JP2013106837A - Heart rate detection method, heart rate detector, and mental stress measuring apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a heart rate detection method for correctly extracting an acceleration pulse wave signal from a volume pulse wave signal without amplifying only noise components by second order differential processing to obtain the variation amount of a pulse wave signal nor submerging thereby the original pulse wave signal in noise.SOLUTION: The heart rate detection method for detecting the heart rate by intensity of a light transmitting through or reflecting from a living body includes: a first process of converting the transmitted or reflected light into an electric signal, and subjecting the electric signal to a low frequency transmission filtering; a second process of subjecting the signal after passing through the first process to differential processing; a third process of subjecting the signal after passing through the second process to the low frequency transmission filtering; a fourth process of subjecting the signal after passing through the third process to the differential processing; a fifth process of subjecting the signal after passing through the fourth process to the low frequency transmission filtering; and a sixth process of detecting the peak value from the signal after passing through the fifth process, and assuming the interval of the peak value as a heartbeat interval.

Description

  The present invention relates to a heartbeat detection method, a heartbeat detection device, and a mental stress measurement device that process biological measurement data and measure a heartbeat waveform.

  When obtaining an index representing physical stress, it is known to use temporal fluctuations in the peak interval of the heartbeat waveform (this is referred to as RR interval). The frequency of the RR interval is analyzed, and the ratio of power in two different frequency ranges set in advance is used as an index representing stress. As a method of acquiring a heartbeat waveform, there is a method of measuring an action potential of a heart muscle, whereby a pulse wave for stress measurement can be acquired. As a simple measuring method, there is a method using light transmittance or reflectance of a specific part of the body. For example, as shown in FIG. 9, there is a method using the reflectance when the fingertip is irradiated with infrared rays or the transmittance when the earlobe is irradiated with infrared rays. In FIG. 9, 101 represents a human fingertip, 102 represents an infrared projector, and 103 represents an infrared receiver. The infrared light receiver 103 detects infrared light that is irradiated from the infrared light projector 102 and reflected by the fingertip. The infrared transmittance or reflectance of these specific parts is related to the amount of blood flowing through that part, and the waveform obtained by measuring the infrared transmittance or reflectance of these specific parts is called a volume pulse wave. As an alternative to a heartbeat signal representing myocardial activity, this volume pulse wave is second-order differentiated and an acceleration pulse wave representing the blood flow acceleration component is used.

  The above heart rate detection means is used in inventions such as a mental stress evaluation apparatus in Patent Document 1, an autonomic nervous system function evaluation method and system in Patent Document 2, and a stress sensor system in Patent Document 3.

  An example of the waveform of the volume pulse wave is shown in FIG. In FIG. 9, when the pulse wave signal obtained by the infrared receiver 103 is actually measured, the waveform as shown in FIG. 10 is often not obtained. This is because the change in reflected light due to blood flow is very small and susceptible to noise. Further, the pulse wave signal measured by the infrared light receiver 103 may be converted into digital data by A / D conversion. In this case, a conversion error or a quantization error is added. An example of an actual waveform is shown in FIG. Differentiation processing is performed to obtain the amount of change of the pulse wave signal in a minute time (Δt). However, when differentiation is performed, the noise component of the original signal is amplified in actual signal processing. Further, if the differential process for amplifying the noise is continuously performed twice, only the noise component may be amplified, and the original pulse wave signal is buried in the noise. For example, when the signal of FIG. 11 is differentiated, it becomes as shown in FIG. 12, and the noise is amplified and the signal component of the pulse wave is buried.

  Therefore, the present invention amplifies only the noise component by the second-order differentiation process for obtaining the amount of change of the pulse wave signal, and the volume pulse wave signal so that the original pulse wave signal is not buried in the noise. An object of the present invention is to provide a heartbeat detection method for correctly extracting an acceleration pulse wave signal from a heartbeat.

  In order to solve the above-described problems, a heartbeat detection method of the present invention is a heartbeat detection method for detecting a heartbeat based on the intensity of light transmitted or reflected through a living body, and converts the transmitted or reflected light into an electrical signal, A first step of applying a low-frequency transmission filter to the signal; a second step of performing a differentiation process on the signal after passing through the first step; and applying a low-frequency transmission filter to the signal after passing through the second step A third step, a fourth step for performing differentiation on the signal after passing through the third step, a fifth step for applying a low-frequency transmission filter to the signal after passing through the fourth step, And a sixth step of detecting a peak value from the signal after passing through the step and setting the interval between the peak values as a heartbeat interval.

  According to the present invention, by passing through a low-frequency transmission filter in a timely manner, only the noise component is amplified by the second-order differentiation process for obtaining the amount of change in the heartbeat signal, and the original heartbeat signal is buried in the noise. Can be prevented.

It is a flowchart which shows the procedure of the heart rate detection method of embodiment of this invention. It is a heartbeat waveform figure for every procedure processed by the heartbeat detection method of the embodiment of the present invention. It is a heartbeat waveform figure for every procedure when the cutoff frequency is 20 Hz processed by the heartbeat detection method of the embodiment of the present invention. It is a heart rate waveform figure for every procedure when the cutoff frequency is 3 Hz processed by the heart rate detection method of the embodiment of the present invention. It is the figure which showed the example which calculates | requires RR space | interval in the procedure of the heart rate detection method of embodiment of this invention. It is a flowchart which shows the procedure of the stress measurement based on the RR space | interval obtained by the heartbeat detection method of embodiment of this invention. It is a block diagram which shows the outline | summary of the heart rate detection apparatus of embodiment of this invention. It is a block diagram which shows the outline | summary of the stress measuring device of embodiment of this invention. It is a figure which shows the actual example of the heart rate waveform measurement method using the infrared reflectance to a person's fingertip. It is an example of the waveform figure of a volume pulse wave. It is an example of the actual waveform figure of a volume pulse wave. It is an example of a waveform diagram when the volume pulse wave signal is subjected to second-order differential processing.

  A heartbeat detection method according to an embodiment of the present invention will be described below with reference to FIG. Information obtained by blood flow measurement (step S1) is converted into digital data by an A / D converter (step S2). The low-pass filter 1 applies a low-frequency transmission filter to the digitized data (step S3). Next, differential processing 1 is performed (step S4), and a low-frequency transmission filter is applied by the low-pass filter 2 (step S5). Thereafter, differentiation processing 2 is performed (step S6), and a low-frequency transmission filter is applied by the low-pass filter 3 (step S7). The time interval between the waveform peaks is output based on the peak detection output from the second-order differentiated signal (step S8).

  An example of this processing is shown in FIG. When differentiation is performed, noise components are amplified. If low-pass filter processing is not performed after differentiation processing 1, noise will be amplified during the next differentiation. Therefore, noise components can be minimized even when multi-stage differential processing is performed by passing a low-pass filter after differential processing. That is, it is possible to prevent noise components from being amplified.

  Consider the cutoff frequency of the low-frequency transmission filter. If the cut-off frequency is too high, noise cannot be removed, and if the cut-off frequency is too low, the peak waveform is dulled and the accuracy of the peak interval is lowered. As an example, FIG. 3 shows waveforms obtained by the steps S1 to S8 described above when the cutoff frequency is 20 Hz. Moreover, the waveform obtained by each process when a cutoff frequency is 3 Hz is shown in FIG.

  A person's heartbeat interval is about 1 second, and the time of a peak showing the peak of an acceleration pulse wave as shown in FIG. 2 is about 1/10 of the heartbeat interval. Therefore, it is desirable that the cutoff frequency is 10 Hz. This cutoff frequency may be changed at any time by measuring the peak interval. For example, as the time (t) obtained by averaging the peak intervals for the past 10 beats, low-pass filter processing is performed using about 1/10 of the time (t / 10) as the cutoff frequency (fc = 10 / t). In this way, it is possible to more accurately detect the peak interval of the heartbeat whose interval changes. The processing may be simplified by setting the cutoff frequency to a fixed frequency of about 10 Hz. A peak value from the acceleration pulse wave signal is detected and the peak interval is obtained. An example is shown in FIG. The peak intervals t1, t2, t3,... Are sequentially obtained, and the intervals are output as RR intervals.

  As described above, by setting the cut-off frequency of the low-frequency transmission filter to 1/10 of the heartbeat interval, the correct acceleration pulse wave can be obtained without impairing the accuracy of peak position detection of the acceleration pulse wave while properly removing noise. Peak time (RR interval) can be obtained. In the present embodiment, the low-pass filter and differentiation processing are performed after A / D conversion, but the low-pass filter and differentiation processing may be performed in an analog manner before A / D conversion.

  The procedure of mental stress measurement according to the embodiment of the present invention will be described below. That is, stress measurement is performed based on the RR interval detected by the above procedure.

  The procedure for stress measurement is shown in FIG. The measured RR interval is converted into data in time series (step S11). For example, the data is converted into a value T1 at t2 seconds, a value T2 at t2 + t3 seconds, and a value T3 at t2 + t3 + t4 seconds. Note that this data is not equal in time interval. Therefore, the time-series data is converted so as to have equal time intervals (step S12). This is because when the power spectrum is obtained by performing frequency analysis (step S13), it is desirable that the data time intervals are equal data strings in order to perform frequency analysis at high speed. Thereafter, power in two frequency bands is calculated from the calculated power spectrum data (step S14). For example, one band LF (Low Frequency) has a band whose upper limit is 0.15 Hz to 0.2 Hz from which the DC component is removed, and another band HF (High Frequency) has an upper limit of 0.5 Hz from the upper limit of LF. This is the bandwidth to be used. The powers of the LF and HF bands are obtained respectively. For example, LF / HF is calculated as the ratio of LF to HF (step S15). The value calculated in this way is used as an index of stress (step S16).

  By taking the above procedure, a more accurate RR interval can be obtained and stress measurement resistant to noise can be performed. Note that the conversion in step S12 can be realized by up-sampling in which the data points are linearly approximated and the time interval is reduced. Upsampling may be performed by interpolating between data points using a spline function or a higher-order function. By making the time intervals equal, it is possible to perform Fourier transform by calculation processing represented by a fast Fourier transform algorithm.

  A heartbeat detection device 1 according to an embodiment of the present invention includes a heartbeat detection unit 11 that detects a heartbeat based on the intensity of light transmitted or reflected through a living body, and an A / D conversion unit that converts information obtained by the heartbeat detection unit 11 into digital data. 12, low-pass filters 13 to 15 that transmit a low-frequency transmission filter and remove high-frequency components, a first differential processing unit 16 and a second differential processing unit 17 that perform differential processing to obtain acceleration pulse waves, and peak time A peak detection output unit 18 that outputs an interval (RR interval) is provided. Note that the cutoff frequency of the low-frequency transmission filter may be variable (including the frequency variable unit 19), or may be fixed.

  In addition, the mental stress measurement device 50 including the heartbeat detection device 1 according to the embodiment of the present invention includes a time series data conversion unit 51 that converts RR intervals into time series data, and sets the time series data as equal time intervals. An equal time interval conversion unit 52 for converting the power spectrum data, a power spectrum data calculation unit 53 for calculating power spectrum data by frequency analysis, and a frequency band calculation unit 54 for calculating power of two frequency bands from the power spectrum data. A power ratio calculation unit 55 that calculates a power ratio of two frequency bands, and a stress value conversion unit 56 that converts the power ratio to a stress value corresponding to the power ratio are provided.

  Each of the above-described embodiments is a preferred embodiment of the present invention, and various modifications can be made without departing from the scope of the present invention.

DESCRIPTION OF SYMBOLS 1 Heartbeat detection apparatus 11 Heartbeat detection part 12 A / D conversion part 13-15 Low pass filter 1-3
DESCRIPTION OF SYMBOLS 16 1st differential process part 17 2nd differential process part 18 Peak detection output part 19 Frequency variable part 50 Mental stress measuring device 51 Time series data conversion part 52 Equal time interval conversion part 53 Power spectrum data calculation part 54 Frequency band calculation Unit 55 Power ratio calculation unit 56 Stress value conversion unit 101 Human fingertip 102 Infrared light projector 103 Infrared light receiver

Japanese Patent No. 4238321 JP 2003-190109 A JP 2007-289540 A

Claims (5)

A heart rate detection method for detecting a heart rate by the intensity of light transmitted or reflected through a living body,
Converts transmitted or reflected light into an electrical signal,
A first step of applying a low frequency transmission filter to the electrical signal;
A second step of differentiating the signal after passing through the first step;
A third step of applying a low-frequency transmission filter to the signal after passing through the second step;
A fourth step of performing differentiation on the signal after passing through the third step;
A fifth step of applying a low-frequency transmission filter to the signal after passing through the fourth step;
A heart rate detection method comprising: a sixth step of detecting a peak value from the signal after passing through the fifth step and setting an interval between the peak values as a heart rate interval.   The heartbeat detection method according to claim 1, wherein a cut-off frequency of the low-frequency transmission filter is set to a frequency about 1/10 of a heartbeat interval. A heartbeat detection unit that detects a heartbeat signal based on the intensity of light transmitted or reflected through a living body;
An A / D converter that converts the heartbeat signal detected by the heartbeat detector into an electrical signal;
A first low-frequency transmission filter that removes high-frequency components from the electrical signal;
A first differentiation processing unit for differentiating a signal transmitted through the first low-frequency transmission filter;
A second low-frequency filter for removing high-frequency components from the signal subjected to differentiation processing by the first differentiation processing unit;
A second differentiation processor for differentiating the signal transmitted through the second low-frequency filter;
A third low-frequency filter for removing high-frequency components from the signal subjected to differential processing by the second differential processing unit;
A heartbeat detection device comprising: a peak detection output unit that detects a peak value from a signal that has passed through the third low-frequency filter, and uses a peak value interval as a heartbeat interval.   It further comprises a frequency variable unit that sets a cut-off frequency of each of the first low-frequency filter, the second low-frequency filter, and the third low-frequency filter to a frequency that is about 1/10 of a heartbeat interval. The heartbeat detection device according to claim 3. A heartbeat detecting device according to claim 3 or 4,
A time-series data conversion unit that converts the peak value intervals into time-series data, and
An equal time interval conversion unit for converting the data converted into time series into equal time intervals;
A power spectrum data calculation unit for calculating power spectrum data by frequency analysis;
A frequency band calculation unit for calculating power of two frequency bands from the power spectrum data;
A power ratio calculator for calculating the power ratio of the two frequency bands;
A mental stress measurement apparatus comprising a stress value conversion unit that converts a stress value according to the power ratio.

Images