JPH03272732A - Electronic hemadynamometer - Google Patents

Abstract

PURPOSE:To accurately make a judgement on blood pressure only based on the Korotkoff sound signals by providing a setting means for setting the base noise level of detected signals by a detecting means based on the respective values of detected peak and bottom signals. CONSTITUTION:In the detection of base noise, the time of 2 seconds after measurement has been started, is equally divided by 16 so that each window is set up to 125mS, the noise level in each window is then measured, both of the peak and bottom values in each window are detected so as to be stored in a peak register KP and a bottom register KB, respectively, and furthermore, each difference is determined so that it is made to be the noise level in each window. And, an area counter AC counts 0 through 15 with +1 added to each window, and control is performed in such a way that KP and KB of each window of AC is detected, a greater one out of KP and KB is picked up so as to be stored in the peak memory P of the window, and a smaller one is concurrently picked up so as to be stored in the bottom memory B of the window, and the minimum one out of 15 differences is picked up so that it is made to be base noise BN.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an electronic blood pressure monitor, and more particularly to an electronic blood pressure monitor that determines blood pressure using Korotkoff sounds.

[Prior Art] Conventionally, in this type of electronic blood pressure monitor, a Korotkoff sound signal is gated using a pulse wave gate signal obtained by extracting a pulsating component of cuff pressure to prevent misjudgment due to noise or artifacts.

[Problems to be Solved by the Invention] However, for example, when the rate of decrease in cuff pressure is fast, a sufficient pulsation component cannot be extracted from the cuff pressure. Therefore, a pulse wave gate signal could not be formed, and the important Korotkoff sound signal could not be detected, often leading to incorrect measurements.

The present invention eliminates the above-mentioned drawbacks of the prior art, and its purpose is to provide an electronic blood pressure monitor that is capable of uniquely and accurately determining blood pressure based solely on Korotkoff sound signals.

[Means and effects for solving the problems] In order to solve the above problems, the electronic blood pressure monitor of the present invention determines blood pressure using Korotkoff sounds. window setting means for setting a window for a predetermined time after the start; peak-bottom detection means for detecting each peak and bottom of the detection signal by the detection means within the window; The outline of the present invention is to include a setting means for setting the base noise level 2 of the detection signal by the detection means based on the detection means.

As a result, true Korotkoff sound signals and noise are clearly distinguished, and a unique and accurate blood pressure determination is performed based only on the Korotkoff sound signals.

Preferably, one aspect of the window setting means is to set two or more n windows during a predetermined time after the start of measurement.

Preferably, one aspect of the peak-bottom detection means is to detect the peak and bottom of each of the n windows.

Preferably, the setting means calculates a difference between a peak value and a bottom value for each of the n sets of peaks and bottoms detected, and sets the one with the smallest difference as the base noise level. do.

Preferably, the setting means calculates a difference between a peak value and a bottom value for each of the n sets of detected peaks and bottoms, and sets the average value of these differences as the base noise level. do.

Preferably, the setting means calculates the difference between the peak value and the bottom value for each of the detected n sets of peaks and bottoms, and selects m (
One aspect of this is to set the average value up to (man) as the base noise level.

[Description of Embodiments] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram of an electronic blood pressure monitor according to an embodiment, and FIG. 2 is an external view of the electronic blood pressure monitor according to an embodiment. In the figure, this electronic blood pressure monitor includes a main unit 10 for determining blood pressure, a pressurizing unit 40 for pressurizing/depressurizing the cuff, a sound detecting unit 50 for detecting Korotkoff sounds (K sounds), and an external charging unit. It consists of a container 60.

In the main body 10, 11 is a rechargeable battery, and as shown in FIG. 2, four 1.2V rechargeable batteries are connected in series.
．． Supply 8V. A mounting section 11a for the rechargeable battery 11 is provided on the main body 10 side, and after mounting the rechargeable battery 11, the lid 1 is removed.
1b is covered. Returning to FIG. 1, the rechargeable battery 11 can be charged by an external charger 60. Reference numeral 12 denotes a power control, which controls supply/cutoff of 4.8V to the main body 10 in accordance with the operation of the power switch 13. 14 is a basic oscillator (Xi), which generates a basic clock signal.

Reference numeral 15 denotes a mode switch, which is operated to select a usage mode of the electronic blood pressure monitor. For example, mode (A) is a mode for measuring a normal person. Mode (B) is a mode in which the device does not automatically measure blood pressure, but functions as a mercury column in conventional auscultation by doctors, and only displays cuff pressure that changes sequentially.

Mode (C) is a mode for measuring a person with an auscultation gap, and the timing for determining the point at which Korotkoff sounds disappear is extended. 31 is a start switch, and the user can also manually instruct the start of measurement.

A filter amplifier 16 filters and amplifies the Korotov sound (K sound) signal detected by the microphone 51. A reference power supply section 17 supplies a controlled and stabilized reference power supply, reference voltage, etc. to the amplifier circuit and A/D conversion section that require a stabilized power supply. 18 is a pressure detection section that detects cuff pressure. An amplifier 19 amplifies the detected cuff pressure signal. 26 is an A/D converter that samples the amplified cuff pressure signal and converts it into a digital signal. 25 is an external memory, which temporarily stores various data necessary for blood pressure measurement. 2
8 is the exhaust valve of the normal oven, and the drive unit 27
Closed by the drive signal.

29 is a display (LCD), as shown in FIG.
Sequential display of cuff pressure during measurement, determined systolic blood pressure SY
Area 29a that displays S and diastolic blood pressure DIA, the priority of the measuring means that gave these measurement results, and further displays the insufficient pressurization state, pulse rate, etc.
An area that displays a graphical display of the pulse wave signal amplitude during measurement, a microcomputer display of the cuff decompression speed range during measurement, the selected use mode of the blood pressure monitor, and a rapid evacuation mark indicating that rapid evacuation is in progress. 29b is provided. Returning to FIG. 1, 30 is a buzzer that notifies the end of the measurement.

A microprocessor section (LSIC) 20 performs main control and processing of the main body section 10 . Microprocessor section 2
0, 21 is an A/D conversion section which samples the amplified sound signal and converts it into a digital signal. 2
2 is a control unit, which has a built-in CPU unit, which inputs sampled cuff pressure signals and sound signals, exchanges processing data with an external memory 25, and controls the display via the display drive unit 24. Controls the display of the LCD 29, etc.

23 is a Korotkoff sound/pulse wave recognition unit which, together with the control unit 22, executes the processing programs shown in FIGS. 3 to 10, which will be described later.

In the pressurizing part 40, 41 is a cuff that is wrapped around a person's upper arm;
is a constant speed (low speed) exhaust valve, 43 is a manual exhaust valve, 4
4 is a rubber bulb, and 70 is a tube forming an air flow path. As shown in FIG. 2, the manual exhaust valve 43 includes a knob 43a for adjusting a low exhaust speed and a full-open switch 43b for manually performing rapid exhaust.

FIG. 3 is a flowchart of main processing including oscillometric measurement in the embodiment. In step S301, "pressurization and detection" is performed.

This pressure detection is performed, for example, by the method shown in FIG. 11(A). The horizontal axis of the figure is time t, and the vertical axis is cuff pressure P. The constant speed exhaust valve 42 of this embodiment is always open, and since the exhaust valve 28 is a normal oven, the rubber bulb 44
When pressurization is performed, the cuff pressure detection signal PS is approximately zero at first, but as shown in the figure, the cuff pressure detection signal PS increases by pulsating up and down as shown in the figure in response to the pump operation by the rubber bulb 44, and each bottom detection signal BM increases accordingly. The value of (i) also increases for a while. Therefore, the control unit 22 controls the content of each BM(f) to be continuously n (n
= 3 or 4) times, it is determined that "pressure has been detected".

FIG. 11(B) shows the case where the arm wrapped around the cuff 41 is simply moved without performing a pump operation using the rubber ball 44. In this case, since each bottom detection signal BM(i) cannot rise continuously n times, pressurization is not detected. The above details will be described later according to the flowchart shown in FIG.

In this embodiment, the contents of the signal BM (i) are compared, but the comparison is not limited thereto. For example, peak detection signal PM (0)
, PM (1), . . . may be obtained and compared.

Further, as shown in FIG. 11(C), when the pressure route is well sealed, the inflection point IPO-IP6 can be detected. In this case, the inflection point IPOIP2. IP4. ...or inflection point IPI, IF3°IF5. ..., etc. may be compared.

The point of inflection can be detected by detecting the gradient of increase in cuff pressure per unit time and finding the point at which the gradient changes.

Returning to FIG. 3, in step 5302, the exhaust valve 28 is closed. In step 5303, the system waits for "pressurization to stop". That is, the operator operates the pump until the desired cuff pressure is reached, and then stops applying pressure. Then constant speed exhaust valve 4
2, the cuff pressure decreases at the current adjustment rate (e.g. 2-4 mmHg/S). For example, as shown in FIG. 12, the control unit 22 controls the cuff pressure signal PS to be 0.5.
By detecting a constant rate of decrease over a period of seconds,
It is determined that the pressure has stopped.

In step 5304, measurement is started.

With this start of measurement, in addition to the oscillometric measurements after step 5305, "sound measurement with gate" and "sound measurement without gate", which will be described later, are performed in parallel. In step 5305, initial settings for oscillometric measurement are performed. That is, a flag UF holds that the cuff pressure signal PS is rising, and a flag U holds that the cuff pressure signal PS is falling after rising.
DF, the bottom counter BC1 that counts the number of bottom occurrences of the pulse wave signal component of the cuff pressure, and the gate signal KG for sound detection are all reset.

In step 5306, "pulse wave detection" is performed.

This pulse wave detection is performed, for example, by the method shown in FIG. That is, when the point where the cuff pressure signal PS first rises is detected after the start of measurement, this is the starting point of the first pulse wave signal component, and the bottom signal BM(0) at that point is stored. At the same time, open the sound gate KG for Korotkoff sound detection. Next, the top (peak) of the cuff pressure signal PS is detected, the top signal TM (0) at that time is stored, and the bottom counter BC is incremented by 1. Next, a point where the cuff pressure signal PS becomes equal to the bottom signal BM (0) is detected and the sound gate KG is closed. The above is one cycle of pulse wave signal component detection, and the same pulse wave detection process is repeated thereafter. The above details will be described later according to the flowchart shown in FIG.

Returning to FIG. 3, in step 5307, the bottom counter B
It is determined whether C is 3 or more, and if it is not 3 or more, it is determined in step 5309 whether the signal *END is true. *EN
D is a signal that becomes true when both "sound measurement with gate" and "sound measurement without gate" are in the completed state. At this point, signal *END is not true, so step 5
At 310, it is checked whether PS<20mmHg.

Here, 20 mmHg is the lowest cuff pressure at which blood pressure measurement can be performed. At this point, since PS<20 mmHg is not established, the process returns to step 5306.

In this way, multiple pulse wave detections are performed, and eventually step 53
If BC≧3 is satisfied in step 07, the process proceeds to step 5308, where BM (C-3)<BM (C-2)<BM (C-
1) Determine whether <BM (C). Here, (C) is the address of the external memory 25 pointed to by the bottom counter BC. That is, in this process, the bottom signal BM (C-3) from the previous three bottom signals to the current bottom signal BM (
C) is checked, and if these increase continuously, it means that the pressurized state by the rubber ball 44 has been detected again during the measurement, and the control returns to step 5303. Further, if it is determined in step 5308 that there is no continuous rise, the process advances to step 5309. Thus, the signal *END is true in step 5309, or PS<20 in step 5310.
Once mmHg is determined, the process advances to step 5311 and "oscillometric analysis" is executed.

This oscillometric analysis is performed, for example, by the method shown in FIG. 13(A). In the figure, when the cuff pressure PS starts to decrease, the pulse wave signal PA appears before the systolic blood pressure SYS, and the pulse wave signal PA reaches the maximum value PAmax (average blood pressure) at the position of ■.
, and then decreases for a while toward the diastolic blood pressure DIA. The control unit 22 examines the pulse wave signal PA stored in the external memory 25 in advance, first detects the maximum value PAmax, and determines whether the amplitude of the pulse wave signal PA has exceeded the threshold TH1=0.5XPAmax before this. The cuff pressure PS at time ■ is determined as the systolic blood pressure SYS, and the maximum value P
After Amax, the moment immediately before the amplitude of the pulse wave signal PA falls below the threshold TH2=0.8XPAmax■
The cuff pressure PS is determined as the diastolic blood pressure DIA.

Note that the method for determining SYS and DIA in this embodiment has some characteristics, the details of which will be described later with reference to the flowchart of FIG.

Returning to FIG. 3, the buzzer is sounded in step 5312, and the exhaust valve 28 is opened in step 5313, thereby completing the measurement.

In the above, a priority is assigned to the display of the results of each measurement method. For example, the priority is lower in the order of "sound measurement with gate,""r oscillometric measurement," and "sound measurement without gate." ing. Therefore, normally the sound measurement results with gates are displayed preferentially. However, by pressing the mode switch 15, it is possible to preferentially display the results of oscillometric measurements or sound measurements without gates.

Alternatively, after the measurement is completed, it is also possible to simply change the display to arbitrarily select these measurement results.

Note that in this embodiment, the rate of decrease in cuff pressure can be adjusted, and when the rate of decrease in cuff pressure is fast, it becomes difficult to extract the pulse wave signal PA. It may come. However, even in such a case, according to this embodiment, sound measurement is performed without a gate, so accurate measurement results can be obtained.

FIG. 4 is a flowchart showing details of pressurization detection in the embodiment. UF in step S401.

Clear UDF and BC, and further clear register PR holding the sampled cuff pressure signal PS. In step 5402, the cuff pressure signal PS is sampled.

This sampling may be performed at a period of 32 mS, for example. In step 5403, (PR+Δp)-ps is calculated. Here, ΔP is a small threshold value. (PR+ΔP
)-PS≧0, the process advances to step 5406 and (PR
-ΔP)-PS is calculated. (PR-△p)-ps≦O
When , ps is within the range of ΔP compared to PR (initially O), so nothing is done and the process returns to step 5402.

Also, in the calculation at step 5403, (PR+Δp)−ps
When <o, it is recognized that the cuff pressure PS has increased by more than the threshold value ΔP, so the process advances to step 5404 and UP is checked. Until now, the cuff pressure has continued to decline at a constant rate, so at this point UP=O. Control is step 5
Proceed to 41O, set UF, and reset UDF. That is, the cuff pressure starts to rise. Step 5411
Then, the contents of PR are stored in the bottom memory BM (C) of the external memory 25. Here, (C) is the address pointed to by the bottom counter BC.

In step 5412, it is determined whether BC=3. Initially, BC=O, so the process advances to step 5405 and updates the content of PS which is the content of PR.

Thus, while the PS continues to rise, step 54
Loop 02, 5403, 5404, 5405.

If (PR-ΔP)ps>o in the calculation in step 5406, a decrease of more than the threshold value ΔP is recognized, so the process proceeds to step 5407 and examines UF. Since UF=1 here, the process advances to step 5408, where OF is reset and UDF is set. This is the start of descent after the cuff pressure rises.

In step 5409, the bottom counter BC is incremented by 1.

In this way, while detecting the pulsating rise due to the pressurizing operation by the rubber ball 44, in step 5412 BC=
3, the process advances to step 5413 and BM (0
)<BM (1)<BM(2)<BM (3). If YES, it means "pressure detection", so the process exits.

If the answer is No, BM (0) 4- is sent in step 5414.
A downshift of BM (1) - BM (2) - BM (3) is performed, and in step 5415, the bottom counter BC is incremented by 1. Thereby, from now on, it is determined whether or not pressurization is detected every time the bottom BM is detected.

FIG. 5 is a flowchart of pulse wave detection according to the embodiment. This process is similar to the pressurization detection process shown in FIG. 4, except for a part. Initialization is performed in step 5305 of FIG.

In step 5501, the cuff pressure signal PS is sampled, and in step 5502, tJDF is checked. At first U.D.
Since F=O, the process advances to step 5505 and (PR+
Δp)-ps is calculated. When (PR+ΔP)-PSkO, (PR-Δp)-ps is calculated in step 5510. When (PR-△P)-PS≦O, the sampled cuff pressure PS is within the range of ±ΔP compared to the contents of the register PR (here, the previous value is held), so nothing can be done. Otherwise, the process returns to step S501.

When (PR-Δp)-ps>o in the calculation at step 5510, it is recognized that the cuff pressure PS has decreased beyond the threshold value ΔP. Control continues to step 5511 and examines the UF. Since the cuff pressure is originally decreasing at the time of starting measurement, LJF=O. Control continues to step 5515;
Simply update the contents of register PR with the contents of PS.

When (PR+Δp)−ps<o in the calculation in step 5505, an increase in the cuff pressure PS exceeding the threshold value ΔP is recognized. Control continues to step 5506 and examines the UF. U
If P=O, it means that a rise in cuff pressure has been detected from a fall, so proceed to step 5507, set the UF, and set the UD.
Reset F. In step 8508, the contents of PR are stored in the bottom memory BM (C). step 550
At step 9, the gate signal KG for sound detection is opened. Thereafter, steps 5505 and 50 are performed while the cuff pressure PS continues to rise.
Loop 6,515 routes.

When (PR-ΔP)ps>o in the calculation at step 5510, a decrease in the cuff pressure exceeding the threshold value ΔP is recognized. Control continues to step 5511 and examines the UF. This time U
Since F=1, the process advances to step 55L2, where UF is reset and tJDF is set. After the cuff pressure has increased, it begins to decrease. In step 5513, the top memory TM
The contents of the PR are stored in (C). In step 5514, the bottom counter BC is incremented by 1.

When UDF=1, that is, the start of a decrease after the increase in cuff pressure is detected, the determination in step 5502 becomes YES, and the control proceeds to step 5503, where it is determined whether PS<BM (C-1). When ps<BM(C-1), the process proceeds to step 5504, and the gate signal KG for K sound detection is closed.

In this way, the external memory 25 stores data BM (C) and TM (C) as many as necessary for later oscillometric analysis, and at the same time, a gate signal KG for sound detection is generated in real time.

FIG. 6 is a flowchart of the oscillometric analysis of the embodiment. In step 5601, the pulse wave signal PA is obtained from the data BM (] and TM (C). In FIG. 12, the cuff decompression rate at that point is -K m m
Assuming that Hg/s, for example, the amplitude of the pulse wave signal PA3 can be found as (TM(3)-BM(3)+Δt).

Here, Δt is the time from BM (C) to TM (C).

Similarly, the amplitude PAn of other pulse wave signals is determined. In step 5602, PAmax is detected. 1st
In FIG. 3(A), PAmax is the pulse wave signal that has the largest amplitude PA. PAmax can be easily detected by checking the amplitude PA of the pulse wave signal starting from the lowest address in the external memory 25. In step 5603, the systolic blood pressure SYS is determined. There are two methods for determining the systolic blood pressure SYS in this embodiment.

FIG. 13(B) is a diagram illustrating the first method for determining the systolic blood pressure SYS. This first method is a method in which the amplitude PA of the pulse wave signal in the external memory 25 is read out from the address with the lowest address. Assuming that the threshold value TH1=0.5XPAmax, first the pulse wave signal (■) exceeding the threshold value THI is detected. Then, the amplitude PA3 of the next pulse wave signal ■ is the threshold value (PA2-C
4), the cuff pressure P corresponding to the pulse wave signal ■
S is determined to be the systolic blood pressure SYS.

Here, 04 is a fourth predetermined value, which is statistically determined in advance based on the variation in the amplitude PA of the actual pulse wave signal. In addition, when the amplitude PA3 of the pulse wave signal ■ does not exceed (PA2-C4), it is confirmed that the amplitude PA4 of the next pulse wave signal ■ exceeds the threshold THI, and then the pulse wave signal ■ and the pulse wave Signal ■
If the difference is greater than or equal to the fifth predetermined value 05, the cuff pressure PS corresponding to the pulse wave signal ■ is determined to be the systolic blood pressure SYS. Similarly, the fifth predetermined value C5 is also statistically determined in advance based on the variation in the amplitude of the actual pulse wave signal PA.

That is, the amplitude PA3 of the pulse wave signal ■ is (PA2-C4)
without exceeding the threshold value TH, the amplitude PA4 of the next pulse wave signal ■
I, and when the difference in amplitude between the pulse wave signal ■ and the pulse wave signal ■ is greater than or equal to the fifth predetermined value 05, the pulse wave signal ■
is determined to be noise. On the other hand, when the amplitude PA3 of the pulse wave signal ■ does not exceed (PA2-C4) and the amplitude PA4 of the next pulse wave signal ■ does not exceed the threshold TH1, or when the amplitude of the pulse wave signal ■ and the pulse wave signal ■ When the difference is less than the fifth predetermined value 05, the pulse wave signal ■ is determined to be noise.

In this way, even if there are variations in the amplitude PA of the pulse wave signal, which often occurs in practice, the systolic blood pressure SY can be reliably adjusted.
S can be determined.

FIG. 13(C) is a diagram illustrating a second method for determining the systolic blood pressure SYS. This second method is a method in which the amplitude PA of the pulse wave signal in the external memory 25 is read out from the position of PAmax toward smaller addresses. Similarly, the threshold T
If H1=0.5XPAmax, first the threshold T
Detect pulse wave signal ■ below HI. Then, when the amplitude PA2 of the next pulse wave signal ■ does not exceed the threshold value (PAS 106), it is determined that the systolic blood pressure SYS is based on the cuff pressure PS corresponding to the pulse wave signal ■. Here, C6 is the sixth predetermined value, which is statistically determined in advance based on the variation in the amplitude PA of the actual pulse wave signal. Furthermore, when the amplitude PA2 of the pulse wave signal ■ exceeds the threshold value (PA3+C6), the amplitude P of the pulse wave signal ■
When AI is below THI, pulse wave signal ■ and pulse wave signal ■
If the difference is greater than or equal to the seventh predetermined value C7, the cuff pressure PS corresponding to the pulse wave signal ■ is determined to be the systolic blood pressure SYS. Similarly, the seventh predetermined value C7 is also statistically determined in advance based on the variation in the amplitude PA of the actual pulse wave signal.

That is, the amplitude PA2 of the pulse wave signal ■ is equal to the threshold (PA3+C
6), the amplitude PAL of the pulse wave signal ■ is below THI, and the difference in amplitude between the pulse wave signal ■ and the pulse wave signal ■ is
is greater than the predetermined value 07, the pulse wave signal ■ is determined to be noise.

On the other hand, the amplitude PA2 of the pulse wave signal ■ is the threshold (PA30C6)
exceeds THI, or when the difference in amplitude between the pulse wave signal ■ and the pulse wave signal ■ exceeds the seventh predetermined value 0.
When it is less than 7, the pulse wave signal ■ is determined to be noise.

In this way, even if there are variations in the amplitude PA of the pulse wave signal, it is possible to accurately determine the systolic blood pressure SYS, and since it is only necessary to trace back the external memory 25 from the position where PAmax is detected, it is easy to control the successive output of pulse wave signal data. .

Returning to FIG. 6, in step 5604, the diastolic blood pressure DIA is determined.

FIG. 13(D) is a diagram illustrating a method for determining the diastolic blood pressure DIA in the example. In this method, the amplitude PA of the pulse wave signal in the external memory 25 is read backward from the position of PAmax.

Assuming that the threshold value TH2=0.8xPAmax, first the pulse wave signal (■) below the threshold value TH2 is detected. Then, when the amplitude of the next pulse wave signal (■) does not exceed the threshold (PAB1C8), it is determined that the diastolic blood pressure DIA is based on the cuff pressure PS corresponding to the pulse wave signal (■). Here, 08 is the eighth predetermined value, which is statistically determined in advance based on the variation in the amplitude PA of the actual pulse wave signal. Furthermore, when the amplitude PA7 of the pulse wave signal ■ exceeds the threshold (PA6+C8), the pulse wave signal ■ becomes TH2.
In the following cases, the difference between the pulse wave signal ■ and the pulse wave signal ■ is calculated, and when the difference is greater than or equal to the ninth predetermined value 09, the cuff pressure PS corresponding to the pulse wave signal ■ is determined to be the lowest. Blood pressure is determined to be DIA. Similarly, the ninth predetermined value C9 is also statistically determined in advance based on the variation in the amplitude of the actual pulse wave signal PA.

That is, the amplitude PA7 of the pulse wave signal ■ is equal to the threshold value (PA6+C
8), and if the pulse wave signal ■ is TH2 or more, or if the difference between the pulse wave signal ■ and the pulse wave signal ■ is less than the ninth predetermined value 09, the pulse wave signal ■ is determined to be noise. on the other hand,
The amplitude PA7 of the pulse wave signal ■ exceeds the threshold (PA6+C8), the pulse wave signal ■ is less than TH2, and the pulse wave signal ■ and the pulse wave signal ■
When the difference between the pulse wave signal and the pulse wave signal is equal to or greater than the ninth predetermined value 09, the pulse wave signal ■ is determined to be noise.

In this way, even if there are variations in the amplitude of the pulse wave signal PA, it is possible to accurately determine the diastolic blood pressure DIA, and also to
) and the second determination method of systolic blood pressure SYS explained in
Since the determination algorithm can be used, programs can be standardized and simplified.

FIG. 12 is a diagram illustrating sound measurement with a gate according to the embodiment. Sound measurement with gate is based on the onset and disappearance of Korotkoff sounds gated with gate signal KG.
This is to determine S and diastolic blood pressure DIA.

Note that it is known to measure Korotkoff sound within a gate. Further, the portion for determining the appearance and extinction of Korotkoff sounds, which is a characteristic portion of this embodiment, is the same as the sound measurement without gates, which will be described later, and therefore the description thereof will be omitted.

FIG. 7 is a flowchart of sound measurement without a gate according to the embodiment. In step S701, initial settings are performed, that is,
Register BN that stores the magnitude of base noise (BN)
, a flag 2ndF that holds that the sound was detected in the first
, a flag DIAF indicating that the control is in the step of detecting the diastolic blood pressure DIA, and a flag KENDF indicating that the sound measurement has been completed without this gate, respectively.

In step 5702, base noise detection rBN detection”
I do. In sound measurement without a gate, the gate signal KG for K sound detection is not used, so unique processing must be used to distinguish between noise and Korotkoff sound. Therefore,
It is necessary to perform BN detection at the beginning of measurement.

This BN detection is performed, for example, by the method shown in FIG. In the figure, the two seconds after the start of measurement are divided into 16 equal parts, each with 12
A 5m5 window is set and the noise level within each window is measured.

The measurement of the noise level within this window is, for example, the first
Perform the method shown in Figure 4. The vertical axis of the figure is the 8-bit A/D conversion output, and the K sound signal KS is the A/D conversion output level O.
Varies within the range of ~255. The horizontal axis is time t, and when measuring within a window, the noise level is measured within an interval of 1=0 to 125 m5. The sound signal KS observed within each window is a signal having various amplitudes that pulsates around the approximate center of the vertical axis. Therefore,
detecting peaks and bottoms within the window and storing the respective values in a peak register KP and a bottom register KB;
Furthermore, the difference (KP-KB) between these values is determined and used as the noise level within the window. Note that if the initial pressurization of the cuff 41 is not sufficient, Korotkoff sounds may be included within the window as shown in the figure.

Returning to FIG. 15, the area counter AC is incremented by 1 for each window and counts from 0 to 15. The control is first performed in a large scale for each window with AC=0.1 (KPO, K
BO), (KPI.

KBI), take the larger of KPO and KPI and store it in the peak memory Po of window O, and also take the smaller of KBO and KBI and store it in window O.
is stored in the bottom memory Bo. Next, the control is AC=1.
(KPI, KBI) (K
P2. KB2) and KPI and KP2 among them.
The larger one of KBI and KB2 is taken and stored in the peak memory P1 of window 1, and the smaller one of KBI and KB2 is taken and stored in the bottom memory B of window 1. Thereafter, up to P 16+ B +5 are obtained in the same manner. Then, for each of these 15 pairs, the difference (Pa Bo) (P + B
, ), ..., (Pus B+s) is calculated and the corresponding 1
Take the smallest of the 5 differences and calculate the base noise BN
shall be. The above details will be described later according to the flowcharts of FIGS. 8 and 9.

Note that the method for determining the base noise BN is not limited to this. The base noise BN may be, for example, an average value of 15 differences. Alternatively, it may be the average value of n smaller differences extracted from among the 15 differences.

Returning to FIG. 7, in step 5703, the systolic blood pressure SYS
Clears the counter SC that determines the determination timing of the Korotkoff sound, and the register T that determines the determination timing of the extinction of the Korotkoff sound, respectively, and the register TH that holds the threshold value for the Korotkoff sound.
K is the threshold value BN+CI for determining systolic blood pressure (for example, 100
mmHg). Here, BN is step 57
02, and CI is the first predetermined value. In step 5704, registers KP, K
B. A control flag R3KF and a timer t for initializing the message detection process midway, which will be described later, are respectively reset.

In step 5705, the sound sampling interrupt (KSIN)
T). Control proceeds to step 5706 and waits (executes IDLE).

FIG. 10 is a flowchart of sound sampling interrupt processing in the embodiment. In step 5100I, the sound signal KS is
In step 51002, the control flag RSKF is checked. Since RSKF=0 at first, the process advances to step S1003 and the contents of timer t are checked.

Since 1=0 at first, the sound signal KS sampled in registers KP and KB in step S1008 is
Set. This is initialization of registers KP and KB.

In step S1006, the timer t is set to +Δt (for example, if the sampling period is 1 mS, Δt=1 m, S
)I do.

Also, if 1=0 is not determined in step S1003,
Proceeding to step S1004, calculation (KP-, S
)I do. When (KP-KS)<○, K against KP
Since an increase in S is recognized, the process advances to step 51005, and the contents of KP are updated with the contents of KS. Also (KP-K
When S)≧○, no increase in KS with respect to KP is recognized, so proceed to step S I 009 and perform the calculation (KB-
KS). When (KB-KS)≦0, a decrease in KS relative to KB is not recognized, so the timer t is simply updated. However, when K, B-KS)>O, it is recognized that KS has decreased relative to KB, so the process advances to step 51010 and updates the contents of KS that are the contents of KB. Thus,
Register KP or KB every time the K sound signal KS is sampled.
The contents of will be updated.

Returning to FIG. 7, when the sound sampling interrupt process is completed, the process is input to step 5707.

Here, KSINT is disabled to prevent duplicate interrupts. In step 5708, (KP-KB)kTH
Determine whether it is K or not. If YES, the first Korotkoff sound has been detected, so proceed to step 5709 and set flag 2.
Check ndF.

Since it is the first Korotkoff sound, 2ndF=O,
Control proceeds to step 5711, where 2ndF is set to 1. In step 5712, the cuff pressure PS at that time is saved in the external memory 25.

In step 5713, DIAF is checked. Since we are not yet in the process of measuring the diastolic blood pressure DIA, DrAF=O,
Control proceeds to step 5715 where counter SC is incremented by +1.
do. In step 5716, the counter SC is checked. Since 5C=1 here, the process returns to step 5704.

In step 5704, registers KP, KB, timers, etc. are reset to detect the next Korotkoff tone.

When the next Korotkoff sound is detected in step 5708, the process advances to step 5709, and this time 2ndF=1. Control proceeds to step 5710, where it is determined whether t≧0.3S. According to medical statistics, the next Korotkoff sound does not occur within 0.3 seconds of the previous Korotkoff sound.

Therefore, the sound signal in this section is ignored to prevent erroneous determination due to noise. Therefore, step 571
When the determination of 0 is NO, the process advances to step 5714, and 1 is set in the control flag R3KF.

Returning to FIG. 10, it is determined in step 51002 that RSKF
When = 1, proceed to step S1007 and RSKF
Reset. In step 5LOO8, register KP
and set the sampled sound signal KS in KB. Thus, the previous contents of KP and KB are canceled and the detection of Korotkoff sounds is restarted anew.

Returning to FIG. 7, it is determined in step 5710 that t≧0.3S
When , a normal Korotkoff sound is detected. Control continues with steps 5712, 5713, 5715, and 5716;
This time, since 5C=2, proceed to step 5717.
Set DIAF to 1. In step 5718, a threshold value BN+C2 for determining the diastolic blood pressure is set in the register THK. Here, C2 is a second predetermined value. Step 5
At step 719, if the display priority allows, the first cuff pressure value PS saved at step 5712 is displayed in the systolic blood pressure SYS section of the display LCD 29.

The next step is to detect the diastolic blood pressure DIA.

If the determination in step 5708 is No, step 57
20, it is determined whether t=23 or not. According to medical statistics, the limit for the next Korotkoff sound to occur is considered to be within 2 seconds from the previous Korotkoff sound. Therefore, t=2S
Until t=23, the process returns to step 5705 and waits for the Korotkoff sound to occur. However, if the Korotkoff sound does not occur even at t=23, the process proceeds to step 5721 and the DIAF
Find out. If DIAF=O, return to step 5705. Such a case may occur from the start of measurement until the first Korotkoff sound is detected. now,
Since DIAF=1, proceed to step 5722,
Perform +t on register T. In the first case, T=23.

In step 5723, it is checked whether it is a Tan beat or not.

Here, in the case of (A) mode measurement, the Korotkoff sound is 2
We would like to detect the missing beat, but since the period of one beat of the Korotkoff sound differs depending on the subject, the period of one beat of the Korotkoff sound is determined in advance by a process not shown, and in step 5723, T> (period of 2×1 beat).

In addition, in the case of (C) mode measurement for a person with an auscultation gap, it is detected that the Korotkoff sound is missing by 5 beats, so in step 5723, it is checked whether T> (period of 5 × 1 beat). . In either case, if the test subject's period for one beat is relatively long (for example, approximately 2 seconds), the process of step 5722 will be performed multiple times, and in the case of the second time, T = 43. In this case, T=63. Thus, if the determination in step 5723 is YES, the process proceeds to step 5724, and if the display priority allows, the last cuff pressure value PS saved in step 5712 is displayed in the diastolic blood pressure DIA section of the display device LCD 29. .

In step 5725, the sound measurement end flag KENDF is set to 1 without gate.

FIG. 8 is a flowchart of BN detection according to the embodiment. In step S801, area counter AC is cleared.

In step 5802, the contents of registers KP, KB, and timer t are cleared.

Step 5803 enables KSINT, and step 5804 waits (executes IDLE).

In the same manner as above, the sound signal K is generated according to the process shown in FIG.
Once S has been sampled, control enters step 5805 of FIG. Here, KSINT is disabled, and the process proceeds to step 8806, where it is determined whether t=125 ms. Since the period until t=125 mS is within one window, control returns to step 5803.

Eventually, when t=125 ms is reached, the process advances to step 5807 and area noise detection (AN detection) is performed.

FIG. 9 is a flowchart of AN detection according to the embodiment. In step S901, the contents of area counter AC are checked. Since AC=O at first, the process proceeds to step 5911, and the contents of registers KP and KB are set in registers PA and BA, respectively.

Returning to FIG. 8, in step 5808, the area counter A
Check C. Since AC is not yet 15, +1 is added to AC in step 5814, and the process returns to step 5802.

Returning to Figure 9, the next time you enter this process, AC=
It is 1. Control continues to step 5904 where register P
Registers KP are placed in a and Ba, respectively.

Set the contents of KB. Now, the peak bottom PA of AC=O, BA and the peak bottom Ps of AC=1, Ba
are all available. In step 5905, the calculation (PA −Pa
), and when (PA Pa)<0, Pa is high, so the process advances to step 5907, and p (
Set the contents of Pa in c). Here, (C) is the address pointed to by area counter AC. Also (pA-pB
)≧O, PA is high, so the process proceeds to step 5906, and the content of PA is set in P (C). Similarly,
In step 5908, calculation (BA-BB) is performed, and (
, BA - BB )>O, since B is low, the process advances to step 5910 and Ba is stored in B (C) of the external memory 25.
Set the contents of. Also, when (BA-B11)≦O, BA is low, so the process advances to step 5909, and B (C
) to set the contents of BA.

Thus, the next time this process is entered, AC will be 2 or more. Control proceeds to step 5903 where register P
The contents of a and B8 are shifted into registers PA and BA, respectively. In step 5904, registers KP are placed in registers Pa and Ba, respectively.

Set the contents of KB. Now, the peak bottom PA, BA of AC=1 and the peak bottom Pa, Ba of AC=2
are now available. Below, area noise P (15
), B (15) are detected.

Returning to FIG. 8, when AC=15 is determined in step 5808, the process proceeds to step 5809, and the base noise BH (
Set the smallest of PJ BJ). This is the detected base noise BM. Here, j is O~15
It is. In step 5810, it is determined whether or not there is a value in (PJ - BJ) that exceeds (BN+C3).

Here, 03 is the third predetermined value, for example (BN+C3
) = 100 mmHg. If the determination is No, the process exits. Further, when YES, it indicates that a signal corresponding to Korotkoff sound was present during the detection of base noise, which indicates that the initial pressure applied by the rubber ball 44 is insufficient. Control is at step 58
Proceeding to step 11, since the systolic blood pressure SYS can no longer be measured, DIAF is set to 1, and further step 5812
"-1111" is displayed preferentially on the systolic blood pressure SYS display section of the display device LCD 29. In this case,
In the process of FIG. 7, the state of D I AF=1 is input to step 5703, but if the measurement is continued as it is, it is possible to ignore the measurement of the systolic blood pressure SYS and only measure the diastolic blood pressure DIA.

Furthermore, as described above, the user can reapply pressure during the measurement.

[Effects of the Invention] As described above, according to the present invention, it is possible to distinguish between a true Korotkoff sound signal and noise based only on the Korotkoff sound signal, and to perform a unique and accurate blood pressure determination.

[Brief explanation of drawings]

Fig. 1 is a block diagram of the electronic blood pressure monitor of the embodiment, Fig. 2 is an external view of the electronic blood pressure monitor of the embodiment, Fig. 3 is a flowchart of main processing including oscillometric measurement of the embodiment, and Fig. 4 is a block diagram of the electronic blood pressure monitor of the embodiment. Flowchart showing details of pressurization detection in the example. Figure 5 is a flowchart for pulse wave detection in the example. Figure 6 is a flowchart for oscillometric analysis in the example. Figure 7 is sound measurement without gate in the example. 8 is a flowchart of BN detection according to the embodiment, FIG. 9 is a flowchart of AN detection according to the embodiment, FIG. 10 is a flowchart of sound sampling interrupt processing according to the embodiment, and FIGS. (C) is a diagram explaining the pressurization detection method of the embodiment, Figure 12 is a diagram explaining the pulse wave detection and sound measurement method with a gate in the embodiment, and Figures 13 (A) to (D). Figure 14 is a diagram explaining the oscillometric analysis of the example. Figure 14 is a diagram explaining the method of measuring the noise level within the window of the example. Figure 15 is a diagram explaining the base noise (BN) detection method of the example. It is a diagram. In the figure, 10...main unit, 11...rechargeable battery, 12...
・Power control, 13...Power switch, 14・
...Reference oscillation section, 15...Mode switch, 16...
・Filter amplifier, 17... Reference power supply section, 18...
Pressure detection section, 19... Amplifier, 20... Microprocessor section (LSIC), 21... A/D conversion section, 2
2... Control unit, 23... Korotkoff sound/pulse wave recognition unit, 24... Display drive unit, 25... External memory, 26
... A/D conversion section, 27... Drive section, 28... Exhaust valve, 29... Display device (LCD), 30...
Buzzer 31... Start switch, 40... Pressure part, 41... Cuff, 42... Constant speed (low speed) exhaust valve, 43... Manual exhaust valve, 43a... Knob,
43b...Full open switch, 44...Rubber bulb, 51.
...Microphone, 60...Charger, 70...Tube. Figure 6 Figure 9 Figure 10 Figure 12 Figure 13 (A) m-''-m- YS Figure 13 (C) 204- l-m- YS Figure 13 (B) IA Figure (D>

Claims (6)

[Claims] (1) An electronic sphygmomanometer that determines blood pressure based on Korotkoff sound, comprising: a detection means for detecting the Korotkoff sound signal; a window setting means for setting a window for a predetermined time after the start of measurement; and a detection means within the window. It is characterized by comprising: a peak-bottom detection means for detecting each value of a peak and a bottom of a detection signal; and a setting means for setting a base noise level of a signal detected by the detection means based on each of the detected peak and bottom values. Electronic blood pressure monitor. (2) The electronic blood pressure monitor according to claim 1, wherein the window setting means sets two or more n windows during a predetermined time after the start of measurement. (3) The electronic blood pressure monitor according to claim 2, wherein the peak-bottom detection means detects each peak and bottom value for each of the n windows. (4) The setting means calculates the difference between the peak value and the bottom value for each of the n sets of detected peak and bottom values, and sets the one with the smallest difference as the base noise level. The electronic blood pressure monitor according to item 3. (5) The setting means calculates the difference between the peak value and the bottom value for each of the n sets of detected peak and bottom values, and sets the average value of these differences as the base noise level. The electronic blood pressure monitor according to item 3. (6) The setting means calculates the difference between the peak value and the bottom value for each of the n sets of detected peak and bottom values, and calculates the average value of m (m<n) from the smallest of these differences. 4. The electronic blood pressure monitor according to claim 3, wherein the base noise level is set as the base noise level.