JPH0767844A - Device and method for detecting qrs wave in long time electrocardiogram - Google Patents

Abstract

PURPOSE:To detect an exact QRS wave from the long time electrocardiogram. CONSTITUTION:An upper arbitrary value than a noise generated in a base line in an electrocardiogram waveform is set temporarily as a threshold (S53), a waveform exceeding the set threshold is detected as an R wave (S54), and from detection result information, whether a threshold level set previously is an appropriate level or not is decided (S55), and in the case it is decided not to be appropriate, a new appropriate threshold level is calculated by considering a detected R wave and an R wave detection processing is executed as a new threshold (S56-S55), an optimal threshold is set and a QRS candidate point is detected. Thereafter, QRS likelihood of the detected QRS candidate point is set (S60), the candidate point whose QRS likelihood is low is excluded from the QRS candidate (S61) and an R-R interval of the remaining candidate points is checked, and in the case the R-R interval does not exist within a prescribed interval range, the candidate point excluded from the QRS candidate previously is restored (S62) and it is set as a final QRS detection point (S63).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a QRS wave detection method in a long-term electrocardiogram.

[0002]

2. Description of the Related Art As is well known, a long-term electrocardiogram is a record of an electrocardiogram of a subject over a long period of time (for example, 24 hours), and is useful for finding an arrhythmia or the like which cannot be detected by only the electrocardiogram in a short time. That is, by analyzing the QRS wave or the like in the long-term electrocardiogram to know what kind of arrhythmia has occurred in which time zone in the daily life, more accurate diagnosis of the subject is possible. It is what

In order to detect an arrhythmia or the like from this long-term electrocardiogram, it is necessary to accurately detect the QRS wave in the long-term electrocardiogram. This is P wave, QR in electrocardiogram
This is because there are S-waves, T-waves, etc., and among these waves, the QRS wave is the largest wave, and various analyzes etc. are performed based on this QRS wave. Conventionally, the following method has been adopted to detect this QRS wave.

That is, with reference to the baseline of the electrocardiographic waveform, noise generated on the baseline is detected, an arbitrary value above the noise is set as a threshold, and a waveform above the threshold is regarded as an R wave. The QRS wave was detected. The threshold value was set to a final value based on the subject's normal waveform and experience.

[0005]

However, the electrocardiogram usually changes according to the change of time, and even a normal person may have a low QRS wave level.
Even if there is no change in the electrocardiogram, a large amount of noise may occur in the electrocardiogram due to body movement or the like. In the above-mentioned conventional QRS wave detection method, since the QRS wave is detected with the set level as a reference after setting the initial level, noise that is originally not the QRS wave is detected at the time as described above.
There was a risk that it would be detected as an RS wave. Also, QRS
Even though it was a wave, there were cases where it was not detected as a QRS wave.

[0006]

The present invention has been made for the purpose of solving the above-mentioned problems, and has the following structure as one means for solving the above-mentioned problems. That is, an R wave detecting means for temporarily setting an arbitrary value above the noise occurring on the baseline in the electrocardiogram waveform as a threshold value and detecting a waveform having a value equal to or higher than the set threshold value as an R wave, and a detection result by the R wave detecting means Based on the information, it is determined whether or not the previously set threshold level is a valid level, and if it is determined that the level is not valid, a new valid threshold level is calculated by considering the detected R wave, and a new threshold level is calculated. Optimum threshold value setting means for executing the detection processing by the R wave detection means as a threshold value, QRS likelihood setting means for setting the QRS likelihood of a QRS candidate point according to the R wave detected by the R wave detection means, and the QRS The candidate point having a low QRS likelihood in the likelihood setting means is excluded from the QRS candidates, and the RR intervals of the remaining candidate points are examined. If the RR interval is not within the predetermined interval range, the candidate that has been excluded from the QRS candidate first. Resurrecting points So it and a QRS detector for the final QRS detection points.

[0007]

With the above-described structure, first, as the initial search, the threshold level is changed to, for example, the upper and lower sides of the electrocardiogram information to set the optimum level to detect the QRS candidate points. The QRS wave can be detected by a method closer to human thought by using the information of (3), and the QRS wave can be detected very accurately. Therefore, by utilizing the QRS wave detection of the present invention, a reliable electrocardiographic analysis can be performed.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment according to the present invention will be described in detail below with reference to the drawings. FIG. 1 is a block diagram of an electrocardiogram analysis recording apparatus according to an embodiment of the present invention, in which 11 is an RO.
Controlling the whole of this embodiment according to the programs stored in M12 according to, for example, the programs shown in FIGS. 4 to 7, and detecting the QRS wave in the electrocardiogram by the QRS wave detection method described later,
The ST level and the like are compared and the electrocardiogram information is analyzed, and the comparison result and the analysis result are sent to the printer control unit 22 and the display control unit 25 for print output or display output. Reference numeral 12 is a ROM that stores various parameters other than the above programs.

Reference numeral 13 is an electrocardiographic recording device (Holter electrocardiograph).
The cassette tape reader 14 reads the electrocardiographic waveform from the cassette tape 30 on which the electrocardiographic waveform is recorded. Reference numeral 14 controls the cassette tape reader 13 to read the electrocardiographic waveform from the cassette tape and binarize it for output to the memory 15. The reading circuit 15 is a memory capable of holding at least several beats of electrocardiographic waveforms from the reading circuit 14. 16 is a memory 15
More electrocardiographic waveform is read, and QRS
It is a QRS wave detection circuit that detects waves, and also detects R wave peaks.

Reference numeral 17 denotes an ECG data compression circuit for time-compressing a long electrocardiographic waveform read from the memory 15 to shape the waveform and vertically writing by dividing it for each minute, for example. It is sent to the unit 22 and the display control unit 25, and is stored in the predetermined compressed waveform storage area of the page buffers 22a and 25a. Reference numeral 18 is an ST deviation sampling circuit for sampling an ST value of an electrocardiographic waveform read from the memory 15 at a predetermined arbitrary timing to create an ST trend graph. Reference numeral 19 is a QRS wave detection circuit 1.
6 is an HR detection circuit for measuring the R wave peak time interval of the QRS waveform detected in 6 and detecting the heart rate.

Reference numeral 22 denotes an analysis result from the control unit 11, which is EC
The compressed electrocardiogram waveform from the G data compression circuit 17, the ST trend graph from the ST deviation sampling circuit 18, the HR trend graph data from the HR detection circuit 19, etc. are placed at the printout position of the page buffer 22a that stores output information. A printer control unit that expands and prints the expanded data from the printer 23. Reference numeral 23 is a printer for printing out the waveform information stored in the page buffer 22a, and reference numeral 24 is a timer circuit.

Numeral 25 is an analysis result from the control unit 11, which is EC
A compressed electrocardiogram waveform from the G data compression circuit 17, an ST trend graph from the ST deviation sampling circuit 18, and
The HR trend graph data from the HR detection circuit 19 is aggregated and expanded at the display output position of the page buffer 25a, and the expanded controller displays the expanded data on the display screen of the display device 21. A display device for displaying data.

The QRS wave detection circuit 16, the ECG data compression circuit 17, and the ST deviation sampling circuit 18
Once activated, the memory 15
The electrocardiographic waveform written inside is always read and each processing is executed. FIG. 2 is a block diagram of an electrocardiographic waveform recording device for recording an electrocardiographic waveform on a cassette tape 30 from a subject, and 31 to 33 are bioinduction electrodes that are fixed to the surface of a living body to derive an electrocardiographic waveform. Is the biological induction electrodes 31 to 3
3, an amplifier circuit for amplifying the electrocardiographic waveform derived from 3 and outputting it to the writing circuit 42, 42 a writing circuit for recording the electrocardiographic waveform from the amplifier circuit 41 in the cassette tape recorder 45, 4
A cassette tape recorder 5 records data from the writing circuit 42 on the cassette tape 30.

The living body guiding electrodes 31 to 3 having the above construction
The electrocardiographic waveform derived by 3 is usually a repetition of the spike wave shown in FIG. 3, and is sequentially named P, Q, R, S, T. P waves are caused by atrial excitation and QRS are caused by ventricular excitation. The T-wave is also produced by the extinction of the ventricles. In addition, following the T wave, there is often a loose start U
Waves may occur.

Then, arrhythmia and impaired conduction of excitation are determined from the temporal relationship of these spike waves, and ischemic heart diseases such as myocardial infarction, myocarditis, pericarditis, etc. are determined from changes in the shape of the spike waves. Diagnosis of left and right atrium and ventricular hypertrophy, and further abnormality of electrolyte, drug action, and internal secret abnormality. In order to make these diagnoses, it is indispensable to reliably detect only the spike wave and clearly separate it from noise, and the QRS wave detection circuit 16 surely detects only the QRS wave important for this diagnosis. The R wave peak is also detected at the same time so that the change in the spike wave change portion can be easily recognized.

From the electrocardiogram, the most frequently used for finding and diagnosing the above-mentioned diseases are the inclination and deviation of the ST site. Therefore, in this embodiment, the crest value at an arbitrary measurement point is measured by the ST deviation sampling circuit 18 for creating the ST trend graph. ST deviation sampling circuit 1
Reference numeral 8 synchronizes with the R wave peak and samples the peak value which is the potential difference from the reference point b at the reference level at the specific point on the specific line indicated by the alternate long and short dash line after the elapse of a predetermined time from the R wave. The specific point may be a ₁ shown by a solid line in FIG. 3 or a ₂ shown by a chain line.

In the figure, the peak value becomes negative in the case of a ₁ and becomes positive in the case of a ₂ . The ST trend graph is obtained by sequentially displaying these peak values as a trend graph. The electrocardiogram information analysis control of the apparatus of this embodiment having the above configuration will be described in detail with reference to the flowchart of FIG.

When the cassette tape 30 on which the electrocardiographic waveform is recorded is inserted into the cassette tape reader 13 of the apparatus and activated, the control of the control unit 11 is controlled by step S1 in FIG.
Proceed to. In step S1, the control unit 11 instructs the reading circuit 14 to control the cassette tape reader 13 so that the long-time electrocardiographic waveforms that are sequentially recorded are read out and at the same time the memory 15 is read.
Start the process of writing in. It is desirable that the memory 15 has a capacity of two beats or more of an electrocardiographic waveform. At this time, the read time information is stored in the timer circuit 24.
Set the time to the same time as the recording. This reading process is continuously performed thereafter. Then, the control unit 11 starts the electrocardiogram information analysis process in parallel with this reading process.

Then, the QRS wave detection circuit 16 is activated in the subsequent step S2, and the ECG data compression circuit 17 is activated in the subsequent step S3. After that, in step S4, the ST deviation sampling circuit 18 and the HR detection circuit 19 are activated,
In step S5, the display controller 20 is activated. Then, in step S6, it is checked whether or not the print instruction is given,
If the print instruction has not been issued, the process proceeds to step S8. On the other hand, if the print instruction is given, step S7
Then, the printer controller 22 is activated and the process proceeds to step S8.

At step S8, it is determined whether the reading of the electrocardiogram information recorded on the cassette tape 30 has been completed or the display / printing process has been instructed by an instruction key (not shown) or the like to end the data. Find out.
If the data has not ended, the process returns to step S6, and if the data has ended, in step S9, the respective circuits that were previously activated are deactivated and the process ends.

In the above process, when each circuit is activated, the process assigned to each circuit is executed independently or by utilizing the interrupt process to the control unit 11 or the like for each circuit. In addition, as the processing for analyzing the electrocardiographic waveform, a predetermined electrocardiographic analysis processing including the above-described processing for detecting a change in the ST waveform due to ischemia is performed. In the analysis process of electrocardiogram waveform, general analysis process except QRS wave detection process and method of generating compressed waveform from the electrocardiogram waveform are described in "Holter long-term electrocardiogram / automatic analysis Since the details are described in "Development of Device", detailed description is omitted.

Next, the QRS wave detection processing of the QRS detection circuit 16, the generation processing of the compressed electrocardiographic waveform in the ECG data compression circuit 17, and the control of the ST deviation sampling circuit 18 and the HR detection circuit 19 will be described. First, the QRS wave detection method of the QRS detection circuit 16 of the present embodiment will be described below with reference to the flowchart of FIG.

When the QRS detection circuit 16 is activated by the control unit 11, the process proceeds from step S51 to step S52 to read the electrocardiographic waveform from the memory 15. This reading continues until the circuit is subsequently deactivated. Subsequently, in step S53, the electrocardiographic waveform is filtered, the baseline generated in the electrocardiographic waveform is used as a reference, noise generated in the baseline is detected, and an arbitrary value above the noise is set as a threshold. Then, a waveform having a threshold value or more set in the subsequent step S54 is regarded as an R wave, and provisional R wave detection (initial search) is performed. Then, in step S55, it is checked from the detection result information whether or not the previously set threshold level is an appropriate level. Here, if it is determined that the threshold level is not appropriate, the process proceeds to step S56, and the appropriate threshold level is calculated by considering the height and interval of the detected peaks, the average value of the past normal peaks, etc., and set as a new threshold. Reset. Then, the process returns to step S53, and the detection process for the next electrocardiographic waveform is performed. By repeating such trial and error, the optimum threshold value is determined, and the QRS of each channel of the read electrocardiogram waveform is determined.
Detect candidate points.

If the threshold level is appropriate, the process proceeds from step S55 to step S60, the signal quality near each candidate point, the noise condition, the presence or absence of another channel candidate, etc. are checked, and the QRS-likeness grade is determined from these information. To provide. Specifically, the QRS corresponding to the detection result of the temporary R wave
Regarding wave candidate points, RR interval, signal quality, noise situation, QRS wave shape (QRS shape, width, height, direction), T wave likeness, and QRS wave candidate points of other channels are checked. For example, a low grade is attached to a noisy signal in the vicinity, an extremely short RR interval, or a QRS wave in which only one (one) channel is detected.

Then, in the next step S61, candidate points having a low grade among the QRS candidate points are once canceled from the detection result, and those having an extremely short RR interval and those considered to be T waves are also canceled. And the following step S6
R in the state where the candidate point in 2 is canceled in step S61
Check R-interval. When the RR interval is extremely long, the possibility of missing detection is estimated, the average RR interval is re-searched around a position where QRS waves are likely to occur, and the candidate points that were once canceled are restored. . In this case, when the QRS wave of the other channel is detected, the QRS wave recovery processing is performed with reference to the QRS wave detection position.

Then, the QRS candidate point determined by the above processing is used as a detected QRS wave, and R in this QRS wave is set.
Let the wave be the detected R wave. Then, this detection result is used by the control unit 1
1, output to the ST deviation sampling circuit 18 and the HR detection circuit 19. Succeedingly, in a step S64, it is determined whether or not the data ends (data display / print processing ends).
If the data has not ended, the process returns to step S52, and the QRS wave detection process for the next electrocardiographic waveform is performed. On the other hand, if it is the end of data, the process ends.

By adopting the QRS detection method as described above, the threshold level in the initial search is changed up and down to set the optimum level to detect the QRS candidate point, and further, regarding the candidate point, QRS waves can be detected by a method closer to human thinking by using various kinds of information, and very accurate QRS waves can be detected. QRS by QRS wave detection circuit 16 of this embodiment
The flow of detection is shown in FIG.

Next, the outline of the processing for generating the compressed electrocardiographic waveform in the ECG data compression circuit 17 will be described below with reference to the flowchart of FIG. ECG data compression circuit 20
Is activated by the control unit 11, step S11
By proceeding to step S12, the electrocardiographic waveform is read from the memory 15. This reading continues until the circuit is subsequently deactivated. Then, in step S13, first, the initially designated electrocardiographic waveform is read out, the read electrocardiographic waveform is compressed to a predetermined scale by a known method, and the internal memory 20a is stored.
Will be stored in sequence. Then, in step S14, it is checked whether or not the compressed waveform for one row has been generated. When the compressed waveform generation for one row is not completed, the process proceeds to step S15, and it is checked whether or not there is a deenergizing instruction. If the power-off instruction has not been received, the process returns to step S12. The above-mentioned initialized waveforms are, for example, two measurement waveforms of one channel.
If there are two channel waveforms, the measurement waveform for one channel is used first.

On the other hand, when the compressed electrocardiographic waveform for one row can be generated in step S14, the process proceeds to step S16, and the generated compressed electrocardiographic waveform is sent to the printer control unit 22 and the display control unit 25. Printer control unit 22 and display control unit 25
Expands to the time axis position corresponding to the display output position of the page buffers 22a and 25a which stores a predetermined amount of output information containing this compressed electrocardiographic waveform, returns to step S12, and shifts to processing of the next electrocardiographic waveform. To do.

If there is a power-off instruction in step S15, the flow advances to step S17 to send it to the printer controller 22 and the display controller 25. The printer control unit 22 and the display control unit 25 expand the output information containing the compressed electrocardiographic waveform to the corresponding time axis positions of the display output positions of the page buffers 22a and 25a that store a predetermined amount and end the processing.

In the above description, the control is performed so that the compressed data for one row is written into the page buffers 22a and 25a every time the compressed waveform is generated. However, the memory 20a is not provided and the page buffers 22a and 25a are provided for each compressed waveform. It may be controlled to write data in the page buffers 22a and 25a every five rows of one waveform.

Further, another ST deviation sampling circuit 1
8. The control of the HR detection circuit 19 will be described with reference to the flowchart of FIG. Although FIG. 7 shows that the processing of each circuit is performed in time series, the processing of each circuit is actually performed independently. In the following description, for simplification of the description, the operation of each circuit will be described in time series. The QRS wave detection circuit 16 performs the QRS wave detection processing and the R wave peak point detection processing (feature point detection processing) shown in FIG. 5 described above in step S23. Then, the detection result is output to each circuit.

Therefore, the HR detection circuit 19 measures the time from the detection of the R wave peak one beat before in the subsequent step S24. This is done by reading the clock data of the timer circuit 24. Then, in the subsequent step S30, the heart rate is calculated from the detection time interval of the characteristic point (R wave peak) measured in step S24, and the heart rate calculated in step S31 is displayed simultaneously with the time information in the display controller 20 and the printer controller 22. Output to. The display control unit 20 and the printer control unit 22 develop the output information containing the HR value on the time axis corresponding to the display output positions of the page buffers 20a and 22a that store a predetermined amount.

Then, in the next step S35, the ST deviation sampling circuit 18 is activated to obtain the ST crest value (the crest value from the reference level) at a position where a predetermined time has passed from the characteristic point as described above, and the ST crest value is obtained at that position. The ST deviation value of is sampled. Then, in the subsequent step S36, the ST deviation sampling circuit 18 sends the time information to the printer control unit 22 and the display control unit 25 at the same time. Printer control unit 22
The display control unit 25 also stores a predetermined amount of output information containing the ST deviation sampling value in the page buffer 22.
It is expanded on the time axis corresponding to the display output positions of a and 25a. For example, in the present embodiment, as shown in FIG. 9, this ST deviation and HR are expressed on the same time axis with the time axis as the vertical axis of the same time, and the display position is changed.

Then, in step S40, it is checked whether or not the data has ended (data display / print processing has ended). If the data has not ended, the process returns to step S23 and the next electrocardiographic waveform analysis process is performed. On the other hand, if it is the end of the data, the process proceeds to step S46, under the control of the display control unit 25, the data for one page (for one column) is displayed on the display device 26,
Under the control of the printer controller 22, one page of data is printed out on the printer 23.

When the data display and printout are completed, the process is completed. In FIG. 9, the compressed electrocardiogram waveform is recorded on, for example, the right two-thirds portion of the recording paper, and the ST trend and the heart rate HR corresponding to the recorded compressed electrocardiogram are recorded on the left portion of the recording paper. There is. According to the output method of FIG. 9, the ST trend and the HR are
And the compressed electrocardiogram are recorded in correspondence with each other, and if the abnormal portion of the S-T trend is traced to the right, the position of the abnormal electrocardiogram can be immediately known, and there is an advantage that diagnosis can be facilitated.

[0037]

As described above, according to the present invention, the threshold level is first changed up and down as the initial search for the electrocardiogram information to set the optimum level, and the QRS candidate point is detected. With respect to the candidate point, the QRS wave can be detected by a method closer to human thought by using various kinds of information, and the QRS wave can be detected very accurately. Therefore, by utilizing the QRS wave detection of the present invention, a reliable electrocardiographic analysis can be performed.

[Brief description of drawings]

FIG. 1 is a block diagram of an electrocardiogram analysis recording device according to an embodiment of the present invention.

FIG. 2 is a block diagram of an electrocardiographic waveform recording device for recording an electrocardiographic waveform used in this embodiment.

FIG. 3 is a diagram for explaining an electrocardiographic waveform.

FIG. 4 is a flow chart showing an electrocardiogram analysis process in this embodiment.

FIG. 5 is a flowchart showing a QRS detection process of this embodiment.

FIG. 6 is a flowchart showing an electrocardiogram waveform compression process by the ECG data compression circuit according to the present embodiment.

FIG. 7 is a flowchart showing output control of this embodiment.

FIG. 8 is a diagram showing a flow of QRS detection by the QRS wave detection circuit of the present embodiment.

FIG. 9 is a diagram showing an output example of an electrocardiographic waveform processing result of the present embodiment.

[Explanation of symbols]

 11 control section 12 ROM 13 cassette tape reader 14 reading circuit 15 memory 16 QRS wave detection circuit 17 ECG data compression circuit 18 ST deviation sampling circuit 19 HR detection circuit 22 printer control section 22a, 25a page buffer 23 printer 25 display control section 26 display Device 30 Cassette tape 31-33 Bio-induction electrode 41 Amplifier circuit 42 Writing circuit 45 Cassette tape recorder

Claims (3)

[Claims] 1. An R wave detection means for temporarily setting a threshold value to an arbitrary value above a noise occurring on a baseline in an electrocardiogram waveform and detecting a waveform equal to or higher than the set threshold value as an R wave, and the R wave detection means. It is determined whether the previously set threshold level is a valid level from the detection result information of 1., and when it is determined that it is not valid, a new valid threshold level is calculated in consideration of the detected R wave. An optimum threshold value setting means for executing the detection processing by the R wave detection means as a new threshold value, and a QRS likelihood setting means for setting the QRS likelihood of a QRS candidate point according to the detected R wave by the R wave detection means. , The candidate point having a low QRS-likeness in the QRS-likeness setting means is excluded from the QRS candidates, and the RR intervals of the remaining candidate points are examined. If the RR interval is not within the predetermined interval range, the QR is first detected.
A QRS wave detection device in a long-term electrocardiogram, comprising: a QRS detection unit that restores a candidate point removed from the S candidate to be a final QRS detection point. 2. The QRS likelihood setting means is at least R
-Q with reference to R interval, noise condition, and height of QRS wave
The QRS wave detection device for a long-term electrocardiogram according to claim 1, wherein RS-likeness is set. 3. An arbitrary value above noise generated on a baseline in an electrocardiogram waveform is provisionally set as a threshold value, a waveform equal to or higher than the set threshold value is detected as an R wave, and the detection result information is set first. It is determined whether or not the threshold level that is present is a proper level, and if it is determined that the threshold level is not appropriate, the detected R
Waves are taken into account to newly calculate an appropriate threshold level, the R wave detection processing is executed as a new threshold value, and an initial search step of setting an optimum threshold value; and an R wave converted in the initial search step are used. A QRS-likeness setting step of setting the QRS-likeness of the following QRS candidate points, a candidate point having a low QRS-likeness in the QRS-likeness setting step is excluded from the QRS candidates, and the R-R intervals of the remaining candidate points are examined, and R- If the R interval is not within the specified interval range, QR
A QRS detection method for a long-term electrocardiogram, comprising: a QRS detection step of restoring candidate points removed from the S candidates to obtain a final QRS detection point.