JP3039236B2 - Pulse wave analyzer - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a pulse wave analyzer for analyzing a pulse wave.

[0002]

2. Description of the Related Art In recent years, pulse diagnosis has attracted attention, and along with this, research for finding the health condition of a living body based on a pulse wave has been actively conducted. As a general waveform analysis method, there is a frequency analysis such as FFT, and a frequency analysis of a pulse wave using this type of frequency analysis technology is being studied.

[0003]

By the way, the waveform of the pulse wave is not the same shape, but changes from moment to moment, and the wavelength of each wave is not constant. Absent. A method of performing a Fourier transform or the like by regarding such a chaotic pulse wave as a waveform having a very long cycle is considered. This type of technique can obtain the spectrum of the pulse wave in detail, but requires a huge amount of calculation, and thus is not suitable for applications in which the spectrum of the pulse wave generated every moment is quickly obtained. . 1 that constitutes a pulse wave
If the waveform parameters representing the characteristics of one wave can be continuously obtained, more information about the living body can be obtained, but there has been no device that meets such a demand.

[0004] The present invention has been made in view of the above circumstances, and can quickly analyze a pulse wave, and can analyze waveform parameters for each wave constituting a pulse wave. It is an object to provide a pulse wave analyzer.

[0005]

The invention according to claim 1 is
Waveform storage means for sequentially accumulating pulse wave time-series data in accordance with a predetermined writing speed; and dividing the time-series data into time-series data corresponding to one beat, and dividing the divided time-series data into the waveform storage means. from a reproducing means for reading, said comprising the analysis means on the basis of the time-series data corresponding to each beat and calculates the waveform parameters of the pulse wave corresponding to each beat are read by the reproducing means, the pulse wave for each beat Wave pattern
A pulse wave analyzer that outputs parameters
To the effect. The invention according to claim 2 is a pulse wave time-series data
Waveform storage means for sequentially accumulating data according to a predetermined writing speed
And the time series data corresponding to one beat each.
And each of the divided time-series data is recorded in the waveform
Reproducing means for reading from the storage means ;
Based on the time series data corresponding to each beat read out,
Analysis means for calculating pulse wave waveform parameters corresponding to the beat
And a waveform pattern corresponding to each of the beats in units of a predetermined number of beats.
Calculates and outputs the average value of the parameters
A pulse wave analyzer is provided. The invention according to claim 3 is
Time-series pulse wave data is sequentially written according to the specified writing speed.
Waveform storing means for accumulating, and storing the time-series data for one beat
It is divided into considerable time series data, and each divided time series
Reproducing means for reading data from said waveform storage means, before
Time series corresponding to each beat read by the recording / reproducing means
Pulse waveform parameters corresponding to each beat based on the data
Comprising an analyzing means for calculating, corresponding to each beat wave
Calculates and outputs the moving average value of shape parameters
The gist of the pulse wave analyzer is as follows. Invention according to claim 4
Is to write the time series data of pulse wave according to a predetermined writing speed.
A waveform storage means for sequentially accumulating the time-series data;
Divided into one beat corresponding time series data, for each beat,-out the specification
Time-series data that is higher than the
Multiple different readings, each proportional to the data length of the data
Output speed is selected in sequence, and the
And stores the time-series data corresponding to the beat in the waveform storage means.
And repeatedly read playback means from, depending on the playback unit
Based on the time series data corresponding to each beat read out
Fundamental and harmonic spectrum of pulse wave corresponding to each beat
Is an analysis means for calculating a specific pass frequency band.
Having time series data read by the reproducing means.
And a plurality of types of
For each of the read speeds
The time-series data read out is read by the band-pass filter.
Of the band-pass filter obtained when input to
From the output signal value, the fundamental wave and height of the pulse wave corresponding to the beat
By comprising analysis means for determining the spectrum of the harmonic
The feature of the invention is a pulse wave analyzer. According to claim 5
The invention relates to a period in which the time series data corresponding to one beat is read out.
A sine wave that outputs a sine wave signal with a frequency that is a fraction of an integer between
A generator based on the sine wave signal.
And detecting the phase of the spectrum.
The gist of the present invention is a pulse wave analysis device. Claim 6
The invention relates to the fundamental and harmonic spectra of a pulse wave.
6. The method according to claim 4, wherein the output is performed for each beat.
The pulse wave analyzer described above is a gist. Invention according to claim 7
Is the spectrum of the pulse wave corresponding to each of the above beats in units of a predetermined number of beats
Calculating and outputting the average value of the
The gist is the pulse wave analyzer according to item 4 or 5. Claim
The invention according to Item 8 is a waveform parameter corresponding to each beat.
Calculating and outputting a moving average value of
The gist is the pulse wave analyzer according to 4 or 5. Claim
9. The invention according to claim 9, wherein the analyzing means is a central part of a human artery system.
Aortic root in electrical circuit that mimics system from head to periphery
When an electric signal corresponding to the pressure wave of
From the time corresponding to the beat reproduced by the reproducing means
So that an output waveform corresponding to the series data is obtained.
That the value of each element in the circuit is calculated for each beat
The method according to any one of claims 1 to 3, wherein
The gist is a pulse wave analyzer. The invention according to claim 10 is
The electric circuit is caused by blood viscosity in the central part of the arterial system.
A first resistance corresponding to the vascular resistance,
Inductance corresponding to the inertia of blood, the central part of the artery
Capacitance corresponding to the viscoelasticity of blood vessels at the periphery, and the periphery
A second resistance corresponding to the vascular resistance at the part,
From the first resistance and inductance between the input terminals
Consisting of a series circuit and the capacitance and the second resistor
Four-element lumped constant in which parallel circuits are inserted in series in series
10. The pulse wave analysis according to claim 9, wherein the pulse wave analysis is a model.
The device is the gist.

[0006]

According to the above inventions, one wave constituting one pulse wave
Waveform parameters such as the spectrum for, or
The average value and moving average of each waveform parameter corresponding to each beat
Average value can be obtained.

[0007]

Embodiments of the present invention will be described below with reference to the drawings.

<First Embodiment> A. First Embodiment FIG. 1 is a block diagram showing a configuration of a pulse wave analyzer according to the present embodiment. As shown in FIG. 1, the pulse wave analyzer includes a pulse wave detector 1, an input unit 2, an output unit 3, a waveform extraction storage unit 4,
It comprises a frequency analysis unit 5 and a microcomputer 6 for controlling these. The pulse wave detector 1
It is constituted by a strain gauge or the like, and is pressed against a patient's radial artery or the like, detects the pressure and outputs it as a pulse wave signal (analog signal). The input unit 2 is a unit provided for inputting a command to the microcomputer 6, and is constituted by, for example, a keyboard.
The output unit 3 includes a printer, a display device, and the like. These devices record and display a spectrum of a pulse wave obtained from a patient under the control of the microcomputer 6. The waveform extraction storage unit 4 includes a microcomputer 6
, The pulse wave signal output from the pulse wave detection unit 1 is sequentially stored, and a switching point in the pulse wave signal, that is, from a pulse wave corresponding to a certain beat to a pulse wave corresponding to the next beat. 2 is extracted and stored, and the detailed configuration is shown in FIG. The frequency analysis unit 5 repeatedly reproduces the pulse wave signal stored in the waveform extraction storage unit 4 at a high speed for each beat, and obtains and outputs a spectrum constituting a pulse wave for each beat. The detailed configuration will be described. The spectrum of the pulse wave for each beat obtained by the frequency analysis unit 5 is captured by the microcomputer 6 and output from the output unit 3.

(1) Configuration of the Waveform Extraction Storage 4 Next, the configuration of the waveform extraction storage 4 will be described with reference to FIG.
In FIG. 2, reference numeral 101 denotes A / D (analog / digital).
The converter converts a pulse wave signal output by the pulse wave detector 1 into a digital signal in accordance with a sampling clock φ having a constant period, and outputs the digital signal. Reference numeral 102 denotes a low-pass filter which performs a process of removing components having a predetermined cut-off frequency or higher from a digital signal sequentially output from the A / D converter 101, and sequentially outputs the result as a waveform value W. Reference numeral 103 denotes a waveform memory constituted by a RAM, which sequentially stores the waveform values W supplied via the low-pass filter 102. 111 is a waveform value address counter, which is a waveform sampling instruction ST from the microcomputer 6.
While the ART is being output, the sampling clock φ
And outputs the count result as the waveform address ADR1 where the waveform value W is to be written. Reference numeral 112 denotes a selector, which alternately selects a waveform address ADR1 and a read address ADR4 output from the microcomputer 6 according to a select signal S1, and supplies the selected address to the address input terminal of the waveform memory 103 as a write address or a read address. Here, the waveform address ADR1 is monitored by the microcomputer 6.

Reference numeral 121 denotes a differentiating circuit which calculates and outputs the time derivative of the waveform value W sequentially output from the low-pass filter 102. Reference numeral 122 denotes a zero cross detection circuit, which outputs a zero cross detection pulse Z when the time derivative of the waveform value W becomes zero. More specifically, the zero cross detection circuit 1
Reference numeral 22 denotes a peak point P in the pulse wave waveform illustrated in FIG.
A circuit provided for detecting 1, P2,...
When a waveform value W corresponding to these peak points is input, a zero cross detection pulse Z is output. Reference numeral 123 denotes a peak address counter, which counts the zero-cross detection pulse Z while the microcomputer 6 outputs the waveform sampling instruction START, and outputs the count result as a peak address ADR2. Reference numeral 124 denotes a moving average calculation circuit that calculates the average value of the time differential values of a predetermined number of past waveform values W output from the differentiating circuit 121 up to the present time, and calculates the result as the pulse wave gradient up to the current time. Is output as inclination information SLP representing Reference numeral 125 denotes a peak information memory provided for storing peak information.
Here, the peak information is a general term for the information listed below, detected or created by the microcomputer 6, and written into the peak information memory 125 in accordance with the table format shown in FIG.

(Contents of peak information) Waveform value address ADR1: Write address ADR1 output from waveform address counter 111 when waveform value W output from low-pass filter 102 reaches a maximum value or a minimum value, that is, This is a write address in the waveform memory 103 of the waveform value W corresponding to the maximum value or the minimum value. Peak type B / T: information indicating whether the waveform value W written in the waveform value address ADR1 is a local maximum value T (Top) or a local minimum value B (Bottom). Waveform value W: a waveform value corresponding to the maximum value or the minimum value. Stroke STRK: A change in waveform value from the immediately preceding peak value to the peak value. Slope information SLP: Average value of time derivative of a predetermined number of waveform values in the past up to the peak value.

(2) Configuration of Frequency Analysis Unit 5 Next, a detailed configuration of the frequency analysis unit 5 will be described with reference to FIG. The frequency analysis unit 5 receives the waveform value WD of the pulse wave in beat units from the waveform memory 103 of the waveform extraction storage unit 4 via the microcomputer 6 and repeatedly reproduces the received waveform value WD at a high speed. The spectrum constituting the pulse wave is calculated by performing frequency analysis. The frequency analysis unit 5 firstly calculates the basic spectrum of the pulse wave,
Next, the second harmonic spectrum is calculated in a time-division manner for each spectrum constituting the pulse wave, such as.

When the microcomputer 6 outputs the first waveform value WD of the pulse wave for one beat to the frequency analysis unit 5, the microcomputer 6 outputs the synchronization signal SYNC and the waveform value W contained in the beat.
It outputs the number N of D and switches the select signal S2. While outputting the waveform value WD for one beat, the microcomputer 6 synchronizes with the delivery of each waveform value WD and changes the write address ADR5 that changes from 0 to N-1.
Are sequentially output.

The buffer memories 201 and 202 are memories provided for accumulating the waveform values WD output from the microcomputer 6 in this manner. The distributor 221 outputs the waveform value WD of the pulse wave supplied from the waveform extraction storage unit 4 via the microcomputer 6 to the buffer memory 201 or 202 to the one specified by the select signal S2. Further, the selector 222 selects the select signal S of the buffer memory 201 or 202.
2 selects the buffer memory designated by 2 and outputs the waveform value WH read from the buffer memory to the high-speed reproduction unit 230 described later. Selectors 211 and 212
Selects a write address ADR5 or a read address ADR6 (described later) generated by the high-speed playback unit 230 according to the select signal S2, and supplies the selected address to the buffer memories 201 and 202, respectively.

The distributor 221 and the selector 22 described above
2, 201 and 202 are controlled to be switched based on the select signal S2, so that the buffer memory 201
While data is being written to the
2, the data is read from the buffer memory 201 and supplied to the high-speed reproduction unit 230. While the data is being written to the buffer memory 202, the data is read from the buffer memory 201 and supplied to the high-speed reproduction unit 230.

The high-speed reproducing unit 230 is provided with a buffer memory 20
This is a means for reading waveform values corresponding to each beat from 1 and 202, and outputs the read address ADR6 while changing it in the range of 0 to N-1 (where N is the number of waveform values to be read). More specifically, the high-speed playback unit 23
0 indicates that the read address ADR6 is generated during the period when each waveform value WD corresponding to a certain beat is written to one buffer memory, and all the waveform values WD corresponding to the beat before that beat are stored in the other buffer memory. Is repeatedly read out a plurality of times. At that time, the total waveform value WD corresponding to one beat is
The generation of the read address ADR6 is controlled so that all data is always read within a certain period. The period during which all waveform values corresponding to one beat are read out is switched in accordance with the order of the spectrum to be detected. When detecting the fundamental wave spectrum, T is used. When the second harmonic spectrum is used, 2T is used. In the case of a wave spectrum, it is switched to 3T,. The high-speed reproducing unit 230 has a built-in interpolator, and interpolates the waveform value WH read from the buffer memory 201 or 202 to obtain a predetermined sampling frequency m /
It is output as a waveform value of T (m is a predetermined integer).

The band pass filter 250 is a band pass filter having a center frequency of a pass band of a predetermined value 1 / T. The sine wave generator 240 is a frequency-variable waveform generator, which corresponds to the order of the spectrum to be detected under the control of the microcomputer 6 and has a cycle of T, 2T, 3
The sine waves of T, 4T, 5T, and 6T are sequentially output. The spectrum detector 260 determines the amplitude H ₁ of each spectrum of the pulse wave based on the output signal level of the band-pass filter 250.
To H ₆ detects a bandpass phase theta ₁ through? ₆ of each spectrum based on the difference of the sine wave of phase the phase and sine wave generator 240 outputs the output signal of the filter 250 detects the.

C. Operation of Embodiment The operation of the embodiment will be described below.

(1) Waveform division First, when a frequency analysis start command is input from the input unit 1, the microcomputer 6 inputs a waveform sampling instruction S
TART is output, and the waveform address counter 111 and the peak address counter 123 in the waveform extraction storage unit 4 are output.
Is reset. As a result, the count of the sampling clock φ is started by the waveform address counter 111, and the count value is supplied to the waveform memory 103 as the waveform address ADR1. Then, the radial artery waveform detected by the pulse wave detector 200 is input to the A / D converter 101, sequentially converted into a digital signal according to the sampling clock φ, and sequentially output as a waveform value W via the low-pass filter 102. The waveform values W output in this manner are sequentially supplied to the waveform memory 103, and are written into a storage area specified by the waveform address ADR1 at that time. By the above operation, FIG.
A series of waveform values W corresponding to the radial artery waveform illustrated in FIG.

On the other hand, in parallel with the above operation, detection of peak information and writing to the peak information memory 125 are performed as described below. First, the low-pass filter 102
Is calculated by the differentiating circuit 121, and the time differential is calculated by the zero-cross detection circuit 12
2 and the moving average calculation circuit 124. The moving average calculation circuit calculates the average value (ie, moving average value) of a predetermined number of time differential values in the past each time the time differential value of the waveform value W is supplied in this way, and outputs the calculated result to the slope information S.
Output as LP. Here, when the waveform value W is rising or has finished rising and is in the maximum state, the inclination information S
A positive value is output as LP, and a negative value is output as the slope information SLP when it is in the minimum state during or after the descent.

For example, when the waveform value W corresponding to the local maximum point P1 shown in FIG. 4 is output from the low-pass filter 102, 0 is output from the differentiating circuit 121 as the time differential, and the zero-cross detecting circuit 122 detects the zero-cross. Pulse Z
Is output. As a result, the microcomputer 6 causes the waveform address ADR1, which is the count value of the waveform value address counter 111, the waveform value W, and the peak address A, which is the count value of the peak address counter, at that time.
DR2 (in this case, ADR2 = 0) and inclination information SL
P is taken in. Further, the count value ADR2 of the peak address counter 123 becomes 2 by outputting the zero cross detection signal Z.

The microcomputer 6 determines the peak type B / P based on the sign of the acquired inclination information SLP.
Create T. As in this case, the waveform value W of the maximum value P1
Is output at this point in time, the microcomputer 6 sets the value of the peak information B / T to correspond to the maximum value. Then, the microcomputer 6 writes the peak address ADR2 (ADR2 = 0 in this case) fetched from the peak address counter 123 as it is into the write address ADR3.
And the waveform value W, the waveform address ADR1 corresponding to the waveform value W, the peak type B / T, and the slope information SLP
As the first peak information, the peak information memory 125
Write to. In the case of the first writing of the peak information, the stroke information is not created or written because there is no immediately preceding peak information.

Thereafter, when the waveform value W corresponding to the minimum point P2 shown in FIG.
As described above, the zero cross detection pulse Z is output, and the write address ADR1, the waveform value W, and the peak address ADR2
(= 1), the inclination information SLP (<0) is taken in by the microcomputer 6. Then, similarly to the above, the microcomputer 6 determines the peak type B / T (in this case, bottom B) based on the inclination information SLP. The microcomputer 6 supplies an address smaller than the peak address ADR2 by one as the read address ADR3 to the peak information memory 125, and reads the first written waveform value W.
Then, the microcomputer 6 calculates a difference between the waveform value W fetched this time from the low-pass filter 102 and the first waveform value W read from the peak information memory 125 to obtain stroke information STRK. Then, the peak type B / T and the stroke information STRK obtained in this manner are used as other information ADR1, W, and SLP.
And the peak information memory 1 as the second peak information
The data is written to the storage area corresponding to the 25 peak addresses ADR3 = 1. Here, when the stroke information STRK is equal to or greater than a predetermined value, specifically, when the stroke information is a large stroke that can be regarded as corresponding to the rising portion of the pulse wave (for example, STRKM in FIG. 4), the microcomputer 6 further determines the stroke. (For example, in FIG. 4, the start point P of STRKM in FIG. 4)
6) is read from the peak information memory 125 and written into the built-in shift register. Thereafter, the peak points P3, P4,
.. Are detected, the same operation is performed.

(2) Delivering the Waveform In parallel with the above operation, the microcomputer 6 sequentially reads out the waveform values from the waveform memory 103 in the waveform extracting and storing unit 4 and delivers them to the frequency analyzing unit 5 as the waveform data WD. Hereinafter, this operation will be described with reference to FIGS.

As shown in FIG. 7, the select signal S1 is switched in synchronization with the clock φ, and the waveform memory 103 is switched between the write mode and the read mode in synchronization with this.

In FIG. 6, when the waveform value of the pulse wave Wn for one beat corresponding to a certain beat is input to the waveform memory 103, first, the first minimum value of the pulse wave corresponding to the beat is input. zero cross detection signal Z is generated at the time, its waveform address ADR1 = a ₀ is written to the peak information memory 125 (see FIG. 7). Then, the local maximum (address A
When 1) is input into the waveform extraction storage unit 4, the zero-cross detection signal Z is generated again (see FIG. 7), and the stroke between this maximum value and the immediately preceding minimum value (address A ₀ ) is set to a predetermined value. If the value is not less than the value, the minimum address A ₀ is written into a shift register (not shown) in the microcomputer 6. The waveform address written in this way is
After that, it is output from the shift register with a delay of two beats,
The waveform value WD for one beat to be delivered to the frequency analysis unit 5 is taken into the microcomputer 6 as a start address. That is, in FIG. 6, the pulse wave W corresponding to a certain beat
When the address Wn of the local maximum value of n is written to the shift register, 2
The start address (address of the first minimum value) of the pulse wave Wn-2 before the beat is output from the shift register and detected by the microcomputer 6.

At this point, the microcomputer 6 refers to the contents of the shift register and calculates the waveform address of the first minimum value of the pulse wave Wn-2 from the waveform address of the first minimum value of the next pulse wave Wn-1. , That is, the number N of waveform values included in the pulse wave Wn-1 for one beat is obtained, and is output to the frequency analysis unit 5 together with the synchronization signal SYNC. Further, the select signal S is synchronized with the synchronization signal SYNC.
2 is switched, and the internal connection state of the distributor 221, the selectors 211 and 212, and the selector 221 is, for example, as shown in FIG.
At the state indicated by the solid line.

Then, the microcomputer 6 sequentially increases the read address ADR4 from the first minimum waveform address of the pulse wave Wn-2 and supplies the read address ADR4 to the waveform memory 103 via the selector 112. Here, the read address ADR4 is changed at a higher speed (for example, twice as fast) as the write address ADR1. This is the pulse wave W
The maximum value of the pulse wave Wn + 1 of the pulse next to n is stored in the waveform extraction storage unit 4.
This is for reading out all the waveform values corresponding to the pulse wave Wn-2 before the pulse wave Wn-1 before being input to.
In parallel with the accumulation of the pulse wave Wn in the waveform memory 103 in this manner, the microcomputer 6 reads out the waveform value WD of the pulse wave Wn-2 two beats earlier from the waveform memory 103 and sends it to the frequency analysis unit 5. It is delivered and sequentially supplied to the buffer memory 201 via the distributor 221. In this manner, in synchronization with the sequential supply of the waveform value WD to the buffer memory 201, the write address ADR5 becomes 0 to N-1.
The write address ADR5 is sequentially supplied to the buffer memory 201 via the selector 211.
As a result, addresses 0 to N−1 of the buffer memory 201
Each waveform value WD corresponding to the pulse wave Wn-2 is accumulated in each of the storage areas.

(3) High-speed playback On the other hand, in parallel with the above operation, the read address ADR6 is output by the high-speed playback unit 230 and supplied to the buffer memory 202 via the selector 212. As a result, each waveform value W corresponding to the pulse wave Wn-3 one beat before the pulse wave Wn-2 is obtained.
D is read from the buffer memory 202 and the selector 2
The data is taken into the high-speed reproduction unit 230 via the control unit 22.

Here, the pulse wave W in the buffer memory 202
Each waveform value WD corresponding to n-3 is stored in the buffer memory 20.
Each waveform value corresponding to the pulse wave Wn-2 is repeatedly read out a plurality of times at a higher speed than that stored in the same. At this time, the waveform value WD corresponding to the pulse wave Wn-3 is set to the read address AD such that all of the waveform value WD is read within the fixed period T.
The rate of increase of R6 is controlled. That is, the high-speed playback unit 2
30 increases the read address ADR6 at a high speed when the number of waveform values WD to be read from the buffer memory 202 is a large value N1 as illustrated in FIG. 8, and conversely, decreases the read address ADR6 as illustrated in FIG. When the value is N2, the read address ADR6 is increased at a low speed, and the read address ADR6 becomes 0 to N1-1 within a certain period T.
Alternatively, the section from 0 to N2-1 is changed. Then, the waveform values WD sequentially read out in this way are subjected to an interpolation operation in the high-speed reproducing section 230, and supplied to the bandpass filter 250 as a waveform value WH of a constant sampling frequency m / T.

(4) Spectrum Detection The band-pass filter 250 selects and passes a signal having a frequency of 1 / T from the time-series data based on the waveform value WH, and supplies the selected signal to the spectrum detection unit 260. On the other hand, the sine wave generator 240 generates a sine wave having a period T as shown in FIG. Spectrum detector 260 detects over the output signal level of the bandpass filter 250 to several waves, and outputs the representative value as the amplitude H ₁ of the fundamental wave spectrum of pulse wave Wn-3, the band-pass filter 250 The phase difference between the phase of the output signal and the phase of the sine wave output from the sine wave generator 240 is detected over several waves, and the representative value is defined as the phase θ ₁ of the fundamental wave spectrum of the pulse wave Wn-3. Output. For each representative value, for example, a moving average value of the output signal level and the phase difference corresponding to each wave immediately before outputting the fundamental wave spectrum is calculated.

Next, the high-speed reproducing section 230 sets the increasing speed of the read address ADR6 so that all the waveform values of the pulse wave Wn-3 can be read within the predetermined period 2T, which is 1/2 of that in the case of detecting the fundamental wave spectrum. Then, the waveform value WH corresponding to the pulse wave Wn-3 is repeatedly read and supplied to the band-pass filter 250 (see FIG. 10). Then, the waveform value WH
A signal having a frequency of 1 / T, that is, a signal corresponding to the second harmonic of the pulse wave Wn-3 passes through the band-pass filter 250 and the spectrum detection unit 2
60. As a result, the spectrum detector 260
The amplitude H ₂ of the second harmonic spectrum of the pulse wave Wn-3
Is detected and output. On the other hand, the sine wave generator 240
Generates a sine wave having a period of 2T and supplies it to the spectrum detection unit 260 (see FIG. 10). As a result, the spectrum detector 260 outputs the phase θ ₂ of the fundamental wave spectrum of the pulse wave Wn-3.

Thereafter, the increasing speed of the read address ADR6 is 1/3, 1/4, 1 /
5, 1/6, and the period of the sine wave generated by the sine wave generator 240 is 3T,
4T, 5T, sequentially switched and 6T, the similar operation is performed, harmonic spectrum up to the third order 6 order amplitude H ₃ to H ₆ and the phase theta ₃ through? ₆ is the spectrum detection unit 260
Output from The pulse wave Wn− thus obtained
Each spectrum of 3 is taken into the microcomputer 6. Then, the pulse wave Wn is calculated by the microcomputer 6.
The frequency f = 1 / (Nτ) of the fundamental wave is calculated using the number N of the waveform values WD corresponding to −3 and the period τ of the clock φ, and is output from the output unit 3 together with the spectrum.

Thereafter, a pulse wave Wn which is one beat later than the pulse wave Wn
When +1 rises and the first maximum value is input into the waveform extraction storage unit 4, the microcomputer 6 generates a synchronization signal SYNC and outputs the number N of waveform values WD included in the pulse wave Wn-2. You. Further, the select signal S2 is inverted, and the internal connection state of the distributor 221, the selectors 211 and 212, and the selector 221 is changed to the state shown by the broken line in FIG. And the pulse wave Wn +
In parallel with the accumulation in the waveform memory 103, the microcomputer 6 calculates the waveform value W of the pulse wave Wn-1 two beats earlier by the microcomputer 6.
D is read from the waveform memory 103 and the frequency analysis unit 5
To the buffer memory 2 via the distributor 221.
02 sequentially. On the other hand, in parallel with this operation, the pulse wave Wn-
Each waveform value WD corresponding to 2 is read from the buffer memory 201, interpolated by the high-speed reproduction unit 230, and output as the waveform value WH. Then, the same processing as that of pulse wave Wn-3 is performed on waveform value WH corresponding to pulse wave Wn-2, and its spectrum is obtained.

Thereafter, the same processing as described above is performed for each of the sequentially arriving pulse waves, the spectrum of each pulse wave is continuously obtained, and output from the output unit 3 as waveform parameters corresponding to individual beats.

<Second Embodiment> In the first embodiment, the waveform data stored in the waveform memory 103 is reproduced in beat units, and the spectrum of the pulse wave is calculated in each beat unit. In contrast,
In this embodiment, each element of a four-element lumped parameter model in which the circulatory dynamics of a patient is modeled based on a pulse wave obtained from the patient using the method proposed by the present applicant in Japanese Patent Application No. 5-1431. Is obtained, and the result is used as a status display parameter. Here, the four-element lumped parameter model is based on the factors that determine the behavior of the circulatory system of the human body, the inertia of blood in the central part of the arterial system, the vascular resistance (viscous resistance) of blood viscosity in the central part, Focusing on four parameters of blood vessel compliance (viscoelasticity) and peripheral blood vessel resistance (viscous resistance), these are modeled as an electric circuit.

FIG. 11 is a circuit diagram of a four-element lumped parameter model. Hereinafter, the correspondence between each element constituting the four-element lumped parameter model and each of the above parameters will be described. Inductance L: inertia of blood in the central part of arterial system [dy
n · s ² / cm ⁵ ] Capacitance C: compliance (viscoelasticity) of blood vessels at the central part of the arterial system [cm ⁵ / dyn] Note that compliance is a quantity representing the softness of blood vessels, That is. Electric resistance R _c : vascular resistance due to blood viscosity at central part of arterial system [dyn · s / cm ⁵ ] Electric resistance R _p : vascular resistance due to blood viscosity at peripheral part of arterial system [dyn · s / cm ⁵ ] Current i, i _P , _ic flowing through each part in this electric circuit
Corresponds to the blood flow [cm ³ / s] flowing through each corresponding part. The input voltage e applied to this electric circuit corresponds to the pressure [dyn / cm ² ] at the aortic root. The terminal voltage v _P of the capacitance C is the pressure [dyn / cm
² ].

In this embodiment, the microcomputer 6 stores the waveform data corresponding to each beat in the buffer memory 20.
While sequentially writing data into the buffer memory 1 and 202, the waveform data corresponding to one beat is read out from the buffer memory where the writing has not been performed via the selector 222. Then, the operation of the 4-element lumped parameter model when an electric signal corresponding to the pressure wave at the aortic root is given is simulated, and an output waveform corresponding to the waveform data read from the buffer memory 201 or 202 is obtained. Thus, the value of each element of the four-element lumped parameter model is calculated, and the calculation result is output as a waveform parameter. The value of each element of the four-element lumped model can be obtained by trial and error by repeating the operation simulation while changing the value of each element, but is described in Japanese Patent Application No. 5-1431. May be used.

<Modifications> (1) In the above-described first embodiment, the frequency analysis of the pulse wave was executed by hardware, but the present invention is not limited to this, and the frequency is analyzed by software executed by the microcomputer 6. An analysis may be performed. Various methods such as DFT (Discrete Fourier Transform) and FFT (Fast Fourier Transform) can be applied to the frequency analysis. (2) In each of the above embodiments, the waveform parameters corresponding to each beat are output in real time at the time when each is obtained, but the method of outputting the waveform parameters is not limited to this. For example, the microcomputer 6 may calculate and output the average value of the waveform parameters for a predetermined number of beats. Alternatively, the microcomputer 6 may calculate the average value of the waveform parameters for a predetermined number of beats in the past, that is, the moving average value of the waveform parameters, and output the moving average value to the real time. (3) Although the apparatus for analyzing the radial artery is described in each of the above embodiments, the analysis target of the present invention is not limited to the radial artery, and other types of pulse waves such as a fingertip pulse wave may be used. May be applied to the analysis. (4) As the waveform parameter of the pulse wave, various ones can be considered in addition to those described in the above embodiments. When the pulse wave analyzer according to the present invention is used for diagnosis, Deformation is performed to find the optimal waveform parameter for diagnosis. For example, the present applicant has filed Japanese Patent Application No. 5-1975.
No. 69 proposes a device for obtaining a stress level of a patient based on the amplitude value and phase of a peak point appearing in a pulse wave. With the device according to the above embodiment, information on the peak point may be obtained from the pulse wave corresponding to each beat, and the obtained information may be used for evaluating the stress level.

[0040]

As described above, according to the present invention, a pulse wave can be quickly analyzed, and one wave constituting the pulse wave can be analyzed.
Waveform parameters of a wave can be continuously analyzed, and an effect that much information about a living body can be obtained is obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a pulse wave analyzer according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of a waveform extraction storage unit 4 in the embodiment.

FIG. 3 is a block diagram showing a configuration of a frequency analysis unit 5 in the embodiment.

FIG. 4 is a diagram illustrating a pulse wave written in a waveform memory 103 according to the embodiment.

FIG. 5 is a diagram showing storage contents of a peak information memory 125 in the embodiment.

FIG. 6 is a diagram illustrating the timing of delivering a waveform from the waveform extraction storage unit 4 to the frequency analysis unit 5 in the embodiment.

FIG. 7 is a time chart showing an operation in a waveform extraction storage unit 4 in the embodiment.

FIG. 8 is a diagram illustrating the operation of the high-speed playback unit 230 in the embodiment.

FIG. 9 is a diagram for explaining the operation of the high-speed playback unit 230 in the embodiment.

FIG. 10 is a diagram illustrating the operation of a high-speed reproducing unit 230 and a sine wave generator 240 in the embodiment.

FIG. 11 illustrates a fourth embodiment of the present invention.
It is a circuit diagram showing an element lumped parameter model.

[Explanation of symbols]

4... Waveform extraction storage unit, 6... Frequency analysis unit, 5... Microcomputer.

Claims (10)

(57) [Claims] 1. A waveform storage means for sequentially accumulating time-series data of a pulse wave according to a predetermined writing speed; and dividing the time-series data into time-series data each corresponding to one beat; the comprising a reproducing means for reading from said waveform storage means and analyzing means for calculating waveform parameters of the pulse wave corresponding to each beat based on the time-series data corresponding to each beat are read out by the reproducing means, each beat Pulse wave waveform for each
A pulse wave analyzing device for outputting a parameter . 2. The method according to claim 1, wherein the time series data of the pulse wave is written at a predetermined writing speed.
Waveform storage means for sequentially accumulating according to degree, and dividing the time-series data into time-series data each corresponding to one beat
The divided time-series data is stored in the waveform storage means.
Playback means for reading from the memory and time series corresponding to each beat read by the playback means
Pulse waveform waveform parameters corresponding to each beat based on the column data
Analysis means for calculating the data, and the
Calculates and outputs the average value of the waveform parameters corresponding to each beat
A pulse wave analyzer characterized by applying force. 3. A method for writing pulse-wave time-series data at a predetermined writing speed.
Waveform storage means for sequentially accumulating according to degree, and dividing the time-series data into time-series data each corresponding to one beat
The divided time-series data is stored in the waveform storage means.
Playback means for reading from the memory and time series corresponding to each beat read by the playback means
Pulse waveform waveform parameters corresponding to each beat based on the column data
Analysis means for calculating the data ,
Calculates and outputs the moving average of waveform parameters.
Pulse wave analyzer to be used as a feature. 4. A method for writing time-series data of pulse waves at a predetermined writing speed.
Waveform storage means for sequentially accumulating according to degree, and dividing the time-series data into time-series data each corresponding to one beat
In each beat, the writing speed is higher than the writing speed and
It is proportional to the data length of the time-series data corresponding to the beat .
Select several different reading speeds sequentially and select
Depending on the selected reading speed, the time-series data corresponding to the beat
Reproduction means for repeatedly reading data from the waveform storage means
And a time series corresponding to each beat read by the reproducing means.
Based on the pulse data, the fundamental wave of the pulse wave corresponding to each beat and
Analysis means for calculating the spectrum of harmonics.
Having a pass frequency band of
Band-pass filter to which the input time series data is input
For each of the plurality of read speeds.
The time-series data read at the read speed is
The corresponding bar obtained when input to the bandpass filter
From the output signal value of the
Analytical means for determining the fundamental and harmonic spectra of waves
Pulse wave analyzing apparatus characterized by comprising and. 5. The time series data corresponding to one beat is read out.
Output a sine wave signal with a frequency that is a fraction of the integer period
A sine wave generator, wherein the analyzing means includes a sine wave signal.
Detecting the phase of the spectrum based on
The pulse wave analyzer according to claim 4, wherein 6. A spectrum of a fundamental wave and a harmonic wave of a pulse wave.
Is output for each beat.
3. The pulse wave analyzer according to 1. 7. A pulse wave corresponding to each of said beats in units of a predetermined number of beats.
Calculates and outputs the average of the spectrum
The pulse wave analyzer according to claim 4 or 5, wherein 8. The transfer of a waveform parameter corresponding to each beat.
5. The method according to claim 4, wherein a moving average value is calculated and output.
Or the pulse wave analyzer according to 5. 9. The apparatus according to claim 1, wherein said analyzing means is a central part of a human arterial system.
Electrical circuit that mimics the system from the aorta to the periphery
When an electric signal corresponding to the pressure wave is given,
Time series corresponding to the beat reproduced by the reproducing means.
The electric circuit is controlled so that an output waveform corresponding to the column data is obtained.
The value of each element of the road is calculated for each beat.
The pulse according to any one of claims 1 to 3, wherein
Wave analyzer. 10. The system of claim 10 , wherein said electrical circuit is
A first resistance corresponding to vascular resistance due to blood viscosity of the
Inductin for blood inertia in central arterial system
Capacitance corresponding to the viscoelasticity of blood vessels at the central part of the artery
Volume and a second resistance corresponding to the vascular resistance at the periphery.
A first resistor between the pair of input terminals;
A series circuit consisting of an inductance and the capacitance and
A parallel circuit consisting of a second resistor is inserted in series
A four-element lumped parameter model
Item 10. A pulse wave analyzer according to Item 9.