JP2009195590A - Biological information processing apparatus and method and program for controlling biological information processing apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To carry out a calculation using a pulse rate accurately by reducing an influence caused by different noise components due to different kinds of exercises performed by a user wearing a biological information processing apparatus. <P>SOLUTION: The biological information processing apparatus detects the pulse rate by a pulse sensor 30 and a body motion sensor 302 and sets a filter coefficient in accordance with the kind of the exercise when eliminating the noise by using an adaptive filter in a signal processing circuit 305. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

The present invention relates to a biological information processing apparatus, a biological information processing apparatus control method, and a control program, and more particularly, to a biological information processing apparatus and a biological information processing apparatus capable of estimating calorie consumption more accurately for each subject (user). The present invention relates to a control method and a control program.

2. Description of the Related Art Conventionally, a biological information processing apparatus that is worn by a user and processes biological information such as the user's pulse rate is known (see, for example, Patent Document 1).
Some of such biological information processing apparatuses use an adaptive filter to remove a user's body motion component as noise.
Japanese Patent No. 3250622

In this case, depending on the type of exercise, the noise component (body motion component) also differs. Therefore, if an adaptive filter using a certain parameter is used, the noise cannot be removed successfully depending on the type of exercise. There is a risk that the pulse rate to be inaccurate will be inaccurate, and furthermore, processing based on this pulse rate, for example, calculation of calorie consumption may not be performed.
Therefore, an object of the present invention is to reduce the influence of different noise components included in the pulse wave signal due to different types of exercise performed by the user wearing the biological information processing apparatus,
To provide a biological information processing apparatus, a biological information processing apparatus control method, and a control program capable of accurately detecting a pulse rate and thus performing an accurate calculation using the pulse rate, such as calculating calorie consumption. It is in.

In order to solve the above-described problem, a first aspect of the present invention is a biological information processing apparatus that is worn by a user and detects and processes biological information during exercise by the user, and detects a pulse wave component to detect a pulse. A pulse wave detection unit that outputs a wave detection signal, a body motion component detection unit that detects a body motion component and outputs a body motion component detection signal, a predetermined filter coefficient, the pulse wave detection signal, and the body motion component detection signal An adaptive filter unit that removes the body motion component from the pulse wave detection signal based on the above, an exercise type setting unit for allowing the user to input and set the type of exercise, and according to the set type of exercise And a filter coefficient setting unit for setting the filter coefficient.
According to the above configuration, the pulse wave detection unit detects a pulse wave component and outputs a pulse wave detection signal.
The body motion component detection unit detects the body motion component and outputs a body motion component detection signal.
On the other hand, the exercise type setting unit allows the user to input and set the type of exercise, and the filter coefficient setting unit sets the filter coefficient according to the determined type of exercise.
As a result, the adaptive filter unit removes the body motion component from the pulse wave detection signal using a filter coefficient corresponding to the type of exercise.
Therefore, the influence of different noise components included in the pulse wave detection signal due to different types of exercise can be reduced, and the body motion component can be accurately removed.
According to a second aspect, in the first aspect, the filter coefficient set in the filter coefficient setting unit may be a step size.
According to a third aspect, in the first aspect or the second aspect, the sensor constituting the body motion component detection unit is an at least biaxial acceleration sensor worn on a user's arm, and the body motion The component is detected as an acceleration component in the direction in which the arm extends in the acceleration component detected by the acceleration sensor, and an acceleration component in a direction that is parallel to the back of the user's hand and orthogonal to the direction in which the arm extends. May be.
According to a fourth aspect, in the first to third aspects, a pulse rate is calculated based on an output signal of the adaptive filter unit, and an oxygen intake is calculated based on the calculated pulse rate. The biological information may be calculated based on the oxygen intake.
According to a fifth aspect, in the fourth aspect, the calculated biological information may be calorie consumption.
According to a sixth aspect, a pulse wave sensor that detects a pulse wave component and outputs a pulse wave detection signal, a body motion sensor that detects a body motion component and outputs a body motion component detection signal, and the pulse wave detection signal An adaptive filter that removes the body motion component from the biological information processing apparatus that is worn by the user and processes the biological information of the user, and inputs the type of exercise performed by the user, It is characterized by comprising an exercise type setting process for setting and a filter coefficient setting process for setting a filter coefficient of the adaptive filter in accordance with the set type of exercise.
According to the above configuration, the influence of different noise components included in the pulse wave detection signal due to different types of exercise can be reduced, and the body motion component can be accurately removed.
The seventh aspect includes a pulse wave sensor that detects a pulse wave component and outputs a pulse wave detection signal, a body motion sensor that detects a body motion component and outputs a body motion component detection signal, and the pulse wave detection signal An adaptive filter that removes the body motion component from the computer, and a control program for controlling, by a computer, a biological information processing apparatus that is worn by the user and processes the biological information of the user. The type of motion is input and set, and the filter coefficient of the adaptive filter is set according to the set motion type.
According to the above configuration, the influence of different noise components included in the pulse wave detection signal due to different types of exercise can be reduced, and the body motion component can be accurately removed.

According to the present invention, it is possible to reduce the influence on the biological information processing due to the difference in the noise component included in the pulse wave signal due to the different types of exercises performed by the user, and to accurately calculate the pulse rate. Thus, calculation using the pulse rate can be accurately performed, such as calorie consumption calculation.

Next, a preferred embodiment of the present invention will be described.
FIG. 1 is an explanatory diagram illustrating a configuration of the biological information processing apparatus according to the embodiment.
FIG. 2 is a cross-sectional view of the vicinity of the pulse sensor of the biological information processing apparatus.
In the present embodiment, the pulse is used as the biological information, the pulse rate is detected as the biological information value, and the calorie consumption is calculated from the detected pulse rate.
The biological information processing apparatus 1 is roughly classified into a wristwatch-type apparatus main body 10, a cable 20 connected to the apparatus main body 10, and a tip side of the cable 20, and is closely attached to a little finger by a sensor fixing band 40. And a pulse sensor 30 (see FIG. 2) fixed so as to be configured.
The apparatus main body 10 is provided with a wristband 12 that is wound around the arm from the twelve o'clock direction of the wristwatch and fixed in the six o'clock direction. By this wristband 12, the apparatus main body 1
0 is detachably attached to the arm.

As shown in FIG. 2, the pulse sensor 30 is mounted between the base of the little finger and the finger joint while being shielded from light by the sensor fixing band 40. Thus, by attaching the pulse sensor 30 to the base of the finger, the cable 20 can be shortened, so the cable 20 does not get in the way during running. Further, when the distribution of the body temperature from the palm to the fingertip is measured, the temperature of the fingertip is remarkably lowered when it is cold, but the temperature at the base of the finger is not relatively lowered. Therefore, if the pulse sensor 30 is attached to the base of the finger, the pulse rate and the like can be accurately measured even when running outdoors on a cold day. The finger wearing the pulse sensor 30 is not limited to the little finger, and may be another finger.

FIG. 3 is a plan view showing the main body 10 of the biological information processing apparatus 1 with the wristband, cable, etc. removed, and FIG. 4 is a side view of the biological information processing apparatus 1 as viewed from the 3 o'clock direction on the wristwatch. is there.
In FIG. 3, the apparatus main body 10 includes a resin watch case 11 (main body case). On the surface side of the watch case 11, in addition to the current time and date, a liquid crystal display device 13 (display device) with an EL backlight that displays pulse wave information such as the pitch and the pulse rate during running and walking, etc. Is provided.
The liquid crystal display device 13 includes a first segment display area 131 located on the upper left side of the display surface,
A second segment display area 132 located on the upper right side, a third segment display area 133 located on the lower right side, and a dot display area 134 located on the lower left side are configured. This information can be displayed graphically.
A body motion sensor 302 (see FIG. 6) for obtaining a pitch is built in the watch case 11, and an acceleration sensor or the like can be used as the body motion sensor 302.

In addition, a control unit 5 that performs various types of control and data processing is provided inside the watch case 11.
The control unit 5 detects the detection result (body motion signal) by the body motion sensor 302 and the pulse sensor 30.
Based on the detection result (pulse wave signal), the pulse rate, and thus the calorie consumption, is calculated, and the calorie consumption of the user who is the subject is displayed on the liquid crystal display device 13.
In this case, since the control unit 5 is also configured with a timer circuit, the normal time can be displayed on the liquid crystal display device 13.
Further, as shown in FIG. 4, button switches 111 to 115 for performing external operations such as time adjustment and display mode switching are provided on the outer peripheral portion (side surface portion) of the watch case 11. In addition, large button switches 116 and 117 that are assumed to be operated during exercise are configured on the front surface of the watch case.

The power source of the biological information processing apparatus 1 is a button-shaped small battery 59 (see FIG. 3) built in the watch case 11, and the cable 20 supplies power from the battery 59 to the pulse sensor 30, and The detection result of the sensor 30 is input to the control unit 5 of the watch case 11.
In the biological information processing apparatus 1, it is necessary to increase the size of the apparatus main body 10 as its functions are increased. However, since the apparatus main body 10 is restricted to be worn on the arm, the apparatus main body 10 cannot be expanded toward the 6 o'clock and 12 o'clock directions on the wristwatch.
Therefore, in the present embodiment, as shown in FIG. 4, the apparatus main body 10 has a horizontally long watch case in which the length in the 3 o'clock and 9 o'clock directions is longer than the length in the 6 o'clock and 12 o'clock directions. 11 is used.
In this case, since the wristband 12 is connected at a position biased toward the 3 o'clock direction, when viewed from the wristband 12, the 9 o'clock direction of the wristwatch is different from the 3 o'clock direction. A portion 101 is provided. Therefore, instead of using the horizontally long watch case 11, the wrist can be bent freely, and even if it falls, the back of the hand does not hit the watch case 11.

Inside the watch case 11, a buzzer flat piezoelectric element 58 is arranged in the direction of 9 o'clock with respect to the battery 59 as shown by a one-dot chain line in FIG. The battery 59 includes the piezoelectric element 5
Since it is heavier than 8, the position of the center of gravity of the apparatus main body 10 is biased toward the 3 o'clock direction. Since the wristband 12 is connected to the side where the center of gravity is biased, the apparatus main body 10 can be attached to the arm in a stable state. Further, since the battery 59 and the piezoelectric element 58 are arranged in the plane direction, the apparatus main body 10 can be thinned.
At the same time, as shown in FIG. 4, the user can easily replace the battery 59 by providing a battery lid 118 on the back surface portion 119.
In FIG. 4, a connecting portion 105 for holding a stop shaft 121 attached to the end of the wristband 12 is formed in the 12 o'clock direction of the watch case 11. Watch case 11
In the 6 o'clock direction, the wristband 12 wound around the arm is folded back at an intermediate position in the length direction, and a receiving portion 106 to which a fastener 122 for holding the intermediate position is attached is formed. .

In the 6 o'clock direction of the apparatus body 10, a portion from the back surface portion 119 to the receiving portion 106 is formed integrally with the watch case 11 and forms an angle of about 115 ° with respect to the back surface portion 119. It has become. That is, when the apparatus main body 10 is mounted by the wristband 12 so as to be positioned on the upper surface L1 (back of the hand) of the left wrist L (arm), the watch case 11
The back surface portion 119 closely contacts the upper surface portion L1 of the wrist L. In parallel with this, the rotation stopper 108
Is in contact with the side surface portion L2 having the rib R.
In this state, the back surface portion 119 of the apparatus main body 10 feels like straddling the radius R and the ulna U. At the same time, it feels to come into contact with the rib R from the bent portion 109 of the rotation stop portion 108 and the back surface portion 119 to the rotation stop portion 108. Thus, the rotation stopper 108 and the back surface 119
Is an anatomically ideal angle of about 115 °, so that even if the device body 10 is rotated in the direction of arrow A or arrow B, the device body 10 is unnecessary around the arm. There is no deviation.
In addition, the apparatus main body 1 is provided at two positions on one side around the arm by the back surface portion 119 and the rotation stop portion 108.
It only regulates zero rotation. For this reason, even if the arm is thin, the back surface portion 119 and the rotation stop portion 108 are surely in contact with the arm, so that the rotation stop effect can be reliably obtained. Furthermore, even if the arm is thick, there is no cramped feeling.

FIG. 5 is a cross-sectional view of the pulse sensor 30 of the embodiment.
In FIG. 5, the pulse sensor 30 has a back cover 4 on the back side of a sensor frame 36 as a case body.
By covering 08, the component storage space 400 is formed inside. A circuit board 35 is arranged inside the component storage space 400. The circuit board 35 includes an LED 31,
A phototransistor 32 and other electronic components are mounted. An end of the cable 20 is fixed to the pulse sensor 30 by a bush 493, and each wiring of the cable 20 is soldered onto a pattern of each circuit board 35. Here, the pulse sensor 30 is connected to the cable 2
It is attached to the finger so that 0 is pulled out from the base side of the finger to the apparatus main body 10 side. Therefore, the LED 31 and the phototransistor 32 are arranged along the length direction of the finger. Among them, the LED 31 is located on the tip side of the finger, and the phototransistor 32 is located on the base of the finger. Such an arrangement has an effect that it is difficult for external light to reach the phototransistor 32.

In the pulse sensor 30, a light transmission window is formed by a light transmission plate 34 made of a glass plate on the upper surface portion (substantially pulse wave signal detection unit) of the sensor frame 36. The LED 31 and the phototransistor 32 have a light emitting surface and a light receiving surface on the light transmitting plate 34, respectively.
For those who are For this reason, when the finger surface is brought into close contact with the outer surface 441 (contact surface / sensor surface with the finger surface) of the translucent plate 34, the LED 31 emits light toward the finger surface side. At the same time, the phototransistor 32 can receive light reflected from the finger side of the light emitted from the LED 31. Here, in order to improve the adhesion between the outer surface 441 of the translucent plate 34 and the finger surface, the outer surface 441 of the translucent plate 34 has a structure protruding from the peripheral portion 461 thereof.

In the present embodiment, the LED 31 is an InGaN system (indium-gallium-nitrogen system).
Blue LED, and its emission spectrum has an emission peak at 450 nm. Furthermore, the emission wavelength region of the LED 31 is in the range from 350 nm to 600 nm.
In this example, a GaAsP-based (gallium-arsenic-phosphorus-based) phototransistor is used as the phototransistor 32 in correspondence with the LED 31 having such light emission characteristics. The light receiving wavelength region of the phototransistor 32 itself has a main sensitivity region of 300 nm to 600 nm.
There is also a sensitivity region at 300 nm or less.

When the pulse sensor 30 configured in this way is attached to the base of the finger by the sensor fixing band 40 and light is emitted from the LED 31 toward the finger in this state, the light reaches the blood vessel and is caused by hemoglobin in the blood. Part of the light is absorbed and part is reflected. The light reflected from the finger (blood vessel) is received by the phototransistor 32, and the change in the amount of received light changes the blood volume (
Corresponding to blood pulse wave). That is, when the blood volume is large, the reflected light is weakened, while when the blood volume is small, the reflected light becomes strong. Therefore, if a change in the reflected light intensity is detected, various biological information including the pulse rate can be measured.

In the present embodiment, the wavelength region from about 300 nm to about 600 nm, which is the overlapping region of the light emission wavelength region of the LED 31 and the light reception wavelength region of the phototransistor 32, that is,
Biological information is displayed based on the detection result in a wavelength region of about 700 nm or less.
The reason for adopting such a configuration is that, even when external light is exposed to the finger, light having a wavelength region of 700 nm or less out of the light included in the external light has the finger as a light guide and the phototransistor 32.
This is because it does not reach (light receiving part). This is because the wavelength region included in external light is 700 nm.
This is because the following light tends to hardly pass through the finger. Therefore, even if external light is applied to the finger portion not covered with the sensor fixing band 40, the phototransistor 3 passes through the finger.
It does not reach 2 and does not affect the measurement result.
Further, since pulse wave information is obtained using light in a wavelength region of about 700 nm or less, the S / N ratio of the pulse wave signal based on the blood volume change is high. The reason for this is that hemoglobin in blood has an absorption coefficient of 8 for light having a wavelength of 300 nm to 700 nm, which is a conventional detection light.
This is considered to be several times to about 100 times or more larger than the extinction coefficient for light of 80 nm.
Therefore, since the blood volume changes sensitively, the pulse wave detection rate based on the blood volume change (S / N
Ratio) is considered to be high.

FIG. 6 is a schematic configuration block diagram around the control unit.
The control unit 5 is roughly divided into a pulse wave data processing unit 500 that calculates a pulse rate and the like based on an input result from the pulse sensor 30, and a pitch data processing unit that calculates a pitch based on an input result from the body motion sensor 302. 501; a clock generation unit 502 that generates an operation clock signal;
And a control unit 503 for controlling the entire control unit.
The pulse wave data processing unit 500 is roughly divided into a pulse wave signal amplifying circuit 303 and a pulse wave waveform shaping circuit 306, and is shared with the pitch data processing unit 501 to share a signal processing circuit (DSP).
305.
The pulse wave signal amplification circuit 303 amplifies the pulse wave signal that is the output of the pulse sensor 30 and outputs the pulse wave amplification signal to the signal processing circuit 305 and the pulse wave waveform shaping circuit 306.
The pulse wave waveform shaping circuit 306 shapes the waveform of the pulse wave amplification signal and outputs it to the control unit 503.

The signal processing circuit 305 in the pulse wave data processing unit 500 performs signal processing of the pulse wave amplification signal and outputs the signal as pulse wave data to the control unit 503.
The pitch data processing unit 501 roughly includes a body motion signal amplifying circuit 304 and a body motion waveform shaping circuit 307, and is shared with the pulse wave data processing unit 500 as described above to share the signal processing circuit 305. It has.
The body motion signal amplification circuit 304 amplifies the body motion signal that is the output of the body motion sensor 302 and outputs the body motion amplification signal to the signal processing circuit 305 and the body motion waveform shaping circuit 307.
The body motion waveform shaping circuit 307 performs waveform shaping of the body motion amplification signal and outputs it to the control unit 503.
The signal processing circuit 305 in the pitch data processing unit 501 performs signal processing of the body motion amplification signal and outputs it to the control unit 503 as body motion data.

The clock generation unit 502 includes an oscillation circuit 312 and a frequency dividing circuit 313, when roughly classified.
The oscillation circuit 312 includes a crystal oscillator and the like, and supplies the clock signal to the control unit 503 as a reference operation clock and supplies it to the frequency dividing circuit 313 so as to generate a clock signal for timing from the clock signal.
The frequency dividing circuit 313 divides the supplied clock signal, generates various clock signals for timing, and supplies them to the control unit 503.
The control unit 503 is roughly divided into an MPU 308, a RAM 309, and a ROM 31.
0 and a communication unit 311.

The MPU 308 controls the entire control unit 5 and thus the entire biological information processing apparatus 1 based on a control program stored in the ROM 310.
The RAM 309 temporarily stores various data including pulse wave data and body motion data, and is used as a work area.
The ROM 310 stores in advance a control program for controlling the MPU 308 and, consequently, the biological information processing apparatus 1 as a whole.
The communication unit 311 transmits / receives data to / from an external device connected via a communication connector under the control of the MPU 308. That is, it is possible to output measurement data to an external device or input setting data of the biological information processing apparatus 1 from the external device.

Here, the signal processing circuit 305 will be described in detail.
FIG. 7 is a schematic configuration block diagram of the signal processing circuit 305.
The signal processing circuit 305 is broadly classified into an A / D conversion circuit 30 that performs A / D conversion of the input pulse wave amplification signal or body motion amplification signal and outputs it as original pulse wave data or body motion data.
5A and an adaptive filter circuit 305B that receives the original pulse wave data and the body motion data, removes a noise component (body motion component) included in the original pulse wave data based on the body motion data, and outputs it as pulse wave data; It is equipped with.

FIG. 8 is a block diagram showing the principle configuration of the adaptive filter circuit.
The adaptive filter circuit 305B includes noise estimated to be included in body motion data x (k) input at a certain timing and original pulse data m (k) input at the same timing based on the impulse response h. That is, the filter unit 401 that generates the estimated noise data y (k) and the estimated noise data y (k) are subtracted from the input original pulse wave data m and output as the pulse wave data e (k) that is the target data. And a filter coefficient setting unit 403 that sets a filter coefficient including the step size μ based on the pulse wave data e (k).
With the above configuration, the filter unit 401 of the adaptive filter circuit 305B has noise estimated to be included in the original pulse data m (k) based on the input body motion data x (k) and the impulse response h, that is, , Estimated noise data y (k) is generated and output to the subtraction unit 402.
The subtraction unit 402 subtracts the estimated noise data y (k) from the input original pulse wave data m and outputs it as pulse wave data e (k) that is target data.
In parallel with this, the filter coefficient setting unit 403 sets a filter coefficient including the step size μ based on the pulse wave data e (k).

Here, the setting of the step size μ in the filter coefficient setting unit 403 will be considered.
As step size μ becomes smaller, the correction of the filter weight will be smaller for each sample, and the error will slowly decrease. On the other hand, when the step size μ is increased, the weight for each step increases and the error decreases rapidly.
The resulting error may not be close to the ideal solution, i.e. unfocused.
Therefore, in order to ensure a good convergence ratio and stability, it is necessary to select an optimal step size μ within a predetermined range.
Therefore, in order to examine the setting of the optimum step size μ, the inventors tap the tap coefficient parameter TAP, which is a weight when taking a weighted average, based on data sampled for each of a plurality of types of exercises described later. While appropriately changing the parameter αx and parameter βx corresponding to the X-axis acceleration for setting the step size μ and the parameter αyz and parameter βyz corresponding to the Y-axis / Z-axis combined acceleration for setting the step size μ, the pulse The detection rate and the S / N improvement rate by the filter were examined.

Details will be described below.
The following five types of exercise were examined.
(Exercise Type 1) Normal Walking Normal walking exercise in which the pulse and pitch (body motion) do not overlap.
(Exercise type 2) Normal running This is a normal running exercise in which the pulse and pitch (body motion) do not overlap.
(Exercise type 3) Overlapping movements Overlapping movements in which the pulse and pitch (body motion) overlap, such as walking movements,
This includes the exercise of a user having a sports heart.
(Exercise type 4) Gripping exercise When an acceleration sensor, which is a body motion sensor, is worn on the arm, the exercise is performed with the arm fixed, and the noise does not appear in the acceleration and the pulse contains noise It is exercise. For example,
This includes exercises by motorcycle rowing and handrail treadmills.
(Exercise type 5) Pulse or pitch sudden change exercise This is an exercise in which the pulse rate or pitch (body motion) changes abruptly and a characteristic spectrum is hard to appear. For example, a motion on a slope, an interval motion, and a motion of climbing up and down stairs are included.
In this study, we examined the interval motion and the stair climbing / falling motion for the sudden movement of the pulse or pitch.

Next, the step size μ will be described.
The step size μ is expressed by the following equation.
μ = α / (β + P (n))
Where P (n) = Σx (n) ^{^ 2}
P is power, n is the number of data, x is a noise component, and n = 0, that is,
P (0) = TAP.

FIG. 9 is an explanatory diagram of the setting values of each parameter.
In this study, tap coefficient parameter TAP, parameter αx, parameter βx,
After grasping a rough tendency by quality engineering in advance for each of the parameter αyz and the parameter βyz, as shown in FIG. 9, three values are set, and the pulse detection rate and the S / N improvement rate by the filter are set. The data was measured.
The pulse detection rate is a ratio at which a correct pulse is detected with respect to the entire measurement data, and is expressed by the following equation.
Detection rate = (number of pulse detection data / number of measurement data)
Also, the S / N improvement rate (improvement rate) by the filter is
Improvement ratio = power ratio SN3 after filtering / power ratio SN before filtering
3
Here, the power ratio SN3 is expressed by the following equation based on the result of FFT.
SN3 = the sum of the power of the frequency corresponding to the pulse rate and the frequencies before and after that

FIG. 10 is an explanatory diagram of the detection rate during normal walking.
FIG. 11 is an explanatory diagram of the S / N improvement rate (improvement rate) during normal walking.
FIG. 12 is an explanatory diagram of the detection rate during normal running.
FIG. 13 is an explanatory diagram of an improvement rate (improvement rate) of S / N during normal running.
FIG. 14 is an explanatory diagram of the detection rate during the overlapping movement.
FIG. 15 is an explanatory diagram of the S / N improvement rate (improvement rate) during the overlapping movement.
FIG. 16 is an explanatory diagram of a detection rate during a gripping exercise.
FIG. 17 is an explanatory diagram of an improvement ratio (improvement rate) of S / N during a gripping exercise.
FIG. 18 is an explanatory diagram of a detection rate during a pulse or pitch (body motion) sudden change exercise (interval).
FIG. 19 is an explanatory diagram of an improvement ratio of S / N during a pulse or pitch (body motion) sudden change exercise (interval).
FIG. 20 is an explanatory diagram of a detection rate during a pulse or pitch (body motion) sudden change exercise (step).
FIG. 21 is an explanatory diagram of an improvement ratio of S / N during a pulse or pitch (body motion) sudden change exercise (stairs).

For example, as shown in FIG. 12, the detection rate of the tap coefficient parameter TAP is improved when the value is small, and the detection rate is decreased when the values of the parameter αx and the parameter βx are small, as shown in FIG. It can be seen that the detection rate of the parameter αyz and the parameter βyz is most improved with respect to the parameter αyz and the parameter βyz.
In addition, the value described as old in FIG. 12 is a conventional comparative example when the tap coefficient parameter TAP is similarly changed.
As for the S / N improvement rate (improvement rate), as shown in FIG. 13, during normal running, the intermediate value for the tap coefficient parameter TAP has the highest improvement rate, and the parameter αx and the parameter βx As for the parameter α, as the value becomes smaller, the improvement rate decreases, and for the parameter αyz and the parameter βyz, it can be seen that the intermediate value has the highest detection rate.
In addition, the value described as old in FIG. 13 is a conventional comparative example when the tap coefficient parameter TAP is similarly changed.

Based on the results of FIGS. 10 to 21, it was qualitatively understood that the detection rate and the improvement rate are improved in the following cases except for the gripping motion.
Parameter αx → Small Parameter βx → Large Parameter αyz → Large Parameter βyz → Large In addition, it was found that in the gripping motion, the detection rate was improved in the following cases.
Parameter αx → Large Parameter βx → Small Parameter αyz → Small Parameter βyz → Small

Therefore, the filter coefficient setting unit 403 appropriately sets the tap coefficient parameter TAP, parameter αx, parameter βx, parameter αyz, and parameter βyz to the type of exercise set by the user based on the results shown in FIGS. It is set accordingly.
As a result, the noise-removed pulse wave data e (k) output from the adaptive filter circuit 305B can be obtained as more accurate pulse wave data because noise as an error is reduced according to the type of exercise. It is.

FIG. 22 is an explanatory diagram of the operation of the adaptive filter circuit.
FIG. 22A is an explanatory diagram of the original pulse wave data m (k).
In FIG. 22A, the upper part is a waveform diagram of a pulse wave signal corresponding to the original pulse wave data m (k), and the lower part is a spectrum diagram (vertical) of the original pulse wave data m (k) after FFT processing. Axis power, horizontal axis frequency).
FIG. 22B shows X-axis direction estimated noise data y (k) generated by the filter unit.
) It is explanatory drawing of x.
In FIG. 22B, the upper part is a waveform diagram of an X-axis direction estimated noise signal corresponding to X-axis direction estimated noise data y (k) x, and the lower part is estimated noise data y (k) after FFT processing.
It is a spectrum figure (vertical axis power, horizontal axis frequency) of x.
The filter unit 401 of the adaptive filter circuit generates X-axis direction estimated noise data y (k) x shown in FIG. 22B and outputs it to the subtractor 402.
As a result, the subtraction unit 402 calculates the X-axis direction estimated noise data y from the original pulse wave data m (k).
(K) x is subtracted to obtain X-axis direction noise-removed pulse wave data e (k) x.

FIG. 22C is an explanatory diagram of the X-axis direction noise-removed pulse wave data e (k) x.
In FIG. 22C, the upper part is a waveform diagram of a pulse wave signal corresponding to X-axis direction noise-removed pulse wave data e (k) x, and the lower part is X-axis direction noise-removed pulse wave data after FFT processing. e (k
) X spectrum diagram (vertical axis power, horizontal axis frequency).
FIG. 22D is an explanatory diagram of the noise-removed pulse wave data e (k).
In FIG. 22D, the upper part is a waveform diagram of a pulse wave signal corresponding to the noise-removed pulse wave data e (k), and the lower part is a spectrum diagram of the noise-removed pulse wave data e (k) after FFT processing. (
Vertical axis power, horizontal axis frequency).
In the same manner as described above, in the adaptive filter circuit 305B, the Y-axis direction estimated noise data y (k) y and the Z-axis direction estimated noise data y (k) z (or Y-axis-Z-axis combined estimated noise data y ( k) yz) is removed from the X-axis direction noise-removed pulse wave data e (k) x to obtain the final noise-removed pulse wave data e (k) as shown in FIG. It will be.

Here, prior to description of a specific calculation operation of calorie consumption, the calorie consumption calculation method of the present embodiment will be described.
First, it is assumed that the following requirements are satisfied as prerequisites in the application of the present embodiment.
-In this embodiment, since the calorie consumption is calculated from the oxygen intake, the pulse rate is within a range that can be handled as being proportional to the exercise intensity.

If the pulse rate is low or high, it may not be proportional. If the exercise rate is low and the pulse rate is low, the pulse rate has a significant psychological effect and exceeds the limit of oxygen uptake capacity. This is because the pulse rate may increase.
・ There are individual differences in pulse rate due to age, gender, exercise ability, etc.
Further, in the following description, the resting pulse rate is the lowest pulse rate measured in a resting position at the start of measurement and in a stable pulse rate state. However, the present invention is not limited to this as long as it can be achieved stably under the same conditions.

Next, the operation of the embodiment will be described.
FIG. 23 is a flowchart of calorie consumption calculation processing of the biological information processing apparatus according to the embodiment.
First, the MPU 308 of the biological information processing apparatus 1 measures the pulse rate HR (step S11).
).

Here, the measurement of the pulse rate HR will be described in detail.
First, the MPU 308 acquires output signals from the pulse sensor 30 and the body motion sensor 302.
Specifically, the pulse sensor 30 detects a pulse wave from the living body and outputs the detected pulse wave signal to the pulse wave signal amplification circuit 303. The pulse wave signal amplifier circuit 303 amplifies the input pulse wave signal and outputs it to the signal processing circuit 305 and the pulse wave waveform shaping circuit 306. The pulse wave shaping circuit 306
The pulse wave signal is shaped and output to the MPU 308.
On the other hand, the body motion sensor 302 detects the movement of the user and outputs the detected body motion signal to the body motion signal amplification circuit 304. The body motion signal amplifier circuit 304 amplifies the body motion signal, and the signal processing circuit 3
05 and the body motion waveform shaping circuit 307. The body motion waveform shaping circuit 307 generates a body motion signal Sx.
Is output to the MPU 308.
On the other hand, the signal processing circuit 305 performs A / D conversion on the pulse wave signal Sm and the body motion signal Sx by the A / D conversion circuit 305A, respectively, and the obtained original pulse wave data m (k) and body motion data x (
k) is output to the adaptive filter circuit 305B, and the body motion data x (k) is output to the MPU 30.
8 is output.
Thereby, the adaptive filter circuit 305B performs the original pulse wave data m (k) by the above-described procedure.
The estimated noise data y (k) is removed from the noise, and the noise-removed pulse wave data e (k) is converted into MPU3.
Output to 08.

Subsequently, the MPU 308 performs fast Fourier transform (FFT) processing based on the noise-removed pulse wave data e (k) and the body motion data x (k), and the noise-removed pulse wave data e (k) and the body motion data x ( The pulse component Fm and the body motion component Ft are extracted from the result of the FFT processing in k).
Next, the MPU 308 determines whether or not the amount of the body motion component is larger than a predetermined threshold value for determining whether or not the pulse rate can be calculated.
And, if the amount of body motion component is larger than a predetermined threshold for determining whether or not the pulse rate can be calculated, since the current pulse rate is too much, it is impossible. Measurement is impossible.
On the other hand, when the amount of the body motion component is equal to or less than a predetermined threshold for determining whether or not the pulse rate can be calculated, the original pulse component Fm is obtained by removing the body motion component Ft from the pulse component Fm. .

In particular,
Fm = Fm-Ft
Perform the process. That is, a frequency component that exists only in the pulse wave signal is extracted.

And let the largest frequency component in the extracted pulse component Fm be a pulse spectrum.
Next, the MPU 308 calculates the pulse rate HR based on the extracted pulse spectrum frequency.
Subsequently, the MPU 308 calculates the current oxygen intake VO ₂ by the equation (1) (step S13). In this case, when the MPU 308 calculates the current oxygen intake VO ₂ by the equation (1), the MPU 308 calculates the resting oxygen intake VO ₂ res by the Harris-Benedict equation described above.
t (= basal metabolic rate) is calculated.

続いて、ＭＰＵ３０８は、現在の酸素摂取量ＶＯ₂から消費カロリー量Ｃを算出する（
ステップＳ１７）。
具体的には、（２）式により消費カロリー量Ｃを算出する。

Based on the calculated resting oxygen intake VO ₂ rest, the current oxygen intake VO ₂
Is calculated by equation (4).

Subsequently, the MPU 308 calculates the calorie consumption C from the current oxygen intake VO ₂ (
Step S17).
Specifically, the calorie consumption C is calculated by equation (2).

And when this exercise state is continued for 1 hour, as the total calorie consumption CT,
CT = C × 60
, And the total calorie consumption CT (kcal / h) as the calculation result is displayed on the liquid crystal display device 13 as shown in FIG.

As described above, according to the present embodiment, when removing a body motion component (body motion data; noise component) from a pulse signal (pulse data), an optimum adaptive filter circuit parameter is determined depending on the type of motion. Since it is set, it is possible to calculate an accurate pulse rate regardless of the type of exercise.
As a result, when automatically calculating calorie consumption associated with exercise from the pulse rate, the user is more accurate and without the need for specialized knowledge and complicated and expensive measuring equipment, Calorie consumption can be grasped.
Further, if the pulse rate can be measured, the calorie consumption can be calculated only by the arithmetic processing, so that the apparatus can be downsized and the manufacturing cost can be reduced.
Furthermore, since the relative oxygen intake is used to calculate the calorie consumption, it is possible to calculate the calorie consumption with little error in consideration of individual differences among users, that is, differences in physical strength among users.

In the above description, the case where the control program for controlling the biological information processing apparatus is stored in the ROM in advance has been described. However, the control program is recorded in advance on a recording medium such as various magnetic disks, optical disks, and memory cards. However, it is also possible to read and install from these recording media. It is also possible to provide a communication interface, download the control program via a communication network such as the Internet or LAN, install and execute the program.

As described above, according to the present embodiment, the calorie consumption associated with exercise is automatically calculated from the pulse rate, so that the user does not need specialized knowledge and complicated and expensive measuring equipment. You can easily grasp the calorie consumption of the exercise you are doing.
Further, if the pulse rate can be measured, the calorie consumption can be calculated only by the arithmetic processing, so that the apparatus can be downsized and the manufacturing cost can be reduced.
Furthermore, since the relative oxygen intake is used to calculate the calorie consumption, it is possible to calculate the calorie consumption with little error in consideration of individual differences among users, that is, differences in physical strength among users.

It is explanatory drawing which shows the structure of the biometric information processing apparatus of embodiment. It is sectional drawing of the vicinity of the pulse sensor of a biological information processing apparatus. It is a top view which shows the apparatus main body of a biological information processing apparatus in the state which removed the wristband, the cable, etc. It is the side view which looked at the biological information processor from the direction of 3 o'clock in a wristwatch. It is sectional drawing of the pulse sensor of embodiment. It is a general | schematic structure block diagram around a control part. FIG. 2 is a schematic configuration block diagram of a signal processing circuit 305. It is a principle block diagram of an adaptive filter circuit. It is explanatory drawing of the setting value of each parameter. It is explanatory drawing of the detection rate at the time of normal walking. It is explanatory drawing of the improvement ratio (improvement rate) of S / N at the time of normal walking. It is explanatory drawing of the detection rate at the time of normal driving | running | working. It is explanatory drawing of the improvement ratio (improvement rate) of S / N at the time of normal driving | running | working. It is explanatory drawing of the detection rate at the time of an overlap motion. It is explanatory drawing of the improvement ratio (improvement rate) of S / N at the time of an overlap exercise | movement. It is explanatory drawing of the detection rate at the time of a grip exercise. It is explanatory drawing of the improvement ratio (improvement rate) of S / N at the time of a grip exercise. It is explanatory drawing of the detection rate at the time of a pulse or pitch (body motion) sudden change exercise (interval). It is explanatory drawing of the improvement ratio of S / N at the time of a pulse or pitch (body motion) sudden change exercise (interval). It is explanatory drawing of the detection rate at the time of a pulse or pitch (body motion) sudden change motion (stairs). It is explanatory drawing of the improvement ratio of S / N at the time of a pulse or a pitch (body motion) sudden change exercise | movement (stairs). It is operation | movement explanatory drawing of an adaptive filter circuit. It is a calorie consumption calculation process flowchart of the biological information processing apparatus of the embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Biological information processing apparatus, 5 ... Control part, 10 ... Apparatus main body, 11 ... Clock case, 12 ... Wristband, 13 ... Liquid crystal display device, 20 ... Cable, 30 ... Pulse sensor, 302 ... Body motion sensor, 303 ... Pulse wave signal amplifier circuit 304 ... body motion signal amplifier circuit 305 ... signal processing circuit 3
05A ... A / D conversion circuit, 305B ... adaptive filter circuit, 306 ... pulse waveform shaping circuit, 3
07 ... Body motion waveform shaping circuit, 308 ... MPU, 309 ... RAM, 310 ... ROM, 311 ...
Oscillating circuit, 312... Frequency dividing circuit, 401... Filtering unit, 402... Subtracting unit, 403 ... Filter coefficient setting unit, 500 ... Pulse wave data processing unit, 501 ... Pitch data processing unit, 502 ... Clock generation unit, 503 ... Control Department.

Claims (7)

A biological information processing apparatus that is worn by a user and detects and processes biological information during exercise by the user,
A pulse wave detector that detects a pulse wave component and outputs a pulse wave detection signal;
A body motion component detector that detects the body motion component and outputs a body motion component detection signal;
An adaptive filter unit for removing the body motion component from the pulse wave detection signal based on a predetermined filter coefficient, the pulse wave detection signal and the body motion component detection signal;
An exercise type setting unit for allowing the user to input and set the type of exercise;
A filter coefficient setting unit for setting the filter coefficient according to the set type of exercise;
A biological information processing apparatus comprising: The biological information processing apparatus according to claim 1,
The biological information processing apparatus, wherein the filter coefficient set in the filter coefficient setting unit is a step size. The biological information processing apparatus according to claim 1 or 2,
The sensor constituting the body movement component detection unit is an at least biaxial acceleration sensor worn on the user's arm,
The body motion component is detected as an acceleration component in the extension direction of the arm, an acceleration component parallel to the back of the user's hand and perpendicular to the extension direction of the arm among the acceleration components detected by the acceleration sensor. A biological information processing apparatus characterized by: The biological information processing apparatus according to any one of claims 1 to 3,
Calculating a pulse rate based on an output signal of the adaptive filter unit; calculating an oxygen intake amount based on the calculated pulse rate; and calculating biological information based on the oxygen intake amount Information processing device. The biological information processing apparatus according to claim 4,
The biological information processing apparatus characterized in that the calculated biological information is calorie consumption. A pulse wave sensor that detects a pulse wave component and outputs a pulse wave detection signal, a body motion sensor that detects a body motion component and outputs a body motion component detection signal, and removes the body motion component from the pulse wave detection signal An adaptive filter, and a control method of a biological information processing apparatus that is worn by a user and processes the biological information of the user,
An exercise type setting process for inputting and setting the type of exercise performed by the user;
A filter coefficient setting process for setting a filter coefficient of the adaptive filter according to the set type of exercise;
A control method for a biological information processing apparatus. A pulse wave sensor that detects a pulse wave component and outputs a pulse wave detection signal, a body motion sensor that detects a body motion component and outputs a body motion component detection signal, and removes the body motion component from the pulse wave detection signal A control program for controlling, by a computer, a biological information processing apparatus that is equipped with an adaptive filter and is mounted on a user and processes the biological information of the user,
Let the user input and set the type of exercise to be performed,
According to the set type of exercise, the filter coefficient of the adaptive filter is set.
A control program characterized by that.

Images