JP2018099209A - Biological value arithmetic system, biological value arithmetic device, program, and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve a device and a system for measuring a biological value such as blood pressure on a living body.SOLUTION: In a biological value arithmetic system 1, a sensor device 2 detects a value that changes according to a life activity of a living body. An arithmetic device 3 acquires a detection biological value by performing data communication with the sensor device 2, and calculates an evaluation biological value on the living body by arithmetic processing using a plurality of feature values indicating features of a waveform indicated by a plurality of the detection biological values, and a plurality of regression arithmetic formulae. An output device 4 outputs the evaluation biological value acquired by performing data communication with the arithmetic device 3 or a result of determination on the evaluation biological value.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a biological value calculation system, a biological value calculation device, a biological value calculation method, a program, and a recording medium that calculate an evaluation biological value related to a living organism such as blood pressure.

  For example, an evaluation biological value related to a living body such as blood pressure is useful information for determining the health state of the living body. And in the life crisis notification system of patent document 1, a user's blood pressure etc. are detected, and when there exists a predetermined change, an emergency call is performed. Thus, for example, as in the emergency notification system of Patent Document 2, an emergency call can be made even when the user is not conscious as compared to the case where the user makes an emergency call by pressing a button switch.

JP 2006-334369 A JP 2006-135742 A

However, in Patent Document 1, it is necessary to use a life crisis reporting device designed exclusively for the life crisis reporting system. In particular, the life crisis reporting device of Patent Document 1 has a multifunctional device such as a liquid crystal monitor, a built-in speaker, a built-in microphone, and the like in addition to the blood pressure measurement unit. For this reason, the life crisis reporting device is expensive, and the weight and size are also increased.
In particular, when used for medical purposes, it is necessary to use a blood pressure measuring unit that can measure blood pressure with high accuracy. For this reason, the blood pressure measurement unit is also expensive and complicated.
Since these are all burdens on the user, a system such as Patent Document 1 has not been widely spread.

  Thus, further improvements are required in devices and systems that measure biological values such as blood pressure of living organisms.

  A biological value calculation system according to the present invention includes a sensor device that detects a value that changes with the life activity of a living body, and a detected biological value by performing data communication with the sensor device to obtain a plurality of detected biological values. An arithmetic device that calculates an evaluation biological value related to the living body by arithmetic processing using a plurality of feature values and a plurality of regression arithmetic expressions indicating the characteristics of the waveform represented by: and obtained by data communication with the arithmetic device An output device that outputs the evaluation biometric value or the determination result of the evaluation biometric value.

  Preferably, the calculation device stores a plurality of regression calculation formula data whose coefficients have been determined by regression analysis, and reads the plurality of regression calculation formula data from the storage unit to evaluate the evaluation organism. A calculation unit that executes a value calculation process, wherein the calculation unit uses a plurality of the detected biometric values acquired from the sensor device to obtain characteristics of the waveform represented by the plurality of detected biometric values. A process of generating a plurality of feature values, a process of calculating a plurality of expansion value series by an operation combining a plurality of the feature values, and a target value that is a calculation result of each of the regression calculation formulas in the regression analysis A plurality of the target values are executed by performing a process of calculating using the developed value series or the eigenvalues of which the combination is determined for each regression calculation formula as an explanatory value, and a process of calculating the sum of the target values. Total value The computed as the evaluation biological value, the better.

  Preferably, each regression equation and its coefficient are obtained by substituting a plurality of feature values into a plurality of unfolded development equation groups prepared in advance in the regression analysis. A series of reference biological values or a series of residuals obtained when a plurality of developed value series are calculated by calculating a developed value, and for each of the plurality of developed value series, the evaluation biological value is detected from an organism. And is determined based on the correlation between the developed value series having the highest correlation and the reference biometric value series or residual series. The explanatory value to be used may be the developed value series having the highest correlation or its eigenvalue obtained based on the plurality of detected biological values acquired from the sensor device.

  Preferably, the plurality of regression calculation formulas are combinations of a plurality of calculation formulas selected so as to sequentially reduce a residual with a series of reference biometric values obtained when the evaluation biometric value is detected from a living organism. There should be.

  Preferably, the plurality of regression arithmetic expressions are correlated with a series of reference biological values obtained when the evaluation biological value is detected from a living body from among the plurality of expanded value series in at least a regression analysis. A first regression equation for converting the expanded value series selected to be evaluated as the highest to the reference biological value series, and in regression analysis, the reference biological value is selected from a plurality of the expanded value series. Second regression that converts the expanded value series selected by evaluating that the correlation between the series of residuals and the series of residuals according to the first regression equation is the highest to the residual series The expanded value series used as an explanatory variable in the second regression calculation formula may be different from the expanded value series used as an explanatory variable in the first regression calculation formula.

  Preferably, at least a part of the regression calculation formulas among the plurality of regression calculation formulas uses the developed value series different from the developed value series used in the other regression calculation formulas.

  Preferably, the evaluation biological value is a maximum blood pressure or a minimum blood pressure of an organism, the detected biological value is a pulsating flow value in which blood flow is detected, and the plurality of characteristic values are a plurality of pulsating flow values. The maximum value, minimum value, maximum speed value, minimum speed value, maximum acceleration value, minimum acceleration value, and pulse rate in the waveform represented by

  Preferably, the sensor device may detect a value that changes with the heartbeat of the living body as the value that changes with the life activity of the living body.

  Preferably, the calculation device stores a plurality of regression calculation formula data whose coefficients have been determined by regression analysis, and reads the plurality of regression calculation formula data from the storage unit to evaluate the evaluation organism. A calculation unit that executes a value calculation process, wherein the calculation unit uses a plurality of the detected biometric values acquired from the sensor device to obtain characteristics of the waveform represented by the plurality of detected biometric values. A process for generating a series of characteristic values to be shown, and a target value that is a calculation result of each regression calculation formula, a series of the feature values or combinations of eigenvalues determined for each regression calculation formula in a regression analysis It is good to perform the process calculated using it as an explanatory value, and the process which calculates the sum total of several said target value, and compute the sum total of several said target value as said evaluation biometric value.

  Preferably, the sensor device includes: a sensor that detects a value that changes with the life activity of the living organism; and a wireless communication unit that performs data communication of the detected biological value detected by the sensor to the arithmetic device. It is good to have.

  Preferably, the output device includes a comparison unit that compares the biometric value acquired from the arithmetic device by data communication with an evaluation reference value, and a notification that automatically reports when the comparison result by the comparison unit is bad or not good. It is good to have a part.

  A biological value calculation apparatus according to the present invention acquires a detected biological value as data from a sensor that detects a detected biological value that changes with the life activity of a living organism, and is represented by a plurality of detected biological values. By calculating using a plurality of feature values indicating the characteristics of the waveform and a plurality of regression equations, a calculation unit that calculates an evaluation biological value related to the living organism, and obtaining the evaluation biological value as data from the calculation unit, An output unit that outputs the evaluation biometric value or a determination result of the evaluation biometric value.

  The biological value calculation method according to the present invention is a method of generating an evaluation biological value by detecting a detected biological value that changes with the life activity of a living body by a sensor device, and the detected biological value is obtained from the sensor device. Obtained and calculated an evaluation biometric value related to the living organism by a calculation process using a plurality of feature values indicating a feature of the waveform represented by the acquired plurality of detected biometric values and a plurality of regression calculation formulas. The evaluation biological value or the determination result about the evaluation biological value is output.

  The program according to the present invention includes a step of acquiring a detected biological value that changes with the life activity of a living body, a plurality of feature values indicating a feature of a waveform represented by the plurality of acquired detected biological values, and a plurality of characteristic values The computer is caused to perform a step of calculating an evaluation biometric value related to the living object and a step of outputting the calculated evaluation biometric value or the evaluation biometric value based on a calculation process using a regression equation.

  A computer-readable recording medium according to the present invention records the above-described program.

In the present invention, a sensor device that detects a value that changes with the life activity of a living organism, and a detected biological value obtained by data communication with the sensor device, and a waveform feature represented by a plurality of detected biological values An arithmetic device that calculates an evaluation biological value related to a living organism by an arithmetic process using a plurality of feature values and a plurality of regression calculation expressions, and an evaluation biological value or an evaluation biological value acquired by data communication with the arithmetic device And an output device for outputting the determination result.
Therefore, for example, without directly measuring the evaluation biological value such as blood pressure, the detected biological value such as a blood flow value that changes with the life activity of the living body and based on the detected biological value that has a correlation with the evaluation biological value in the living body Thus, an evaluation biological value such as blood pressure can be obtained. Moreover, the judgment result about this evaluation biometric value can be output as an emergency call, for example.

On the other hand, when a normal method, for example, an evaluation biological value is directly detected and corrected as necessary, a dedicated device for detecting the evaluation biological value is required. In addition, since the accuracy of the detection value directly affects the accuracy of the evaluation value, a dedicated device with high detection accuracy of the detection value is required to obtain an accurate evaluation value.
In the present invention, the evaluation biological value is not detected. For this reason, it is possible to set a biological value that is easy to detect from a living object as a detected biological value. In addition, instead of using the detected biological value as it is, an evaluation biological value is obtained from the detected biological value by using a regression equation based on the waveform characteristics of the detected biological value. Therefore, even if the detected biometric value is detected with a low absolute accuracy using, for example, an inexpensive sensor, an evaluation biometric value with high probability can be obtained.

FIG. 1 is an explanatory diagram of an example of a biological value calculation system. FIG. 2 is an explanatory diagram of an example of the wristband of FIG. FIG. 3 is a circuit diagram of an example of the sensor device of FIG. 4 is a block diagram of an electric circuit provided in the portable terminal of FIG. FIG. 5 is a flowchart illustrating an example of a method for determining a plurality of regression equations and their coefficients. FIG. 6 is an explanatory diagram of an example of a plurality of feature values. FIG. 7 is an explanatory diagram of an example of a first expansion calculation expression group for calculating a plurality of expansion values from a plurality of feature values. FIG. 8 is an explanatory diagram of an example of a plurality of first expansion value series generated by the first expansion arithmetic expression group in FIG. FIG. 9 is an explanatory diagram of an example of the second expansion arithmetic expression group and a plurality of second expansion value series. FIG. 10 is a diagram for explaining that a plurality of regression calculation equations can be obtained by using a plurality of development value series. FIG. 11 is an explanatory diagram of a first regression calculation formula for calculating the systolic blood pressure and a method for determining its coefficient. FIG. 12 is an explanatory diagram of a second regression calculation formula for reducing the residual according to the first regression calculation formula of FIG. 11 and a method for determining its coefficient. FIG. 13 is an explanatory diagram of a third regression calculation formula for reducing the residual by the second regression calculation formula of FIG. 12 and a method for determining its coefficient. FIG. 14 is an explanatory diagram of a fourth regression calculation formula for reducing the residual according to the third regression calculation formula of FIG. 13 and a method for determining its coefficient. FIG. 15 is an explanatory diagram of a fifth regression calculation formula for reducing the residual based on the fourth regression calculation formula of FIG. 14 and a method for determining its coefficient. FIG. 16 is an explanatory diagram of a sixth regression calculation formula for reducing the residual according to the fifth regression calculation formula of FIG. 15 and a method for determining its coefficient. FIG. 17 is an explanatory diagram of a seventh regression calculation formula for reducing the residual by the sixth regression calculation formula of FIG. 16 and a method for determining its coefficient. FIG. 18 is an explanatory diagram of an eighth regression calculation formula for reducing the residual according to the seventh regression calculation formula of FIG. 17 and a method for determining its coefficient. FIG. 19 is a diagram for explaining the calculation formula of the systolic blood pressure (evaluation biological value) and the relationship between the value and the reference biological value. FIG. 20 is a diagram for explaining a plurality of regression calculation formulas for calculating the minimum blood pressure and a calculation formula for the minimum blood pressure. FIG. 21 is a flowchart illustrating an example of an evaluation biometric value calculation method. FIG. 22 is a flowchart illustrating an example of a method for outputting an evaluation biometric value. FIG. 23 is an explanatory diagram of another example of a plurality of feature values. FIG. 24 is an explanatory diagram showing a correlation between each of the plurality of feature values of FIG. 23 and an evaluation biological value (maximum blood pressure). FIG. 25 is an explanatory diagram of one application example of the biological value calculation system of FIG. FIG. 26 is an explanatory diagram of another application example of the biological value calculation system of FIG. FIG. 27 is an explanatory diagram of another application example of the biological value calculation system of FIG.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[First embodiment]
FIG. 1 is an explanatory diagram of an example of a biological value calculation system 1.
The biological value calculation system 1 in FIG. 1 includes a sensor device 2, a calculation device 3, and an output device 4.

The sensor device 2 is mounted on a wristband 6 that is attached to an arm of a human body that is a living body. In addition to this, the sensor device 2 may be provided, for example, on a ring or an earring. Further, the sensor device 2 may be a non-contact type that does not adhere to the human body.
The arithmetic device 3 and the output device 4 are mounted on a portable terminal 7 capable of data communication with the wristband 6. Further, the mobile terminal 7 can communicate with the emergency device 9 via the communication network 8.
And the sensor apparatus 2 detects the value (detection biological value) which changes with the life activity of a living body.
The arithmetic device 3 obtains a detected biometric value by performing data communication with the sensor device 2, and performs arithmetic processing using a plurality of feature values indicating a feature of the waveform represented by the plurality of detected biometric values and a plurality of regression equations. To calculate an evaluation biometric value related to the life form.
The output device 4 outputs an evaluation biometric value or a determination result about the evaluation biometric value acquired by data communication with the arithmetic device 3.
1 shows an example in which the communication network 8 is used for data communication between the portable terminal 7 and the emergency device 9, but the communication network 8 is used for data communication between the sensor device 2 and the portable terminal 7. May be.

FIG. 2 is an explanatory diagram of an example of the wristband 6 of FIG. The wristband 6 is wound around an arm illustrated by hatching.
FIG. 3 is a circuit diagram of an example of the sensor device 2 of FIG. The arms and blood vessels are shown on the left side of the figure.

  The wristband 6 includes a band portion 11 around which the sensor device 2 is wound around a human arm, and a sensor device 2 mounted on the band portion 11.

The sensor device 2 includes a light emitting diode 12, a light receiving transistor 13, an operational amplifier 17, an integrated circuit 20, and a wireless communication unit 21.
The light emitting diode 12 is an LED element, for example, and outputs light. The light is reflected according to the pulsating flow of blood flowing through the blood vessel inside the arm to which the band unit 11 is attached. The amount of reflection varies depending on, for example, the flow rate of hemoglobin. The flow rate of hemoglobin has a favorable correlation with blood pressure and the like. The light emitting diode 12 may be in contact with the skin or may be separated.
The light receiving transistor 13 receives the reflected light, and a current corresponding to the amount of received light flows. This current generates a voltage in the load resistance element 14. The voltage of the load resistance element 14 passes through a high-pass filter including the capacitor 15 and the resistance element 16 and is supplied to the non-inverting input terminal of the operational amplifier 17. Note that the light receiving transistor 13 may receive the transmitted light so as to be opposed to the light emitting diode 12.
A parallel circuit of a capacitor 18 and a resistance element 19 is connected to the operational amplifier 17 between the inverting input terminal and the output terminal. As a result, the operational amplifier 17 amplifies the voltage change input to the non-inverting input terminal and outputs the amplified voltage change from the output terminal.
The output terminal of the operational amplifier 17 is connected to the integrated circuit 20. The integrated circuit 20 samples the input voltage and generates a detected biological value. For example, blood pulsation values and raw data for detecting them are generated. The blood pulsation value and the raw data for detecting it are values that change with the heartbeat of the living organism.
A wireless communication unit 21 is connected to the integrated circuit 20. The wireless communication unit 21 transmits the detected biological value data generated by the integrated circuit 20 to the mobile terminal 7.
As described above, the sensor device 2 optically detects the biological value by the light emitting diode 12 and the light receiving transistor 13 as a sensor, and transmits it to the portable terminal 7.
In addition to this, the sensor device 2 may detect, for example, a value of pressure generated in the living body or a value due to electromagnetic change of the living body.

FIG. 4 is a block diagram of an electric circuit provided in the mobile terminal 7 of FIG. The electric circuit in FIG. 4 is a kind of computer.
4 includes a memory 31 for storing a program, a CPU (Central Processing Unit) 32 for executing the program, a timer 33 as a peripheral device, a wireless communication unit 34, a liquid crystal device 35, and a speaker 36. . The timer 33, the wireless communication unit 34, the liquid crystal device 35, and the speaker 36 exchange data with the CPU 32 through an internal bus.
Note that the program stored in the memory 31 may be partly or entirely installed in a portable terminal 7 that is recorded on a computer-readable recording medium such as a recording medium. It may be installed in the portable terminal 7 through a transmission medium such as a communication network from a server device capable of communicating with the device.

The timer 33 measures time. For example, the timer 33 measures the period designated by the CPU 32 and outputs to the CPU 32 that it has elapsed.
The wireless communication unit 34 transmits / receives data to / from wireless communication units of other devices such as the wireless communication unit 21 and the emergency device 9 of the wristband 6 of FIG. The wireless communication unit 34 outputs the communication data received from the wireless communication unit of the other device to the CPU 32, and transmits the communication data received from the CPU 32 to the wireless communication unit of the other device.
The liquid crystal device 35 displays an image. The liquid crystal device 35 receives image data from the CPU 32 and displays an image.
The speaker 36 outputs sound. The speaker 36 receives sound data from the CPU 32 and outputs sound.
The memory 31 stores data of a plurality of regression calculation formulas whose coefficients are determined by regression analysis as part of the program. The memory 31 accumulates and stores the detected biological values acquired from the sensor control unit by data communication.
The CPU 32 reads a program from the memory 31 and executes it. Thereby, the calculating part 37 is implement | achieved.

  And the calculating part 37 calculates the evaluation biological value regarding a living body by the arithmetic processing using the several feature value which shows the characteristic of the waveform represented by several detection biometric value, and several regression equation.

  Moreover, the calculating part 37 outputs the judgment result about the evaluation biological value acquired by the data communication between programs, or an evaluation biological value. For example, the calculation unit 37 compares the acquired evaluation biometric value with the evaluation reference value, and executes the automatic notification process using the wireless communication unit 34 when the comparison result is bad or not good.

1 may have an electric circuit having the same function as the portable terminal 7 in FIG.
It may be realized on a computer. In this case, even if the program stored in the memory of the emergency device 9 is partly or wholly recorded on a computer-readable recording medium such as a recording medium, the program is installed in the emergency device 9. Alternatively, a server device that can communicate with the emergency device 9 may be installed in the emergency device 9 through a transmission medium such as a communication network.

Next, a plurality of regression calculation formulas used in the biological value calculation system 1 in FIG. 1 and a method for determining coefficients thereof will be described.
FIG. 5 is a flowchart illustrating an example of a method for determining a plurality of regression equations and their coefficients.

As shown in FIG. 5, the calculating part 37 acquires the detection biometric value detected with the sensor apparatus 2 first (step ST1). The calculation unit 37 may acquire an average value of a plurality of detected biological values detected by the sensor device 2. The average value may be, for example, a simple moving average.
Further, the detected biological value may be any value that can find a certain correlation with the reference biological value. For example, a certain correlation can be found between blood pressure and blood flow. If the blood flow is large, the blood pressure tends to be high, and if the blood flow is small, the blood pressure tends to be low.

  And the calculating part 37 produces | generates the several feature value which shows the characteristic of the waveform represented by several detected biometric value using the some detected biometric value (step ST2).

FIG. 6 is an explanatory diagram of an example of a plurality of feature values.
The waveform in FIG. 6A is a pulsating flow waveform represented by a plurality of pulsating flow values obtained by optically detecting blood flow. The pulsating flow value is a detected biological value. The horizontal axis is time, and the vertical axis is the pulsating flow value.
The waveform in FIG. 6B is a pulsating velocity waveform obtained by differentiating the pulsating waveform in FIG. The horizontal axis is time, and the vertical axis is the pulsating velocity.
The waveform in FIG. 6C is a pulsating acceleration waveform obtained by differentiating the pulsating velocity waveform in FIG. The horizontal axis is time, and the vertical axis is pulsating acceleration.
The waveform in FIG. 6D is a frequency distribution obtained by Fourier transforming the pulsating flow waveform in FIG. The horizontal axis is the frequency, and the vertical axis is the component ratio of each frequency.

And the calculating part 37 produces | generates the some feature value which shows the characteristic of a waveform about these waveforms. The feature values X1 to X7 become different values when the waveform changes. That is, the characteristic value takes a unique value corresponding to the waveform.
The feature values X1 to X7 only need to have a certain correlation with the waveform of the evaluation biological value. Preferably, the feature value is less likely to change due to factors other than the evaluation biological value.
The pulsating flow waveform in FIG. 6A is generated as a maximum value X1 and a minimum value X4 of the waveform as feature values indicating the peak of the waveform.
The pulsating velocity waveform in FIG. 6B is generated as a maximum velocity value X2 and a minimum velocity value X5 of the waveform as feature values indicating the peak of the waveform.
The pulsating acceleration waveform of FIG. 6C is generated as a maximum acceleration value X3 and a minimum acceleration value X6 of the waveform as feature values indicating the peak of the waveform.
For example, the pulse value X7 is generated based on the time interval between two adjacent waveform peaks in the pulsating acceleration waveform of FIG. 6C.
For the frequency distribution in FIG. 6D, the component ratio values of the 19 peak frequencies are extracted as feature values Y1 to Y19.
The feature value Y0 = 1 is added to these to obtain a total of 27 feature values.
For Y1 to Y19, an arbitrary value such as a random number may be used instead of the feature value.

Next, the calculation unit 37 calculates a plurality of development values by a calculation combining a plurality of feature values, and generates a plurality of development value series (step ST3).
In the present embodiment, about 10,000 expansion values are generated using the first expansion arithmetic expression group in FIG. 7 and the second expansion arithmetic expression group in FIG. 9A, and are generated as shown in FIG. 9B. Are classified into 15 second expansion value series.
FIG. 7 is an explanatory diagram of an example of a first expansion calculation expression group for calculating a plurality of expansion values from a plurality of feature values.

The first expansion arithmetic expression group in FIG. 7 includes 31 expansion arithmetic expressions from the expansion arithmetic expression F1 to the expansion arithmetic expression F31.
In each expansion calculation expression of FIG. 7, a radian value is used as the substitution value of the trigonometric function. That is, for example, in the expansion calculation formula F1, as shown below, the radians value of “Yn * π / 2” and the radians value of “Xn * Yn” are substituted into the sine function.
F1 = SIN (RADIANS (Yn * π / 2)) * (1 / Yn ² ) * SIN (RADIANS (Xn * Yn))
Each expansion arithmetic expression is composed of a combination of trigonometric functions. In each formula, “* (asterisk)” means multiplication.
By using a combination of different trigonometric functions, each expansion arithmetic expression becomes a series having different characteristics from other expansion arithmetic expressions. It is possible to improve the multicollinearity between a plurality of expansion arithmetic expressions and improve statistical accuracy.

FIG. 8 is an explanatory diagram of an example of a plurality of first expansion value series generated by the first expansion arithmetic expression group in FIG.
As shown in FIG. 8, a total of 3380 expanded values can be calculated using a plurality of feature values X1 to X7, Y0 to Y19 and 31 expanded calculation formulas. The plurality of expansion values can be classified into, for example, 31 expansion value series for each formula.

Furthermore, the calculation unit 37 generates a plurality of second expansion value series in FIG. 9B using the first expansion value series and the second expansion calculation expression group in FIG. 9A.
The group of second expansion arithmetic expressions in FIG. 9A includes an involute curve function Fa, a clothoid curve function (X side) Fb, and a clothoid curve function (Y side) Fc. In addition, a spiral curve function, a cycloid curve function, or the like may be used. By converting and evaluating with these functions, the significance of a plurality of development value series can be improved.
The plurality of second expanded value series in FIG. 9B is 15 series from G1 to G15.
In the subsequent processing in the present embodiment, the plurality of second development value series are used as a plurality of development value series.

Next, the calculation unit 37 selects a development value series to be used for the regression calculation expression from a plurality of development value series, and starts processing for determining coefficients of the plurality of regression calculation expressions.
The plurality of regression equations are combinations of a plurality of equations selected so as to sequentially reduce the residual with the series of reference biological values obtained by directly detecting the evaluation biological values from the human body.
The reference biometric value may be a value measured with a sphygmomanometer, for example.

FIG. 10 is a diagram for explaining that a plurality of regression equations can be obtained by using a plurality of development value series.
FIG. 10A is an explanatory diagram showing a correlation between a series of reference biological values and an A development value series. The vertical axis is the reference biometric value, and the horizontal axis is the development value of the A development value series. The distribution of the plurality of development values and the reference biological value is represented by a plurality of dots. In the case of FIG. 10A, the first regression equation that minimizes the residuals for a plurality of dots can be obtained by regression analysis, for example.
FIG. 10B is an explanatory diagram showing the correlation between the residual series based on the first regression equation and the A-expansion value series. The vertical axis is the residual, and the horizontal axis is the expansion value of the A expansion value series. Note that the residual based on the first regression equation can be obtained by subtracting the target value of the first regression equation from the reference biological value. As shown in FIG. 10B, it is not possible to obtain a second recursive operation expression having an effective slope between the series of residuals based on the first regression operation expression and the A-expansion value series.
FIG. 10C is an explanatory diagram showing the correlation between the residual series based on the first regression equation and the B-expansion value series. The vertical axis is the residual, and the horizontal axis is the expanded value of the B expanded value series. As shown in FIG. 10C, a plurality of dots can be distributed with an effective gradient between the series of residuals based on the first regression equation and the B-expansion value series. As described above, the second regression calculation expression can be obtained by using the B expansion value series different from the A expansion value series for the residual series by the first regression calculation expression.

First, in order to obtain the first regression calculation formula, the calculation unit 37 evaluates the correlation with the reference biometric value series for each of the plurality of development value series by regression analysis (step ST4).
FIG. 11 is an explanatory diagram of a first regression calculation formula for calculating the systolic blood pressure and a method for determining its coefficient.

  FIG. 11A is a table showing a correlation between a plurality of development value series and a series of reference biometric values. Here, a coefficient of an arithmetic expression for calculating a series of reference biological values (maximum blood pressure) from each developed value series is shown as a correlation value. In the figure, the expanded value series G7 has the largest correlation value (potential).

In this case, the computing unit 37 selects the expanded value series G7 showing the highest correlation as the explanatory series (step ST5), and computes the first regression equation and its coefficient in FIG. 11B (step ST6). ).
Thereby, the first regression calculation formula can be obtained.
FIG. 11B is a first regression calculation expression for calculating an evaluation biometric value from the developed value series G7.
A value obtained by converting eigenvalue V1 obtained from a plurality of expansion values constituting expansion value series G7 by cos (2X / 3) is assigned to Vat.
By obtaining the eigenvalue V1 from a plurality of development values, it is possible to calculate a value unique to the development value series G7 corresponding to the feature of the waveform. Here, the eigenvalue V1 refers to a value obtained by multiplying each of a plurality of expansion values constituting the expansion value series G7 by an individual coefficient and calculating the sum.
Further, by converting the eigenvalue V1 with cos (2X / 3), each value in the blood pressure range of 0 to 270 mmHg can be handled as different values normalized to a range of −1 to +1. For example, the eigenvalue corresponding to the waveform at a blood pressure of 179 mmHg and the eigenvalue corresponding to the waveform at a blood pressure of 181 mmHg can be handled as different values. Alternatively, for example, the eigenvalue corresponding to the waveform at a blood pressure of 170 mmHg and the eigenvalue corresponding to the waveform at a blood pressure of 190 mmHg can be handled as different values.
Thereby, the target value Vpa of the first regression calculation formula can be calculated. Generally, a residual remains between the target value Vpa and the evaluation biological value.

  Next, the computing unit 37 computes the residual of the target value of the first regression equation with respect to the reference biological value, and generates a residual series (first residual series) of the first regression equation (step ST7). .

FIG. 12 is an explanatory diagram of a second regression calculation formula for reducing the residual according to the first regression calculation formula of FIG. 11 and a method for determining its coefficient.
Then, as shown in FIG. 12A, for each developed value series, the correlation with the residual series of the first regression calculation formula is calculated (step ST8).
Further, the developed value series G2 showing the highest correlation is selected as an explanatory series (step ST9), and the second regression calculation formula and its coefficient are calculated as shown in FIG. 12B (step ST10).
Thereby, a 2nd regression equation can be obtained.
FIG. 12B is a second regression calculation formula for calculating the residual series of the first regression calculation formula from the developed value series G2.
A value obtained by converting the eigenvalue V2 obtained from a plurality of expansion values constituting the expansion value series G2 by sin (V2-Veam) is substituted for Vbt. Here, Beam is the minimum residual value of the first regression equation Vpa.
By obtaining the eigenvalue V2 from the plurality of development values, it is possible to calculate a value unique to the development value series G2.

Next, the calculating part 37 calculates the residual (second residual series) of the target value of the first regression calculation formula and the target value of the second regression calculation formula with respect to the reference biological value (step ST11).
Then, it is determined whether the maximum residual of the second residual series is equal to or less than a desired allowable residual. If it is equal to or smaller than the desired allowable residual, the regression calculation formula determination process in FIG. 5 is terminated.

On the other hand, if it is not less than the desired allowable residual, the processes from step ST7 to step ST11 are repeated until the desired residual is less than or equal to the regression calculation formula and coefficient determination process.
For example, by repeating the regression calculation formula and coefficient determination process six more times, the sum of the calculation results of the plurality of regression calculation formulas and the residual with respect to the reference biological value can be reduced to ± 1% or less.

FIG. 13 is an explanatory diagram of a third regression calculation formula for reducing the residual by the second regression calculation formula of FIG. 12 and a method for determining its coefficient.
In this case, the computing unit 37 selects the developed value series G7 based on the correlation with the residual series based on the second regression equation for the plurality of developed value series in FIG. The third regression equation of B) is obtained.
In FIG. 13, “e-07” means that 10 ⁻⁷ is applied as a scale. The same applies to the other drawings thereafter.
FIG. 13B is a third regression calculation formula for calculating the residual series of the second regression calculation formula from the developed value series G7.
A value obtained by converting eigenvalue V3 obtained from a plurality of expansion values constituting expansion value series G7 by sin (V3-Vebm) is substituted for Vct. Here, Vebm is the minimum residual value of the second regression equation Vpb.
By obtaining the eigenvalue V3 from the plurality of development values, it is possible to calculate a unique value for the development value series G7.

FIG. 14 is an explanatory diagram of a fourth regression calculation formula for reducing the residual according to the third regression calculation formula of FIG. 13 and a method for determining its coefficient.
In this case, the computing unit 37 selects the developed value series G12 based on the correlation with the residual series based on the third regression equation for the plurality of developed value series in FIG. The fourth regression equation of B) is obtained.
FIG. 14B is a fourth regression calculation formula for calculating the residual series of the third regression calculation formula from the developed value series G12.
A value obtained by converting eigenvalue V4 obtained from a plurality of expansion values constituting expansion value series G12 by sin (V4−Vecm) is substituted for Vdt. Here, Vecm is the minimum residual value of the third regression equation Vpc.
By obtaining the eigenvalue V4 from a plurality of development values, it is possible to calculate a value unique to the development value series G12.

FIG. 15 is an explanatory diagram of a fifth regression calculation formula for reducing the residual based on the fourth regression calculation formula of FIG. 14 and a method for determining its coefficient.
In this case, the computing unit 37 selects the developed value series G7 based on the correlation with the series of residuals according to the fourth regression equation for the plurality of developed value series in FIG. The fifth regression equation of B) is obtained.
FIG. 15B is a fifth regression calculation formula for calculating a residual series of the fourth regression calculation formula from the expanded value series G7.
In Vet, a value obtained by converting eigenvalue V5 obtained from a plurality of expansion values constituting expansion value series G7 by sin (V5-Vedm) is substituted. Here, Vedm is the minimum residual value of the fourth regression equation Vpd.
By obtaining the eigenvalue V5 from the plurality of development values, it is possible to calculate a unique value for the development value series G7.

FIG. 16 is an explanatory diagram of a sixth regression calculation formula for reducing the residual according to the fifth regression calculation formula of FIG. 15 and a method for determining its coefficient.
In this case, the calculation unit 37 selects the expanded value series G2 based on the correlation with the residual series based on the fifth regression calculation formula for the plurality of expanded value series in FIG. The sixth regression equation of B) is obtained.
FIG. 16B is a sixth regression calculation expression for calculating the residual series of the fifth regression calculation expression from the expanded value series G2.
A value obtained by converting eigenvalue V6 obtained from a plurality of expansion values constituting expansion value series G2 with sin (V6-Vefm) is substituted for Vft. Here, Vefm is the minimum residual value of the fifth regression equation Vpe.
By obtaining the eigenvalue V6 from the plurality of development values, it is possible to calculate a unique value for the development value series G2.

FIG. 17 is an explanatory diagram of a seventh regression calculation formula for reducing the residual by the sixth regression calculation formula of FIG. 16 and a method for determining its coefficient.
In this case, the computing unit 37 selects the developed value series G12 based on the correlation with the residual series according to the sixth regression equation for the plurality of developed value series shown in FIG. The seventh regression equation of B) is obtained.
FIG. 17B is a seventh regression operation expression for calculating a residual series of the sixth regression operation expression from the expanded value series G12.
A value obtained by converting eigenvalue V7 obtained from a plurality of expansion values constituting expansion value series G12 by sin (V7−Vegm) is substituted for Vgt. Here, Vegm is the minimum residual value of the sixth regression equation Vpf.
By obtaining the eigenvalue V7 from a plurality of development values, it is possible to calculate a value unique to the development value series G12.

FIG. 18 is an explanatory diagram of an eighth regression calculation formula for reducing the residual according to the seventh regression calculation formula of FIG. 17 and a method for determining its coefficient.
In this case, the computing unit 37 selects the developed value series G7 based on the correlation with the residual series based on the seventh regression equation for the plurality of developed value series in FIG. The eighth regression equation of B) is obtained.
FIG. 18B is an eighth regression calculation formula for calculating a residual series of the seventh regression calculation formula from the expanded value series G7.
A value obtained by converting eigenvalue V8 obtained from a plurality of expansion values constituting expansion value series G7 by sin (V8−Vehm) is substituted for Vht. Here, Vehm is the minimum residual value of the seventh regression equation Vpg.
By obtaining the eigenvalue V8 from the plurality of development values, it is possible to calculate a unique value for the development value series G7.

  As shown in FIGS. 11 to 18, the developed value series used as the explanatory variable in the nth regression calculation formula is different from the developed value series used as the explanatory variable in the (n−1) th regression calculation formula. At least some of the regression calculation expressions in the plurality of regression calculation expressions use a development value series different from the development value series used in the other regression calculation expressions.

FIG. 19 is a diagram for explaining the calculation formula of the systolic blood pressure (evaluation biological value) and the relationship between the reference biological value.
FIG. 19A is an arithmetic expression of the evaluation biological value (maximum blood pressure). The calculation unit 37 adds the target value Vpa of the first regression calculation formula to the target value Vph of the eighth regression calculation formula.
FIG. 19B is an explanatory diagram illustrating a relationship between target values of a plurality of regression calculation formulas and reference biological values.
As shown in the figure, by adding the target values Vpa,..., Vph of a plurality of regression calculation expressions, it is possible to calculate an evaluation biological value that falls within the reference biological value and the allowable residual.
Further, in the figure, ΔVpa is a residual of the target value Vpa of the first regression calculation expression with respect to the reference biological value. ΔVpb is a residual of the sum of the target value Vpa of the first regression calculation formula and the target value Vpb of the second regression calculation formula with respect to the reference biological value. ΔVpc is a residual of the sum of the target value Vpa of the first regression calculation formula, the target value Vpb of the second regression calculation formula, and the target value Vpc of the third regression calculation formula with respect to the reference biological value.

FIG. 20 is a diagram for explaining a plurality of regression calculation formulas for calculating the minimum blood pressure and a calculation formula for the minimum blood pressure.
Nine regression calculation formulas Vpa to Vpi shown in FIG. 20 are determined by the process of FIG. 5 in the same manner as the determination method of a plurality of regression calculation formulas for calculating the systolic blood pressure.
However, when calculating the diastolic blood pressure, a plurality of expansion value eigenvalues constituting the expansion value series are calculated, converted by a trigonometric function, and used in the regression calculation formula.
Then, for example, the calculation unit 37 obtains a plurality of development value series similar to FIG. 9B, and further, a series of reference biological values (diastolic blood pressure) for each development value series as in FIG. 11A. The correlation with is calculated. Then, the eigenvalue of the expanded value series showing the highest correlation is selected as the explanatory series, and the first regression calculation formula Vpa and its coefficient in FIG. 20 are calculated.
FIG. 20 also shows an arithmetic expression for the evaluation biological value (minimum blood pressure). The calculation unit 37 adds the target value Vpa of the first regression calculation formula to the target value Vpi of the ninth regression calculation formula, and calculates an evaluation biological value (minimum blood pressure).

The data of the above arithmetic expression is stored in the memory 31 of FIG. 4 as part of the program. The calculation unit 37 reads the program and executes it to calculate the evaluation biological value in real time using the calculation formula obtained by the above regression analysis.
The sensor device 2 in FIG. 1 detects a pulsating flow in a state where the wristband 6 is worn on a human body, and transmits the pulsating flow value as a detected biological value to the portable terminal 7. The memory 31 of the portable terminal 7 accumulates and stores the received detected biological value (pulsating current value).
Note that the data of the arithmetic expression stored in the memory 31 is not generated by the arithmetic unit 37 described above, but may be generated by another computer device or the like.

  FIG. 21 is a flowchart illustrating an example of an evaluation biometric value calculation method. The calculation unit 37 repeatedly executes the process of FIG.

  In FIG. 21, the calculating part 37 acquires a detected biometric value from the memory 31 (step ST21), and extracts a plurality of feature values shown in FIG. 6 for a waveform represented by a plurality of detected biometric values (step ST22). .

  Next, the calculation unit 37 substitutes the plurality of feature values into the expansion calculation formula of FIG. 7 to generate a plurality of expansion value series (step ST23). In this real-time calculation, the values used in the regression analysis may be used as they are for the feature values Y0 to Y19. Even in this case, since the plurality of feature values X1 to X7 are based on the detected biological values detected in real time, the developed value series can also be based on the detected biological values detected in real time.

  Next, the calculation unit 37 calculates the systolic blood pressure as the evaluation biological value. In this case, the eigenvalues of the expansion value series corresponding to the combinations are substituted as explanatory variables in the regression calculation expressions shown in FIGS. 11 to 18 to calculate a plurality of target values (step ST24). As described above, a value based on the expanded value series having the highest correlation is used as the explanatory value used in the calculation of each regression equation. When the calculation process is completed for all the regression calculation formulas shown in FIGS. 11 to 18, the calculated target values are added to calculate the sum (step ST25). The value of this sum is the maximum blood pressure value as the evaluation biological value.

  Similarly, the calculation unit 37 calculates the minimum blood pressure in the process of FIG. In this case, for each of a plurality of regression calculation expressions shown in FIG. 20, a corresponding combination of expanded value series is substituted as an explanatory variable to calculate a plurality of target values. When the calculation process is completed for all of the regression calculation formulas shown in FIG. 20, the calculated target values are added to calculate the sum. The value of this sum is the minimum blood pressure value as the evaluation biological value.

  The memory 31 may store the evaluation biometric value generated by the calculation unit 37.

FIG. 22 is a flowchart illustrating an example of a method for outputting an evaluation biometric value.
The computing unit 37 repeatedly executes the process of FIG. 22, for example, following the process of FIG.

  The calculating part 37 acquires an evaluation biological value from the memory 31, for example (step ST31). The calculating part 37 displays the acquired evaluation biological value on the liquid crystal device 35 (step ST32).

  Moreover, the calculating part 37 acquires the evaluation reference value about the acquired evaluation biological value from the memory 31, for example, and compares these (step ST33). The evaluation reference value includes, for example, a reference value for maximum blood pressure and a reference value for minimum blood pressure. Further, a reference range may be used instead of the evaluation reference value.

  When the evaluation biological value is an abnormal value exceeding the evaluation reference value (step ST34), the calculation unit 37 executes an emergency communication process (step ST35). The calculation unit 37 causes the wireless communication unit 21 to transmit emergency communication data. The emergency communication data includes, for example, evaluation biometric data. The emergency communication data is received by the emergency device 9 via the communication network 8. The emergency device 9 performs emergency notification. As a result, the ambulance crew can know that an abnormality has occurred in the person wearing the wristband 6. You can rush to the site as needed.

  Thereby, a user's blood pressure etc. are detected, and when there exists a predetermined change, an emergency call is performed. Thus, for example, as in the emergency notification system of Patent Document 2, an emergency call can be made even when the user is not conscious as compared to the case where the user makes an emergency call by pressing a button switch.

  As described above, in the present embodiment, a sensor device 2 that detects a value that changes with the life activity of a living body, and a detected biometric value is acquired by performing data communication with the sensor device 2 to obtain a plurality of detected living organisms. By performing data communication with the computing device 3 and the computing device 3 that computes an evaluation biological value related to a living body by a computation process using a plurality of feature values indicating a waveform feature represented by a value and a plurality of regression computation expressions And an output device 4 that outputs the obtained evaluation biometric value or a determination result about the evaluation biometric value.

  In particular, the arithmetic device 3 stores data of a plurality of regression equations whose coefficients have been determined by regression analysis, and reads the data of the plurality of regression equations from the storage unit and executes the calculation process of the evaluation biological value And an arithmetic unit 37. Then, the calculation unit 37 uses a plurality of acquired biometric values to generate a plurality of feature values indicating the features of the waveform represented by the plurality of detected biometric values, and a calculation that combines the plurality of feature values. Using the eigenvalues of the expanded value series, in which a combination is determined for each regression calculation expression in the regression analysis, as the explanatory value, the process of calculating multiple expanded value series and the target value that is the result of each regression calculation expression And a process of calculating the sum of a plurality of target values, and calculating the sum of the plurality of target values as an evaluation biological value.

  Therefore, for example, without directly measuring the evaluation biological value such as blood pressure, the detected biological value such as a blood flow value that changes with the life activity of the living body and based on the detected biological value that has a correlation with the evaluation biological value in the living body Thus, an evaluation biological value such as blood pressure can be obtained. Moreover, the judgment result about this evaluation biometric value can be output as an emergency call, for example.

  Moreover, the wristband 6 is provided with a band portion 11 for mounting on a living body, a light receiving transistor 13 provided on the band portion 11 for detecting a change occurring in the living body at the mounting portion, and a detection detected by the light receiving transistor 13. And a wireless communication unit 21 that performs data communication of the biometric value to the arithmetic device 3. Therefore, it is not necessary to provide the wristband 6 with devices such as a liquid crystal monitor, a built-in speaker, and a built-in microphone. The wristband 6 can be reduced in weight and size, and the cost can be reduced.

  Moreover, if an elderly person has the portable terminal 7, the biological value calculation system 1 can be utilized only by using the wristband 6. Even if the user is unfamiliar with Internet operation or setting of a new device, the biological value calculation system 1 can be easily used using existing resources.

  In addition, it is difficult to wear a dedicated device such as a blood pressure monitor at all times, but the simple and lightweight wristband 6 can be worn all the time. Constant monitoring is possible.

  Further, since the wristband 6 and the like can be formed at a low cost, it can be used as a triage device that measures the severity of the victim in real time in the event of a disaster.

  In addition, the sensor device 2 does not directly measure blood pressure, for example, but may detect a biological value having a correlation with the blood pressure, for example, blood flow of the arm, and an evaluation biological value such as blood pressure based on the detected biological value. Can be obtained. In addition, in the calculation process for obtaining the evaluation biometric value from the detected biometric value, a calculation process using a plurality of feature values indicating the features of the waveform represented by the plurality of detected biometric values and a plurality of regression equations is used. Therefore, for example, even if the correlation between the detected biological value and the evaluated biological value is not high or the detection accuracy of the detected biological value is not high, an evaluated biological value having a high probability can be obtained.

  On the other hand, when the evaluation biometric value is obtained by a normal method, for example, correction processing for correcting the detected biometric value, the detection accuracy of the detected biometric value directly affects the accuracy of the evaluation biometric value. In this case, in order to obtain an accurate evaluation biometric value, a high correlation is required between the detected biometric value and the evaluation biometric value, and detection accuracy of the detected biometric value is required to be high. In the present embodiment, since the detected biological value is obtained from the detected biological value by using the regression calculation formula based on the characteristic of the waveform instead of using the detected biological value as it is, these restrictions are eased. For example, even if a detected biometric value is obtained using an inexpensive light-receiving transistor 13 with low accuracy, an evaluation biometric value having a high probability can be obtained.

  Thus, in the biological value calculation system 1 of the present embodiment, the burden on the user is reduced. Therefore, it can be expected to spread widely.

[Second Embodiment]
Next, the biological value calculation system 1 according to the second embodiment will be described. In the following description, differences from the first embodiment will be mainly described. Moreover, each structure is demonstrated using the same code | symbol as 1st embodiment.
In the first embodiment, the sensor device 2 detects a pulsating flow value that changes with the heartbeat of the living body as a value that changes with the life activity of the living body in the regression calculation. 19 peak frequencies in the frequency distribution, with maximum, minimum, maximum speed, minimum speed, maximum acceleration, minimum acceleration, and pulse rate in the waveform represented by multiple pulsatile values , Calculated as a feature value. In real-time detection, the 19 peak frequencies in the frequency distribution are fixed feature values.
In addition to this, for example, the computing device 3 computes a plurality of frequency values (component ratio values) in a frequency distribution of a waveform represented by a plurality of pulsating flow values as feature values to be computed in regression computation and real-time detection. May be.

FIG. 23 is an explanatory diagram of another example of a plurality of feature values.
The horizontal axis in FIG. 23 is the frequency, and the vertical axis is the value of the component ratio for each frequency.
Here, the calculation device 3 calculates the values of the component ratios of the six frequencies from a to f that take the maximum value as a plurality of feature values. This is used instead of the above-described X1 to X6.

FIG. 24 is an explanatory diagram showing the correlation between each of the plurality of feature value series of FIG. 23 and a reference biological value (maximum blood pressure).
In FIG. 24, a graph showing a correlation with a reference living body value (maximum blood pressure) for a series of component ratios of six maximum frequencies from a to f is shown. In each graph, the horizontal axis represents the component ratio series value, and the vertical axis represents the reference biological value (maximum blood pressure).
As shown in the figure, each of the six maximum frequency component ratio series from a to f shows a good correlation with the series of reference biological values (maximum blood pressure). In particular, a series of five component ratios from a to e shows a very good high correlation.

Therefore, a plurality of regression calculation equations can be obtained based on between a plurality of maximum frequency component ratio series and a reference biological value (maximum blood pressure) series, as in the first embodiment. And the arithmetic unit 3 can calculate an evaluation biological value (maximum blood pressure) by using such a plurality of regression calculation expressions.
For example, as shown in FIG. 4, the arithmetic unit 3 stores a memory 31 that stores data of a plurality of regression equations whose coefficients are determined by regression analysis, and reads data of the plurality of regression equations from the memory 31. And a calculation unit 37 that executes value calculation processing.
And the calculating part 37 produces | generates the series of the some feature value which shows the characteristic of the waveform represented by the some detected biological value using the some detected biological value acquired from the sensor apparatus 2, and each regression calculation type | formula The target value that is the result of the above calculation is calculated using the characteristic value series or the eigenvalue of which the combination is determined for each regression calculation formula in the regression analysis as an explanatory value, and the sum of a plurality of target values is calculated. In this case, the calculation unit 37 calculates the sum of a plurality of target values as an evaluation biometric value.

Note that the diastolic blood pressure can also be calculated as an evaluation calculation value from the detected biological value by a similar method.
In the first embodiment, since the feature value series having a lower correlation with the reference biometric value series is used than in the second embodiment, the calculation process for expanding the feature value series into the expanded value series is repeatedly performed. ing. As shown in FIG. 24, the feature value series of the second embodiment itself has a high correlation with the reference biometric value series. In this case, as in the second embodiment, it can be expected that a good evaluation biological value is calculated by using a regression calculation expression having the characteristic value series or its eigenvalue as an explanatory value. In addition, if high accuracy of the evaluation biometric value with respect to the reference biometric value is not required, the regression calculation formula is used, which reduces the number of expansion calculation processes or uses a feature value series that has a low correlation with the reference biometric value series as an explanatory value. You can do it. Thus, the presence / absence of the calculation process for expanding the feature value series into the expanded value series and the number of calculations may be determined according to a request for accuracy with respect to the evaluation biological value.

[Application example]
Next, an application example of the biological value calculation system 1 of the above-described embodiment will be described.

FIG. 25 is an explanatory diagram of one application example of the biological value calculation system 1 of FIG.
The biological value calculation system 1 in FIG. 25 includes a wristband 6, a mobile terminal 7, and a communication network 8 as in FIG. 1.
The sensor device 2 is mounted on the wristband 6. The sensor device 2 may be worn anywhere on the human body as long as the waveform associated with the heartbeat can be detected.
The mobile terminal 7 is mounted with the arithmetic device 3 and the output device 4. The calculation device 3 calculates the maximum blood pressure, the minimum blood pressure, and the pulse rate from the detected biological value and displays them on the liquid crystal device 35. Further, when an abnormal value is reached, emergency communication is performed by e-mail or the like. Note that the arithmetic device 3 and the output device 4 only need to have the above functions, and may be a microcomputer board or the like.
For data communication between the wristband 6 and the portable terminal 7, short-range wireless communication such as Bluetooth (registered trademark) may be used. The wristband 6 and the portable terminal 7 may perform data communication by relay-type communication via a network. By enabling relay-type communication, data communication can be performed even if the mobile terminal 7 is remote so that it cannot communicate directly with the mobile terminal 7.
The portable terminal 7 may be connected to the communication network 8 by wireless communication such as WiFi (IEEE 802).

FIG. 26 is an explanatory diagram of another application example of the biological value calculation system 1 of FIG.
In the biological value calculation system 1 of FIG. 26, the sensor device 2, the calculation device 3, and the output device 4 are mounted on the wristband 6. Data communication can be performed between the sensor device 2 and the arithmetic device 3 by inter-program communication or the like. Further, when the evaluation biometric value becomes an abnormal value, emergency communication is performed by e-mail or the like.
For data communication between the wristband 6 and the portable terminal 7, short-range wireless communication such as Bluetooth (registered trademark) may be used. The wristband 6 and the portable terminal 7 may perform data communication by relay-type communication via a network.
The portable terminal 7 may be connected to the communication network 8 by wireless communication such as WiFi.

FIG. 27 is an explanatory diagram of another application example of the biological value calculation system 1 of FIG.
The biological value calculation system 1 in FIG. 27 has a biological value calculation device 41. The biological value calculation device 41 includes the sensor device 2, the calculation device 3, and the output device 4. A light emitting diode 12 and a light receiving transistor 13 as a sensor are provided at the tip of a cord extending from the apparatus main body 42. A sensor detects the detected biological value which changes with the life activity of a living body by being applied to the surface of a human body.

  Although the above embodiment is an example of an embodiment suitable for the present invention, the present invention is not limited to this, and various modifications or changes can be made without departing from the scope of the invention.

DESCRIPTION OF SYMBOLS 1 ... Biometric value calculation system 2 ... Sensor apparatus 3 ... Calculation apparatus 4 ... Output device (output part)
6 ... wristband 7 ... portable terminal 8 ... communication network 9 ... emergency device 11 ... band part (mounting part)
12 ... Light emitting diode 13 ... Light receiving transistor (sensor)
DESCRIPTION OF SYMBOLS 14 ... Load resistance element 15 ... Capacitor 16 ... Resistance element 17 ... Operational amplifier 18 ... Capacitor 19 ... Resistance element 20 ... Integrated circuit 21 ... Wireless communication part 31 ... Memory 32 ... CPU
33 ... Timer 34 ... Wireless communication unit 35 ... Liquid crystal device 36 ... Speaker 37 ... Calculation unit (comparison unit, reporting unit)
41 ... Biometric value calculation device 42 ... Device main body

The present invention is, for example, biological value calculation system for calculating an evaluation biological value related organisms such as blood pressure, the biological value calculation device, program, and a recording medium.

The biological value calculation system according to the present invention includes a sensor device that detects a pulsating flow value that detects blood flow as a value that changes with the life activity of a living body, and a detected biological value by data communication with the sensor device. And calculating the blood pressure of the living body as the evaluation biological value related to the living body by the arithmetic processing using the plurality of characteristic values indicating the characteristics of the waveform represented by the plurality of detected living body values and the plurality of regression equations. an arithmetic unit for the have a, and an output device for outputting a determination result for the arithmetic unit and the evaluation biological value or the evaluation biological value obtained by data communication, the arithmetic unit, the coefficient by regression analysis A storage unit that stores data of the plurality of determined regression equations, and a calculation unit that reads the data of the plurality of regression equations from the storage unit and executes the calculation process of the evaluation biological value And the calculation unit generates a plurality of feature values indicating the characteristics of the waveform represented by the plurality of detected biological values using the plurality of detected biological values acquired from the sensor device; In addition, a combination of a process for calculating a plurality of developed value series by a calculation combining a plurality of the characteristic values and a target value which is a calculation result of each regression calculation formula is determined for each regression calculation formula in the regression analysis. A process of calculating using the expanded value series or its eigenvalue as an explanatory value and a process of calculating the sum of a plurality of target values are calculated, and the sum of the target values is calculated as the evaluation biological value The plurality of regression calculation formulas used for the calculation for obtaining one evaluation biometric value sequentially reduce a residual with a series of reference biometric values obtained when the evaluation biometric value is detected from a living body. Selected Next, it is a combination of a plurality of arithmetic expressions used for the calculation, and the expanded value series used in the calculation of each regression calculation expression is different from the expanded value series used in the calculation of the previous regression calculation expression. Is.

Preferably, the characteristic values of the multiple, the maximum value of the waveform represented by a plurality of pulsating flow, minimum, maximum velocity value, velocity minimum, maximum acceleration, the acceleration minimum, and pulse rate It is good that it is at least one value among them.

Another biological value calculation system according to the present invention is detected by data communication with a sensor device that detects a pulsating flow value that detects blood flow as a value that changes with the life activity of a living body. Blood pressure of a living body is obtained as an evaluation biological value related to the living body through a calculation process using a plurality of feature values indicating a characteristic of a waveform represented by a plurality of detected living body values and a plurality of regression calculation formulas. And an output device that outputs the evaluation biometric value obtained by data communication with the arithmetic device or a determination result about the evaluation biometric value. A storage unit that stores data of a plurality of regression calculation formulas for which coefficients have been determined, and an operation that reads the plurality of regression calculation formula data from the storage unit and executes the calculation process of the evaluation biological value The calculation unit uses a plurality of the detected biometric values acquired from the sensor device to generate a series of feature values indicating a feature of the waveform represented by the plurality of detected biometric values. A process of generating and a target value, which is a calculation result of each regression calculation formula, using the series of feature values or combinations of the characteristic values determined for each regression calculation formula in the regression analysis as explanatory values And a process for calculating a sum of a plurality of target values, calculating a sum of the plurality of target values as the evaluation biometric value, and using a plurality of calculations used to obtain one evaluation biometric value The regression calculation formula is a combination of a plurality of calculation formulas that are selected so as to sequentially reduce a residual with a series of reference biometric values obtained when the evaluation biometric value is detected from a living object and that are sequentially used for the calculation. Yes, each said Sequence of the feature values used in the calculation of the return operation expression is different from said characteristic value sequence used in the calculation of before the regression expression.

The biological value calculation apparatus according to the present invention acquires a detected biological value as data from a sensor that detects a pulsating flow value that detects blood flow as a detected biological value that changes with the life activity of a living organism. A calculation unit that calculates blood pressure of the living body as an evaluation biological value related to the living body by a calculation process using a plurality of feature values indicating the characteristics of the waveform represented by the plurality of detected biological values and a plurality of regression calculation formulas ; acquires the evaluation biometric value as data from the arithmetic unit, have a, and an output unit for outputting a determination result for the evaluation biological value or the evaluation biological value, the calculation unit, the coefficient is determined by regression analysis Using the storage unit that stores the plurality of regression equation data, the plurality of regression equation data is read from the storage unit, and the evaluation biological value calculation process is executed. Using a plurality of the detected biometric values acquired from the device, a process for generating a plurality of feature values indicating the features of the waveform represented by the plurality of detected biometric values, and an operation combining the plurality of the feature values A process for calculating a plurality of expansion value series and a target value, which is the calculation result of each regression calculation expression, is used as an explanatory value for the expansion value series or a unique value thereof that is determined for each regression calculation expression in a regression analysis. And calculating the sum of a plurality of target values, calculating the sum of the plurality of target values as the evaluation biometric value, and obtaining one evaluation biometric value The plurality of regression equations used in the above are selected to sequentially reduce a residual with a series of reference biological values obtained when the evaluation biological value is detected from a living organism, and are sequentially used in the calculation. Performance A combination of the formula, the expansion value sequence used in the operation of each said regression arithmetic expression is different from said expansion value sequence used in the calculation of before the regression expression.

The program according to the present invention includes a step of acquiring a pulsating flow value in which blood flow is detected as a detected biological value that changes with the life activity of a living body, and a waveform represented by the acquired plurality of detected biological values. A step of calculating blood pressure of a living body as an evaluation biological value related to the living body by a calculation process using a plurality of characteristic values indicating features and a plurality of regression equations, and the calculated evaluation biological value or the evaluation biological value And a step of outputting the determination result of the computer to the computer, wherein the step of calculating includes a plurality of regression calculation formula data whose coefficients are determined by regression analysis, from a storage unit that stores a plurality of regression calculation formula data. The data of the regression calculation formula is read and the calculation process of the evaluation biological value is executed, and a plurality of the detected biological values acquired from the sensor device are used. A process for generating a plurality of feature values indicating features of a waveform represented by a plurality of the detected biological values, a process for calculating a plurality of development value series by an operation combining the plurality of feature values, A process of calculating a target value, which is a calculation result of a regression calculation formula, using the expanded value series or a unique value thereof defined as a combination for each regression calculation formula in a regression analysis as an explanatory value, and a plurality of target values A plurality of regression calculation formulas used to calculate one sum of the plurality of target values as the evaluation biological value and obtain one evaluation biological value. A combination of a plurality of arithmetic expressions that are selected so as to sequentially reduce a residual with a series of reference biometric values obtained when a biological value is detected from an organism, and that are sequentially used for the calculation. Performance The expansion value sequences used in is different from said expansion value sequence used in the calculation of before the regression expression.

And the calculating part 37 produces | generates the some feature value which shows the characteristic of a waveform about these waveforms. The feature values X1 to X7 become different values when the waveform changes. That is, the characteristic value takes a unique value corresponding to the waveform.
The feature values X1 to X7 only need to have a certain correlation with the waveform of the evaluation biological value. Preferably, the feature value is less likely to change due to factors other than the evaluation biological value.
The pulsating flow waveform in FIG. 6A is generated as a maximum value X1 and a minimum value X4 of the waveform as feature values indicating the peak of the waveform.
The pulsating velocity waveform in FIG. 6B is generated as a maximum velocity value X2 and a minimum velocity value X5 of the waveform as feature values indicating the peak of the waveform.
The pulsating acceleration waveform of FIG. 6C is generated as a maximum acceleration value X3 and a minimum acceleration value X6 of the waveform as feature values indicating the peak of the waveform.
For example, the pulse value X7 is generated based on the time interval between two adjacent waveform peaks in the pulsating acceleration waveform of FIG. 6C.
For the frequency distribution in FIG. 6D, the component ratio values of the 19 peak frequencies are extracted as feature values Y1 to Y19.
The feature value Y0 = 1 is added to these to obtain a total of 27 feature values .

Furthermore, the calculation unit 37 generates a plurality of second expansion value series in FIG. 9B using the first expansion value series and the second expansion calculation expression group in FIG. 9A.
The group of second expansion arithmetic expressions in FIG. 9A includes an involute curve function Fa, a clothoid curve function (X side) Fb, and a clothoid curve function (Y side) Fc . By evaluating and convert these functions can improve the significance of the plurality of expansion value series.
The plurality of second expanded value series in FIG. 9B is 15 series from G1 to G15.
In the subsequent processing in the present embodiment, the plurality of second development value series are used as a plurality of development value series.

Claims (15)

A sensor device for detecting a value that changes in accordance with the life activity of a living body;
The life value is obtained by performing a calculation process using a plurality of feature values and a plurality of regression calculation expressions indicating a characteristic of a waveform represented by a plurality of detected biological values by acquiring a detected biological value by performing data communication with the sensor device. An arithmetic device for calculating an evaluation biological value related to the body;
An output device that outputs the evaluation biometric value or the evaluation biometric value obtained by data communication with the arithmetic device; and
A biological value calculation system.
The arithmetic unit is:
A storage unit that stores data of a plurality of regression calculation formulas whose coefficients are determined by regression analysis, and a calculation unit that reads the data of the plurality of regression calculation formulas from the storage unit and executes the calculation process of the evaluation biological value And having
The computing unit is
Using a plurality of the detected biological values acquired from the sensor device, a process of generating a plurality of feature values indicating the characteristics of the waveform represented by the plurality of detected biological values;
Processing to calculate a plurality of expansion value series by calculation combining a plurality of the feature values;
A process for calculating a target value, which is a calculation result of each regression calculation formula, using the expanded value series or a unique value thereof as a descriptive value for each regression calculation formula in a regression analysis;
A process of calculating a sum of a plurality of the target values;
To calculate the sum of a plurality of the target values as the evaluation biometric value,
The biological value calculation system according to claim 1.
Each regression equation and its coefficient are:
In the regression analysis,
Substituting a plurality of feature values into a plurality of expansion arithmetic expression groups that are prepared in advance from a plurality of different development arithmetic expressions to generate a plurality of expansion value series by calculating a plurality of expansion values,
For each of the plurality of development value series, evaluate the correlation with the series of reference biological values or the residual series obtained when the evaluation biological value is detected from the living organism,
It is determined based on the correlation between the developed value series having the highest correlation and the series of reference biometric values or the series of residuals,
The explanatory values used for the calculation of each regression equation are as follows:
Obtained based on a plurality of the detected biological values acquired from the sensor device, the development value series having the highest correlation or its eigenvalue,
The biological value calculation system according to claim 2.
The plurality of regression equations are
A combination of a plurality of arithmetic expressions selected so as to sequentially reduce a residual with a series of reference biological values obtained when the evaluation biological value is detected from a living organism.
The biological value calculation system according to claim 2 or 3.
The plurality of regression equations are at least
In regression analysis, the expanded value series selected from the plurality of expanded value series that are evaluated to have the highest correlation with the reference biological value series obtained when the evaluated biological value is detected from the living organism A first regression equation for converting the above into a series of reference biometric values,
In regression analysis, selected from among a plurality of the developed value series, the correlation between the series of reference biological values and the series of residuals of the developed value series by the first regression equation is the highest. A second regression equation for converting the expanded value series to the residual series;
Including
The expanded value series used as explanatory variables in the second regression equation is different from the expanded value series used as explanatory variables in the first regression equation,
The biological value calculation system according to any one of claims 2 to 4.
In at least some of the regression arithmetic expressions among the plurality of regression arithmetic expressions, the expanded value series different from the expanded value series used in the other regression arithmetic expressions is used.
The biological value calculation system according to any one of claims 2 to 5.
The evaluation biological value is the highest blood pressure or the lowest blood pressure of the living body,
The detected biological value is a pulsating flow value in which blood flow is detected,
The plurality of characteristic values are at least one of a maximum value, a minimum value, a maximum speed value, a minimum speed value, a maximum acceleration value, a minimum acceleration value, and a pulse rate in a waveform represented by the plurality of pulsating flow values. Value,
The biological value calculation system according to any one of claims 2 to 5.
The sensor device detects a value that changes with the heartbeat of the living body as the value that changes with the life activity of the living body,
The biological value calculation system according to claim 7.
The arithmetic unit is:
A storage unit that stores data of a plurality of regression calculation formulas whose coefficients are determined by regression analysis, and a calculation unit that reads the data of the plurality of regression calculation formulas from the storage unit and executes the calculation process of the evaluation biological value And having
The computing unit is
Using a plurality of the detected biological values acquired from the sensor device, a process of generating a series of a plurality of feature values indicating the characteristics of the waveform represented by the plurality of detected biological values;
A process of calculating a target value, which is a calculation result of each regression calculation formula, using the series of characteristic values or combinations of the characteristic values determined for each regression calculation formula in regression analysis as explanatory values;
A process of calculating a sum of a plurality of the target values;
To calculate the sum of a plurality of the target values as the evaluation biometric value,
The biological value calculation system according to claim 1.
The sensor device includes:
A sensor for detecting a value that varies with the life activity of the life form;
A wireless communication unit that performs data communication of the detected biological value detected by the sensor to the arithmetic device;
Have
The biological value calculation system according to any one of claims 1 to 9.
The output device is
A comparison unit that compares the biological value acquired by data communication from the arithmetic device with an evaluation reference value;
If the comparison result by the comparison unit is bad or not good, a reporting unit that automatically reports,
Having
The biological value calculation system according to any one of claims 1 to 10.
A sensor for detecting a detected biological value that changes with the life activity of the living body;
The detected biological value is obtained as data from the sensor, and the biological value evaluated with respect to the living organism is obtained by a calculation process using a plurality of feature values indicating a feature of the waveform represented by the plurality of detected biological values and a plurality of regression equations. A computing unit for computing
Obtaining the evaluation biometric value as data from the computing unit, and outputting the evaluation biometric value or a determination result about the evaluation biometric value;
A biological value calculation device having
A method of generating an evaluation biometric value by detecting a detection biometric value that changes with the life activity of a living organism by a sensor device,
Obtaining the detected biological value from the sensor device;
By calculating using a plurality of feature values indicating the characteristics of the waveform represented by the plurality of acquired detected biological values and a plurality of regression calculation formulas, an evaluation biological value related to the living organism is calculated,
Outputting the calculated evaluation biometric value or the evaluation biometric value;
Biological value calculation method.
Obtaining a detected biometric value that changes with the life activity of the life form;
A step of calculating an evaluation biometric value related to the living organism by a calculation process using a plurality of feature values indicating a feature of the waveform represented by the plurality of acquired biometric values acquired and a plurality of regression calculation formulas;
Outputting the calculated evaluation biological value or the evaluation result of the evaluation biological value;
A program that causes a computer to execute.
The computer-readable recording medium which recorded the program of Claim 14.

Images