs a detection signal averaging end pulse 512.

  When the third detection signal averaging end pulse 512 is input to the memory selection unit 544, the averaged detection signal 511 is stored in the comparison detection signal memory 550 as the comparison detection signal 552. Then, the waveform comparison means 570 compares the reference detection signal 562 obtained during the TA period with the comparison detection signal 552 obtained during the TC period.

  After the TC period, the 46th to 60th pulses of the irradiation control signal 104 are generated (T154 to T155). After the 60th pulse of the irradiation control signal 104 is generated, the irradiation period signal 103 shown in FIG. 16 (d) becomes Lo (T155). Here, the period from T154 to T155 is assumed to be TD.

  The averaging means 510 averages 15 detection signals 401 in the TD period, outputs an averaged detection signal 511 and outputs a detection signal averaging end pulse 512.

  When the fourth detection signal averaging end pulse 512 is input to the memory selection unit 544, the averaged detection signal 511 is stored in the comparison detection signal memory 550 as the comparison detection signal 552. Then, the waveform comparison means 570 compares the reference detection signal 562 obtained during the TA period with the comparison detection signal 552 obtained during the TD period.

  As described above, in one measurement cycle, the 60 detection signals 401 detected by the irradiation of 60 times of the pulsed light 201 are divided into four, and the reference detection signal 562 obtained in the period of TA and TB˜ By comparing the waveforms of the comparison detection signals 552 obtained during the TD period, a total of three waveform comparison signals 502 are obtained.

  Then, the estimation unit 521 estimates the blood glucose level using only the averaged detection signal 511 that the waveform comparison unit 570 determines to match, and outputs the feature amount estimation signal 501 to the control unit 10.

  The control means 10 counts the number of times of matching among the three waveform comparison signals 502 and performs the next operation according to the count result.

  When the count result is 3 or 2, since it is equal to or greater than the coincidence comparison threshold value (twice), blood glucose level estimation is performed in the same manner as time 0 minute at time 5 minutes, which is the basic measurement cycle of the next noninvasive blood glucose measurement device 1. (T156).

  On the other hand, when the count result is 1 or 0, it is less than the coincidence comparison threshold value (twice), so the pulse of the measurement cycle signal 101 is generated immediately without waiting for the next basic measurement cycle of the noninvasive blood glucose measurement device 1. Perform re-measurement.

  The above operation is performed every measurement cycle.

  In addition, regarding the determination of the re-continuous measurement in the control means 10, an arbitrary number of times is set by the register, and when the continuous measurement is repeated a predetermined number of times, the operation of the non-invasive blood glucose measuring device 1 is stopped. It is also possible to perform processing such as notifying the user using the feature amount display means 60.

  Here, the averaging process of the TA period in the averaging unit 510 may be performed between the end of the TA and the start of the TB, and the TB measurement and the averaging process may be pipelined after the TA ends. . The same applies to the periods TB, TC, and TD.

  In the noninvasive blood glucose measurement device 1 according to the third embodiment, the averaged detection signal 511 obtained during the TA period is used as the reference detection signal 562, and the averaged detection obtained during the TB, TC, and TD periods. Although the signal 511 is the comparison detection signal 552, the reference detection signal 562 can be created from the averaged detection signal 511 obtained in an arbitrary period (TB, TC, or TD) other than TA.

  In the third embodiment, the number of periods in which the irradiation period signal 103 for the measurement period is Hi is described as four times. However, the number of periods is not limited to this and can be set to an arbitrary number. For example, if the total number of pulses generated in the irradiation control signal 104 in one blood sugar level estimation cycle is m times, the total number of detection signals 401 for creating n averaged detection signals 511 is m times. What is necessary is just to set the frequency | count.

  As described above, according to the noninvasive blood glucose measurement device 1 of the third embodiment, in one estimation of blood glucose level, the irradiation period of pulse light repeated at least once is divided into n (n is a natural number of 2 or more). Then, an averaged detection signal is obtained in each of the n-divided irradiation periods, and one of the n averaged detection signals is used as a reference detection signal 562 (1− n) Each comparison detection signal 552 made up of each of the averaged detection signals 511 is compared with the reference detection signal 562, and when the waveform shapes of the both detection signals match, the measurement is performed normally. If they do not match, it is determined that an abnormality has occurred during the measurement. Therefore, the non-invasive blood glucose measuring device 1 is subjected to an impact from the outside during a plurality of times of pulsed light irradiation. Average in case of misalignment Since it is possible to immediately detect the signal quality deterioration of the digitization detection signal 511, it is possible to improve the measurement accuracy.

  Note that the target to be measured by the noninvasive blood glucose measurement device 1 in the first to third embodiments is not limited to the amount of glucose in the blood vessel 31. That is, any substance that absorbs energy in the wavelength region of the irradiated pulsed light 201 and generates a photoacoustic wave may be used. For example, the amount of glucose contained in the tissue fluid between the surface of the living body 30 and the blood vessel 31 It can also be applied to the amount of hemoglobin.

  Since the noninvasive blood glucose measurement device according to the present invention can detect a measurement abnormality by comparing the characteristics included in the photoacoustic signal, the measurement accuracy of the blood glucose level can be improved.

It is a figure which shows the usage example of the noninvasive blood glucose measuring device 1 of this invention. It is a block diagram which shows the structure of the noninvasive blood glucose measuring device 1 by Embodiment 1 of this invention. It is a block diagram which shows the structure of the feature-value estimation means 50 in the noninvasive blood glucose measuring device of the said Embodiment 1. FIG. It is a timing chart figure which shows operation | movement of the noninvasive blood glucose measuring device 1 of the said Embodiment 1. FIG. It is a timing chart figure which shows operation | movement of the feature-value estimation means 50 in the said Embodiment 1. FIG. It is a figure for demonstrating the 1st comparison method of the waveform comparison means 570 in the said Embodiment 1. FIG. It is a figure for demonstrating the 2nd comparison method of the waveform comparison means 570 in the said Embodiment 1. FIG. It is a figure for demonstrating the 3rd comparison method of the waveform comparison means 570 in the said Embodiment 1. FIG. It is a figure for demonstrating the 4th comparison method of the waveform comparison means 570 in the said Embodiment 1. FIG. It is a figure for demonstrating the 5th comparison method of the waveform comparison means 570 in the said Embodiment 1. FIG. It is a figure for demonstrating the 6th comparison method of the waveform comparison means 570 in the said Embodiment 1. FIG. It is a figure for demonstrating the 7th comparison method of the waveform comparison means 570 in the said Embodiment 1. FIG. It is a block diagram which shows the structure of the feature-value estimation means 51 in the noninvasive blood glucose measuring device of Embodiment 2 of this invention. It is a timing chart figure which shows operation | movement of the noninvasive blood glucose measuring device 1 of the said Embodiment 2. FIG. It is a block diagram which shows the structure of the feature-value estimation means 52 in the noninvasive blood glucose measuring device of Embodiment 3 of this invention. Timing chart showing the operation of the non-invasive blood sugar measurement device 1 according to the third embodiment. It is a figure which shows the waveform of the photoacoustic wave signal in the conventional noninvasive blood glucose measuring device, (a) shows the photoacoustic wave signal at the time of normal measurement, (b) shows the photoacoustic wave signal at the time of abnormal measurement Yes.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Noninvasive blood glucose measuring device 10 Control means 20 Light source 30 Living body 31 Blood vessel 40 Photoacoustic detection means 50, 51, 52 Feature quantity estimation means 60 Feature quantity display means 101 Measurement period signal 102 Mode switching signal 103 Irradiation period signal 104 Irradiation control signal 105 Feature Quantity Display Control Signal 201 Pulsed Light 301 Photoacoustic Wave Signal 401 Detection Signal 501 Feature Quantity Estimation Signal 502 Waveform Comparison Signal 510 Averaging Unit 511 Averaging Detection Signal 512 Detection Signal Averaging End Pulse 520, 521 Estimating Unit 540, 543 544 Memory selection means 541, 542 Average detection signal after selection 550 Comparison detection signal memory 551 Comparison detection signal storage completion signal 552 Comparison detection signal 560, 563 Reference detection signal memory 561 Reference detection signal storage completion signal 562 Detection signal for reference 70 waveform comparison means

Claims (27)

In a non-invasive blood glucose measurement device that estimates a blood glucose level by detecting a photoacoustic wave signal emitted from a specific substance in the living body by absorbing light energy by light irradiated on the living body surface,
Control means for outputting an irradiation control signal for controlling irradiation of the pulsed light that is repeated at least once for one estimation of blood glucose level;
At least one light source that irradiates the surface of the living body with the pulsed light according to the irradiation control signal;
Photoacoustic detection means for outputting a detection signal obtained by sampling the photoacoustic wave signal at a predetermined frequency;
A feature amount estimating means for averaging the detection signals for the number of times of irradiation of the pulsed light repeated at least once and estimating the blood glucose level using the averaged detection signals;
Feature amount display means for displaying the estimated blood glucose level,
The feature amount estimating means creates a reference detection signal as a comparison reference from at least one of the plurality of averaged detection signals obtained by estimating the blood glucose level at least twice. Each of a plurality of averaged detection signals other than the signal used to create the reference detection signal is used as a comparison detection signal, and measurement is performed by comparing the waveform shapes of the comparison detection signal and the reference detection signal. To determine if an abnormality has occurred during
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 1,
The control means is
Setting a first period for estimating blood glucose level at least once and a second period for estimating blood glucose level at least once after the end of the first period; and A switching signal for switching between the second period is output to the feature quantity estimating means,
The feature quantity estimating means is
When the switching signal indicates the first period, any one of the averaged detection signals obtained for the estimated number of times of the blood glucose level at least one time, or a plurality of the same waveform shapes Or a signal obtained by further averaging an averaged detection signal of any number of them as a reference detection signal,
When the switching signal indicates a second period, a comparison detection signal composed of an averaged detection signal obtained for each estimation of the blood glucose level at least once is compared with the reference detection signal; If both detection signals match, it is determined that the measurement is performed normally. If they do not match, it is determined that an abnormality has occurred during the measurement.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 1,
The control means is
A third period during which at least one blood glucose level is estimated at a predetermined time interval, and a part of the third period within the predetermined interval, wherein at least two blood glucose levels are estimated Set a fourth period of continuous estimation,
The feature quantity estimating means is
In the fourth period, the waveform shapes of the averaged detection signals obtained for the estimated number of blood glucose levels of at least two times are compared with each other, and the number of averaged detection signals having the same waveform shape is equal to or greater than a first threshold value. When one of the plurality of averaged detection signals having the same waveform shape, or the plurality of averaged detection signals having the same waveform shape, or any number of the averaged detection signals among them. A signal obtained by further averaging is used as a reference detection signal.
In the third period, when a comparison detection signal composed of an averaged detection signal obtained for each estimation of the blood glucose level at least once is compared with the detection signal for reference, and the two detection signals match Determines that the measurement has been performed normally, and if they do not match, determines that an abnormality has occurred during the measurement.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 1,
The feature quantity estimating means is
In the estimation of the blood glucose level once, the irradiation period of the pulsed light repeated at least once is divided into n (n is a natural number of 2 or more),
An averaged detection signal is obtained in each of the n divided irradiation periods,
One of the n averaged detection signals is used as a reference detection signal, and each comparison detection signal including (1-n) averaged detection signals other than the reference detection signal is referred to as the reference. If the waveform shapes of the two detection signals match with each other, it is determined that the measurement is normally performed, and if they do not match, it is determined that an abnormality has occurred during the measurement.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 1,
The feature quantity estimating means is
Using the reference estimation number setting value indicating the estimated number of blood sugar levels at least once, the averaged detection signal in the estimation of the blood glucose level before the number indicated by the reference estimation number setting value as a reference detection signal,
A comparison detection signal consisting of an averaged detection signal newly obtained by blood glucose level estimation is compared with the reference detection signal, and when the waveform shapes of the two detection signals match, the measurement is performed normally. If it does not match, it is determined that an abnormality has occurred during measurement.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to any one of claims 2 to 5,
The feature amount estimating means is
When the waveform shapes of the comparison detection signal and the reference detection signal are compared, the measurement abnormality determination setting value indicating a natural number of at least 1 or more is used, and the number of times of mismatching is determined as the measurement abnormality determination setting value. It is determined that the measurement is normally performed until reaching the number of times indicated, and when the number of times indicated by the measurement abnormality determination set value is reached, it is determined that an abnormality has occurred during the measurement.
A non-invasive blood sugar measuring device characterized by the above. The non-invasive blood sugar measuring device according to claim 2,
The first period is a calibration mode, and is a period for calculating a calibration formula for calibrating the pre-calibration blood glucose level estimated by the non-invasive blood glucose measurement device with the blood glucose level measured by the invasive blood glucose measurement device ,
A non-invasive blood sugar measuring device characterized by the above. The non-invasive blood sugar measuring device according to claim 2,
The second period is a normal measurement mode, and is a period for estimating a blood glucose level by applying a calibration formula to the blood glucose level before calibration estimated by the non-invasive blood glucose measuring device.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 1,
The feature amount estimating means includes:
In comparing the detection signal for reference and the detection signal for comparison,
Taking the irradiation timing of the pulsed light by the irradiation control signal as a reference position, the reference detection signal and the comparison detection signal are synchronized,
Detecting the maximum peak position where the amplitude absolute value of the reference detection signal is maximum,
The first comparison start position and the first comparison end are the positions at which the polarity of the slope of the reference detection signal changes at the position closest to the maximum peak position in both positive and negative sampling timing directions from the maximum peak position, respectively. Position,
In the comparison period from the first comparison start position to the first comparison end position, the number of change in polarity of the slope of the comparison detection signal is counted, and when the count result is 1, the measurement is performed normally. If the count result is not 1, it is determined that an abnormality has occurred during measurement.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 1,
The feature amount estimating means includes:
In comparing the detection signal for reference and the detection signal for comparison,
Taking the irradiation timing of the pulsed light by the irradiation control signal as a reference position, the reference detection signal and the comparison detection signal are synchronized,
After the pulse light irradiation, in the comparison period in which the shapes of the reference detection signal and the comparison detection signal are compared, the maximum amplitude value of the reference detection signal and the maximum amplitude value of the comparison detection signal are made equal. Standardize,
Calculating the sum of square errors in the amplitude value of the reference detection signal and the amplitude value of the comparison detection signal;
If the total amount of the square error is less than or equal to the second threshold, it is determined that the measurement is normally performed, and if greater than the second threshold, it is determined that an abnormality has occurred during the measurement.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 1,
The feature amount estimating means includes:
In comparing the detection signal for reference and the detection signal for comparison,
Taking the irradiation timing of the pulsed light by the irradiation control signal as a reference position, the reference detection signal and the comparison detection signal are synchronized,
After the pulsed light irradiation, in a comparison period in which the shape of the reference detection signal and the comparison detection signal are compared, a correlation coefficient between the reference detection signal and the comparison detection signal is calculated,
If the correlation coefficient is greater than a third threshold, it is determined that the measurement is normally performed, and if it is equal to or less than the third threshold, it is determined that an abnormality has occurred during the measurement.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 1,
The feature amount estimating means includes:
In comparing the detection signal for reference and the detection signal for comparison,
Taking the irradiation timing of the pulsed light by the irradiation control signal as a reference position, the reference detection signal and the comparison detection signal are synchronized,
After the pulse light irradiation, in the comparison period in which the shapes of the reference detection signal and the comparison detection signal are compared, the maximum amplitude value of the reference detection signal and the maximum amplitude value of the comparison detection signal are made equal. Standardize,
When the number of samplings in which the amplitude levels of the reference detection signal and the comparison detection signal exceed the fourth threshold are counted, and the difference between the respective count results is not less than the fifth threshold and not more than the sixth threshold Judge that the measurement is performed normally, and if it is less than the fifth threshold or greater than the sixth threshold, it is judged that an abnormality has occurred during the measurement.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 1,
The feature amount estimating means includes:
In comparing the detection signal for reference and the detection signal for comparison,
Taking the irradiation timing of the pulsed light by the irradiation control signal as a reference position, the reference detection signal and the comparison detection signal are synchronized,
After the pulse light irradiation, in the comparison period in which the shapes of the reference detection signal and the comparison detection signal are compared, the number of change in polarity of the reference detection signal and the comparison detection signal is counted, and the respective counts are counted. If the difference in results is greater than or equal to the seventh threshold and less than or equal to the eighth threshold, it is determined that the measurement is being performed normally, and if it is less than the seventh threshold or greater than the eighth threshold, an abnormality occurs during measurement. Judge that
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 1,
The feature amount estimating means includes:
In comparing the detection signal for reference and the detection signal for comparison,
Taking the irradiation timing of the pulsed light by the irradiation control signal as a reference position, the reference detection signal and the comparison detection signal are synchronized,
After the pulsed light irradiation, in the comparison period in which the shape of the reference detection signal and the comparison detection signal are compared, the inclinations of the reference detection signal and the comparison detection signal are respectively calculated.
If the position where the inclination of the comparison detection signal is equal to or smaller than the ninth threshold is the same as the position where the inclination of the reference detection signal is equal to or smaller than the ninth threshold, the measurement is normally performed. If it is different, it is determined that an abnormality occurred during measurement.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 1,
The feature amount estimating means includes:
In comparing the detection signal for reference and the detection signal for comparison,
Taking the irradiation timing of the pulsed light by the irradiation control signal as a reference position, the reference detection signal and the comparison detection signal are synchronized,
After the pulsed light irradiation, in the comparison period for comparing the shape of the reference detection signal and the comparison detection signal, the reference detection signal and the comparison detection signal are frequency-converted,
In the frequency-converted reference detection signal and the conversion comparison detection signal, if there is no difference in frequency components whose amplitude is equal to or greater than a tenth threshold, it is determined that the measurement is performed normally, and the frequency component If there is a difference, determine that an abnormality has occurred during measurement.
A non-invasive blood sugar measuring device characterized by the above. The non-invasive blood glucose measurement device according to claim 5,
The feature amount estimating means includes:
A writable register is provided, and the reference estimated number of times setting value can be changed from the outside by the register.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 6,
The feature amount estimating means includes:
A writable register is provided, and the measurement abnormality determination setting value can be changed from the outside by the register.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 3,
The feature amount estimating means includes:
A register capable of writing, and the first threshold value can be changed from the outside by the register;
A non-invasive blood sugar measuring device characterized by the above. The non-invasive blood glucose measurement device according to claim 10,
The feature amount estimating means includes:
A register capable of writing, and the second threshold value can be changed from the outside by the register;
A non-invasive blood sugar measuring device characterized by the above. The non-invasive blood sugar measurement device according to claim 11,
The feature amount estimating means includes:
A register capable of writing, and the third threshold value can be changed from the outside by the register;
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 12,
The feature amount estimating means includes:
A plurality of writable registers, the fourth threshold value, the fifth threshold value and the sixth threshold value can be changed from the outside by each of the registers;
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 13,
The feature amount estimating means includes:
A plurality of writable registers, each of which can change the seventh threshold value and the eighth threshold value from the outside;
A non-invasive blood sugar measuring device characterized by the above. The non-invasive blood glucose measurement device according to claim 14,
The feature amount estimating means includes:
A register capable of writing, and the ninth threshold value can be changed from the outside by the register;
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood sugar measurement device according to claim 15,
The feature amount estimating means includes:
A writable register, and the tenth threshold value can be changed from the outside by the register;
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 9,
The feature quantity estimation means includes:
After performing a filtering process that passes only a specific frequency band on the comparison detection signal in the comparison period, the number of polarity changes in the slope of the comparison detection signal is counted.
A non-invasive blood sugar measuring device characterized by the above. The noninvasive blood glucose measurement device according to claim 13,
The feature quantity estimation means includes:
After filtering the reference detection signal and the comparison detection signal through only a specific frequency band in the comparison period, the number of change in polarity of the slope of the reference detection signal and the comparison detection signal is calculated. To count,
A non-invasive blood sugar measuring device characterized by the above. The non-invasive blood glucose measurement device according to claim 14,
The feature quantity estimation means includes:
After performing filtering that passes only a specific frequency band for the reference detection signal and the comparison detection signal in the comparison period, the inclination of the reference detection signal and the comparison detection signal is calculated.
A non-invasive blood sugar measuring device characterized by the above.

Images