WO2023032119A1

WO2023032119A1 - Heartbeat detection method and heartbeat detection device

Info

Publication number: WO2023032119A1
Application number: PCT/JP2021/032266
Authority: WO
Inventors: 伸昭松浦; 利彦近藤
Original assignee: 日本電信電話株式会社
Priority date: 2021-09-02
Filing date: 2021-09-02
Publication date: 2023-03-09
Also published as: JPWO2023032119A1

Abstract

This heartbeat detection device comprises: a positive/negative inverse value calculation unit (3) for calculating a positive/negative inverse value of time difference of sampling data of an electrocardiographic waveform; an index value calculation unit (5) for calculating a first index value by subtracting, from a positive/negative inverse value at a processing-target time, the maximum value of positive/negative inverse values in first and second time ranges preceding and succeeding said processing-target time; an index value calculation unit (8) for calculating a second index value by subtracting, from the positive/negative inverse value at said processing-target time, the maximum value of positive/negative inverse values in third and fourth time ranges preceding and succeeding said processing-target time; and a heartbeat time determination unit (10) which, in the case where a prescribed condition is fulfilled by both the second index value at a first time when the first index value has exceeded a threshold and the first index value at a second time which is in a prescribed relation to said first time, determines the first time as the heartbeat time.

Description

Heartbeat detection method and heartbeat detection device

The present invention relates to a heartbeat detection method and a heartbeat detection device for detecting heartbeats (R waves) from an electrocardiogram waveform.

The ECG (Electrocardiogram) waveform is an observation of the electrical activity of the heart and consists of a continuous heartbeat waveform. One heartbeat waveform consists of components such as P wave, Q wave, R wave, S wave, T wave, etc., which reflect the activity of the atria and ventricles. Of these waves, the R wave accompanies the contraction of the ventricle and has a large amplitude. Therefore, detection of the heartbeat is often performed based on the peak derived from the R wave.

As a conventional heartbeat detection method, there is a method disclosed in Patent Document 1. The method disclosed in Patent Document 1 uses an index value that focuses on the relationship between the peak of the time difference value of the ECG waveform and the value in a certain time region before and after the peak, and uses a determination logic based on a threshold. to detect the R wave. This method has the advantage that it is possible to give sensitivity not only to the amplitude of the ECG waveform but also to the shape (time width), and it is less likely to mistakenly recognize the T wave, which is wider than the R wave, as the R wave. There is

FIG. 14 is a diagram showing an example of an ECG waveform. The ECG waveform contains one heart beat. The amplitude of the cardiac potential (peak to valley) is about 500 μV, which is smaller than the normal value (2000 to 3000 μV), but the amplitude may be attenuated depending on the position and condition of the electrodes that detect the cardiac potential.

FIG. 15 is a diagram showing a value obtained by taking the time difference of the ECG waveform in FIG. 14 and inverting the positive and negative. By taking the time difference of the ECG waveform, it can be confirmed that an upward peak accompanying a sharp drop in electrocardiographic potential from the R wave to the S wave appears along with the rhythm of the heartbeat.

The index value obtained by subtracting the maximum value of the positive/negative inversion of the time difference in the fixed time ranges A and B before and after the time t in FIG. Equivalent to height. This index value becomes a negative value at times off the peak.

FIG. 16 is a diagram plotting time-series data of index values. According to FIG. 16, the index value has a form in which peaks with time widths smaller than the interval between time ranges A and B are emphasized. When the object to be detected is the R wave, it is preferable to set the time ranges A and B to 100 ms and the interval between the time ranges A and B to 50 ms. As shown in FIG. 16, the R wave can be detected by applying a determination logic based on the threshold TH to the index value. The threshold TH can be obtained by a method such as obtaining the maximum index value in a period including peaks derived from a plurality of heartbeats and multiplying the maximum value by a coefficient of 1 or less.

However, with the method using the index value as described above, there is a possibility that even waveforms that are considered inappropriate as R waves due to their short duration may be erroneously recognized as R waves.

Japanese Patent No. 6527286

The present invention has been made to solve the above problems, and an object of the present invention is to provide a heartbeat detection method and a heartbeat detection device that do not erroneously recognize peaks with too short time widths as heartbeats.

A heartbeat detection method according to the present invention includes a first step of calculating, at each sampling time, a positive/negative inversion value of a time difference of sampled data from a sampled data string of an electrocardiogram waveform of a living body; A second step of detecting a first maximum value of the positive/negative inverted value in the time range of and the positive/negative inverted value in the second time range after the sampling time to be processed at each sampling time; a second maximum value of the positive/negative inverted value in a third time range before the sampling time to be processed and the positive/negative inverted value in a fourth time range after the sampling time to be processed; a third step of detecting at each sampling time; and a fourth step of calculating at each sampling time as a first index value a subtraction value obtained by subtracting the first maximum value from the positive/negative inversion value at the sampling time to be processed. a fifth step of calculating a subtraction value obtained by subtracting the second maximum value from the positive/negative inverted value of the sampling time to be processed as a second index value for each sampling time; the second index value at a first sampling time when the index value exceeds a predetermined threshold, and the first index value at a second sampling time having a predetermined relationship with the first sampling time, and a sixth step of setting the first sampling time to heartbeat time when a predetermined condition is satisfied.

In one configuration example of the heartbeat detection method of the present invention, the timing of the leading edge of the third time range with respect to the sampling time to be processed is the same as the timing of the leading edge of the first time range, The timing of the trailing edge of the fourth time range with respect to the sampling time to be processed is the same as the timing of the trailing edge of the second time range, and the timing of the trailing edge of the third time range and the fourth time range The interval is shorter than the interval between the first time range and the second time range.
Further, in one configuration example of the heartbeat detection method of the present invention, the sixth step includes the step of calculating the first index value at the first sampling time with respect to the first index value before the first sampling time. increases, and when the second index value at the first sampling time and the first index value at the second sampling time satisfy a predetermined condition, the first sampling The method is characterized by including a step of setting the time as the heartbeat time.
In one configuration example of the heartbeat detection method of the present invention, the interval between the third time range and the fourth time range ranges from 20 ms to 30 ms.

Further, the heartbeat detection apparatus of the present invention includes a positive/negative inversion value calculation unit configured to calculate, at each sampling time, a positive/negative inversion value of a time difference of sampling data from a sampled data string of an electrocardiogram waveform of a living body; a first maximum value of the positive/negative inverted value in the first time range before the sampling time and the positive/negative inverted value in the second time range after the sampling time to be processed at each sampling time; a first maximum value detection unit configured to detect, the positive/negative inversion value in a third time range before the sampling time to be processed, and a fourth maximum value after the sampling time to be processed a second maximum value detection unit configured to detect a second maximum value of the positive/negative inverted values of the time range at each sampling time; a first index value calculator configured to calculate a subtraction value obtained by subtracting the maximum value of 1 as a first index value at each sampling time; a second index value calculation unit configured to calculate a subtraction value obtained by subtracting the maximum value of 2 as a second index value at each sampling time; When the second index value at one sampling time and the first index value at a second sampling time having a predetermined relationship with the first sampling time satisfy a predetermined condition, the first and a heartbeat time determination unit configured to determine one sampling time as a heartbeat time.

In one configuration example of the heartbeat detection device of the present invention, the timing of the leading edge of the third time range with respect to the sampling time to be processed is the same as the timing of the leading edge of the first time range, The timing of the trailing edge of the fourth time range with respect to the sampling time to be processed is the same as the timing of the trailing edge of the second time range, and the timing of the trailing edge of the third time range and the fourth time range The interval is shorter than the interval between the first time range and the second time range.
Further, in one configuration example of the heartbeat detection device of the present invention, the heartbeat time determination unit may determine the first index value before the first sampling time at the first sampling time. increases, and when the second index value at the first sampling time and the first index value at the second sampling time satisfy a predetermined condition, the first sampling It is characterized in that the time is the heartbeat time.
In one configuration example of the heartbeat detection device of the present invention, the interval between the third time range and the fourth time range is in the range of 20 ms to 30 ms.

According to the present invention, the first index value obtained by subtracting the maximum value of the time difference positive/negative inversion values in the first and second time ranges before and after the time difference positive/negative inversion value at the sampling time to be processed; and a second index value obtained by subtracting the maximum value of the time difference positive/negative inversion values in the preceding and succeeding third and fourth time ranges from the time difference positive/negative inversion value at the sampling time, and the first index value is the threshold and the first index value at a second sampling time having a predetermined relationship with the first sampling time satisfy a predetermined condition. Let the sampling time of 1 be the heartbeat time. As a result, in the present invention, it is possible to detect peaks of index values that are likely to be R waves, and to exclude peaks of index values that are less credible as R waves and whose duration is too short, thereby accurately detecting heartbeats. It can be performed.

FIG. 1 is a diagram showing time difference positive/negative inversion values of an ECG waveform. FIG. 2 is a diagram showing time-series data of the first index value and the second index value of the ECG waveform. 3 is an enlarged view of FIG. 2. FIG. FIG. 4 is a diagram showing an example of a pseudo waveform imitating an ECG waveform. FIG. 5 is a diagram showing time-series data of the first index value and the second index value of the pseudo waveform. 6A to 6C are diagrams showing time-series data of the first index value and the second index value of the ECG waveform. 7A to 7C are diagrams showing time-series data of the first index value and the second index value of the pseudo waveform. FIG. 8 is a block diagram showing the configuration of a heartbeat detection device according to an embodiment of the present invention. FIG. 9 is a block diagram showing a configuration example of the maximum value detection unit, the index value calculation unit, and the index value storage unit of the heartbeat detection device according to the embodiment of the present invention. FIG. 10 is a flowchart illustrating a heartbeat detection method according to an embodiment of the invention. FIG. 11 is a flowchart for explaining a method of calculating the first index value. FIG. 12 is a flowchart illustrating a method of calculating the second index value. FIG. 13 is a block diagram showing a configuration example of a computer that implements the heartbeat detection device according to the embodiment of the present invention. FIG. 14 is a diagram showing an example of an ECG waveform. FIG. 15 is a diagram showing time difference positive/negative inversion values of the ECG waveform of FIG. FIG. 16 is a diagram showing time-series data of index values calculated by a conventional heartbeat detection method.

[Principle of Invention]
FIG. 1 shows time difference positive/negative inversion values of the same ECG waveform as in FIG. The difference from FIG. 15 is that the second time ranges C and D are shown together. The timing of the leading edge of time range C with respect to the sampling time to be processed (t in the example of FIG. 1) is the same as the timing of the leading edge of time range A. FIG. Also, the timing of the trailing edge of the time range D with respect to the sampling time to be processed is the same as the timing of the trailing edge of the time range B. FIG. On the other hand, the interval between time ranges C and D is shorter than the interval between time ranges A and B. In the examples of FIGS. 2 to 4 below, the interval between time ranges C and D is 24 ms.

The time difference positive/negative inversion value of the ECG waveform is calculated for each sampling time. Then, the maximum value M1 of the time difference positive/negative inversion value in the time range A before the sampling time to be processed and the time difference positive/negative inversion value in the time range B after the sampling time to be processed is A value obtained by subtracting the time difference positive/negative inversion value of the sampling time is set as the first index value x of the sampling time to be processed. Further, the maximum value M2 of the time difference positive/negative inversion value in the time range C before the sampling time to be processed and the time difference positive/negative inversion value in the time range D after the sampling time to be processed is A value obtained by subtracting the time difference positive/negative inversion value of the sampling time is set as the second index value y of the sampling time to be processed.

FIG. 2 is a diagram plotting time-series data of the first index value x and the second index value y of the ECG waveform of FIG. It can be seen that the peak of the second index value y is thinner than the peak of the first index value x.

FIG. 3 is an enlarged view of the vicinity of the peak derived from the R wave in FIG. In FIG. 3, values are plotted at sampling intervals of 1 ms. It can be seen that the first index value x and the second index value y diverge from the apex of the peak. The reason for the difference between the first index value x and the second index value y is that when the time shifts from the center of the peak, the trailing edge of the time range C and the leading edge of the time range D become positive or negative in the time difference of the ECG waveform. This is because interference with the peak of the inversion value increases the maximum value M2 in the time ranges C and D, and the slope of the base of the peak of the second index value y becomes steeper.

On the other hand, FIG. 4 is a diagram showing an example of a pseudo waveform imitating an ECG waveform. The waveforms shown in FIG. 4 are obtained by shortening the time widths of the Q wave, R wave, and S wave by about half, and show an example that should not be detected as the R wave.

FIG. 5 shows plots similar to those in FIG. 3 for the waveforms in FIG. Compared to FIG. 3, the difference between the first index value x and the second index value y is small. The reason why the difference between the first index value x and the second index value y is small is that the width of the peak is narrow with respect to the interval between the time ranges C and D, and the degree of interference between the time ranges C and D and the peak waveform is small. This is because the increase in the maximum value M2 as in the case of FIG. 3 is unlikely to occur.

Comparing FIG. 3 and FIG. 5, in FIG. 3, the difference between the first index value x and the second index value y gradually widens as the distance from the peak increases. and the second index value y is small. In order to use such a relationship between the first index value x and the second index value y for R-wave detection, the second index value y at a certain time and the first index value x at a slightly distant time are It is good to pay attention to the relationship between

In FIG. 3, the first index value x=x3 exceeds the threshold TH at time t3. For example, the second index value y=y3 at the same time t3 and the first index value x=x1 at the time two samplings before, (x1−y3)>a (a is a predetermined positive value) If the above condition is satisfied, the time t3 may be set as the heartbeat time. That is, an upward peak that appears immediately after time t3 can be detected as a peak derived from the R wave.

Under the same conditions, in the case of FIG. 5, the first index value x=x3 is lower than the second index value y=y1, the time t3 is not detected as the heartbeat time, and the peak in FIG. It is not detected as a wave-derived peak.

It should be noted that the above condition may be satisfied in a phase where the first index value x and the second index value y pass the peak and then fall. In order to prevent the condition from being satisfied in the phase where the first index value x and the second index value y decrease in this way, the condition that the first index value x increases at time t3 (x3>x2) is further imposed, it is possible not to detect peaks such as those shown in FIG. 5 as peaks derived from R waves.

FIGS. 6A to 6C are diagrams plotting time-series data of the first index value x and the second index value y of the ECG waveform of FIG. 7A to 7C are diagrams plotting time-series data of the first index value x and the second index value y of the pseudo-waveform of FIG. 6A and 7A show the case where the interval between time ranges C and D is 20 ms, FIGS. 6B and 7B show the case where the interval between time ranges C and D is 24 ms, and FIGS. A case of 30 ms is shown. The interval between time ranges A and B is 50 ms.

As shown in FIG. 7A, when the interval between the time ranges C and D becomes shorter, the first index value x and the second index value y diverge even in the case of pseudo waveforms that should not be detected as R waves. On the other hand, as shown in FIG. 6C, when the interval between the time ranges C and D becomes longer, the first index value x and the second index value y approach each other even in the case of the ECG waveform. The interval between the time ranges C and D for distinguishing between the ECG waveform and the pseudo waveform is preferably set in the range of 20 ms to 30 ms.

[Example]
Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 8 is a block diagram showing the configuration of a heartbeat detection device according to an embodiment of the present invention. The heartbeat detection apparatus includes an electrocardiograph 1 for outputting an ECG waveform sampling data string, a storage unit 2 for storing an ECG waveform sampling data string and sampling time information, and sampling data from the ECG waveform sampling data string. A positive/negative inverted value calculation unit 3 that calculates the positive/negative inverted value of the time difference for each sampling time, the positive/negative inverted value of the time difference in the time range A before the sampling time to be processed, and the time after the sampling time to be processed. A maximum value detection unit 4 for detecting the maximum value M1 of the time difference positive/negative inversion values in the range B at each sampling time, and a first index obtained by subtracting the maximum value M1 from the time difference positive/negative inversion value at the sampling time to be processed. It has an index value calculation unit 5 that calculates the value x at each sampling time, and an index value storage unit 6 that receives the first index value x.

Furthermore, the heartbeat detection device detects the maximum value M2 of the time difference positive/negative inversion value in the time range C before the sampling time to be processed and the time difference positive/negative inversion value in the time range D after the sampling time to be processed. and an index value calculator 8 that calculates, at each sampling time, a second index value y obtained by subtracting the maximum value M2 from the time difference positive/negative inversion value at the sampling time to be processed. , the index value storage unit 9 that receives the second index value y, the second index value y at the first sampling time when the first index value x exceeds the threshold TH, and the first sampling time and a heartbeat time determination unit 10 that sets the first sampling time as the heartbeat time when the first index value x at the second sampling time having a predetermined relationship with the first index value x satisfies a predetermined condition.

FIG. 9 is a block diagram showing a configuration example of the maximum value detection units 4 and 7, the index

value calculation units

5 and 8, and the index

value storage units

6 and 9. As shown in FIG.
The maximum value detection unit 4 includes a FIFO buffer (First In, First Out) 40 to which the time difference positive/negative inversion value calculated by the positive/negative inversion value calculation unit 3 is input, and a FIFO buffer to which the output value of the FIFO buffer 40 is input. 41, a FIFO buffer 42 to which the output value of the FIFO buffer 41 is input, and the maximum value M1 of the positive/negative inverted value of the time difference stored in the FIFO buffer 40 and the inverted positive/negative value of the time difference stored in the FIFO buffer 42. and a detection processing unit 43 that performs detection at each sampling time.

The index value calculation unit 5 detects the maximum value detected by the maximum value detection unit 4 from the FIFO buffer 50 to which the time difference positive/negative inversion value calculated by the positive/negative value calculation unit 3 is input, and the output value of the FIFO buffer 50. and a subtraction processing unit 51 that calculates the first index value x obtained by subtracting M1 at each sampling time.
The index value storage unit 6 is composed of a FIFO buffer 60 to which the first index value x calculated by the subtraction processing unit 51 is input.

The maximum value detection unit 7 includes a FIFO buffer 70 to which the time difference positive/negative inversion value calculated by the positive/negative inversion value calculation unit 3 is input, a FIFO buffer 71 to which the output value of the FIFO buffer 70 is input, and a FIFO buffer 71. A FIFO buffer 72 to which an output value is input, and a detection that detects the maximum value M2 of the time difference positive/negative inversion value stored in the FIFO buffer 70 and the time difference positive/negative inversion value stored in the FIFO buffer 72 at each sampling time. and a processing unit 73 .

The index value calculation unit 8 detects the maximum value detected by the maximum value detection unit 7 from the FIFO buffer 80 to which the time difference positive/negative inversion value calculated by the positive/negative value calculation unit 3 is input, and the output value of the FIFO buffer 80. and a subtraction processing unit 81 that calculates a second index value y obtained by subtracting M2 at each sampling time.
The index value storage unit 9 is composed of a FIFO buffer 90 to which the second index value y calculated by the subtraction processing unit 81 is input.

The heartbeat detection method of this embodiment will be described below with reference to FIGS. 10 to 12. FIG. Here, the procedure from detecting one heartbeat to obtaining the heartbeat time will be described. Time-series data of heartbeat times is obtained by repeating such calculation of heartbeat times over the period of the ECG waveform data.

In this embodiment, D(i) is the data string obtained by sampling the ECG waveform. i (i=1, 2, . . . ) is a number given to data of one sampling. Needless to say, the larger the number i, the later the sampling time.

The electrocardiograph 1 measures an ECG waveform of a living body (human body) (not shown) and outputs a sampling data string D(i) of the ECG waveform. At this time, the electrocardiograph 1 adds sampling time information to each sampling data and outputs the data. Since the specific method of measuring the ECG waveform is a well-known technique, detailed description thereof will be omitted.
The storage unit 2 stores the sampling data sequence D(i) of the ECG waveform output from the electrocardiograph 1 and the information on the sampling time.

In order to calculate the time difference positive/negative inversion value Y(i) of the sampling data D(i), the positive/negative inversion value calculation unit 3 calculates the data D(i+1) after one sampling of the sampling data D(i) and the data D(i+1) one sampling before. Data D(i-1) are obtained from the storage unit 2 (step S1 in FIG. 10). Then, the positive/negative inversion value calculator 3 calculates the time difference positive/negative inversion value Y(i) of the sampling data D(i) for each sampling time as shown in the following equation (step S2 in FIG. 10).
Y(i)=-{D(i+1)-D(i-1)} (1)

A first index value x is calculated by the maximum value detection unit 4 and the index value calculation unit 5 (step S3 in FIG. 10). Specifically, the positive/negative inversion value calculator 3 inputs the calculated time difference positive/negative inversion value Y(i) to the FIFO buffer 50 at each sampling time (step S100 in FIG. 11). The input value is held in the FIFO buffer 50, and after a time corresponding to the size of the FIFO buffer 50 (the delay time from when the positive/negative time difference value is input to the FIFO buffer 50 until it is output), It will be used for subtraction processing.

In addition, the positive/negative inversion value calculation unit 3 inputs the calculated time difference positive/negative inversion value Y(i) to the FIFO buffer 40 at each sampling time (step S101 in FIG. 11). The output of FIFO buffer 40 is input to FIFO buffer 41 (step S102 in FIG. 11), and the output of FIFO buffer 41 is input to FIFO buffer 42 (step S103 in FIG. 11). The FIFO buffers 40 to 42 are for obtaining the maximum value of the time difference positive/negative inversion values in the time ranges A and B. FIG.

The time interval L3 corresponding to the length of the FIFO buffer 41 (the delay time from when the time difference positive/negative inversion value is input to the FIFO buffer 41 until it is output) is the width of the peak derived from the R wave (approximately 10 ms). ), preferably about 50 ms. This interval L3 becomes the interval between the time ranges A and B. Also, a time interval L2 corresponding to the length of the FIFO buffer 40 (a delay time from when the time difference positive/negative inversion value is input to the FIFO buffer 40 until it is output), and a time interval corresponding to the length of the FIFO buffer 42 L4 (the delay time from when the time difference positive/negative inversion value is input to the FIFO buffer 42 until it is output, L2=L4) is appropriately about 100 ms.

Also, the time interval L1 corresponding to the length of the FIFO buffer 50 should be L1=L2+L3/2. Therefore, in the above numerical example, L1 is 125 ms. By setting L1=L2+L3/2 and L2=L4, the time (sampling time to be processed) of the output value b of the FIFO buffer 50 is in the range of -(L2+L3/2) to -(L3/2). The maximum value M1 can be obtained for the range from (L3/2) to (L2+L3/2), and the maximum value M1 can be subtracted from the output value b.

The detection processing unit 43 detects the maximum value M1 of the time difference positive/negative inversion value stored in the FIFO buffer 40 and the time difference positive/negative inversion value stored in the FIFO buffer 42 at each sampling time (step S104 in FIG. 11). .
The subtraction processing unit 51 calculates a first index value x(i)=b−M1 by subtracting the maximum value M1 from the output value b of the FIFO buffer 50 at each sampling time (step S105 in FIG. 11).

Thus, the processing of step S3 is performed by the maximum value detection unit 4 and the index value calculation unit 5, and at the same time the second index value y is calculated by the maximum value detection unit 7 and the index value calculation unit 8 (see FIG. 10). step S4).

Specifically, the positive/negative inversion value calculator 3 inputs the calculated time difference positive/negative inversion value Y(i) to the FIFO buffer 80 at each sampling time (step S200 in FIG. 12). Further, the positive/negative inversion value calculator 3 inputs the calculated time difference positive/negative inversion value Y(i) to the FIFO buffer 70 at each sampling time (step S201 in FIG. 12). The output of FIFO buffer 70 is input to FIFO buffer 71 (step S202 in FIG. 12), and the output of FIFO buffer 71 is input to FIFO buffer 72 (step S203 in FIG. 12). The FIFO buffers 70-72 are for obtaining the maximum value of the time difference positive/negative inversion values in the time ranges C and D. FIG.

A time interval L7 corresponding to the length of the FIFO buffer 71 (a delay time from when the time difference positive/negative inversion value is input to the FIFO buffer 71 until it is output) is 24 ms in the above example. This interval L7 becomes the interval between the time ranges C and D. Also, a time interval L6 corresponding to the length of the FIFO buffer 70 (a delay time from when the time difference positive/negative inversion value is input to the FIFO buffer 70 until it is output), and a time interval corresponding to the length of the FIFO buffer 72 L8 (the delay time from when the time difference positive/negative inversion value is input to the FIFO buffer 72 until it is output, L6=L8) is appropriately about 113 ms. As is clear from the above description, L2+L3+L4=L6+L7+L8.

Also, the time interval L5 corresponding to the length of the FIFO buffer 80 should be L5=L6+L7/2. Therefore, in the above numerical example, L5 is 125 ms. This value is the same as the FIFO buffer 50 length L1. By setting L5=L6+L7/2 and L6=L8, the time (sampling time to be processed) of the output value c of the FIFO buffer 80 is in the range of -(L6+L7/2) to -(L7/2). The maximum value M2 can be obtained for the range from (L7/2) to (L6+L7/2), and the maximum value M2 can be subtracted from the output value c.

The detection processing unit 73 detects the maximum value M2 of the time difference positive/negative inversion value stored in the FIFO buffer 70 and the time difference positive/negative inversion value stored in the FIFO buffer 72 at each sampling time (step S204 in FIG. 12). .
The subtraction processing unit 81 calculates a second index value y(i)=c−M2 by subtracting the maximum value M2 from the output value c of the FIFO buffer 80 at each sampling time (step S205 in FIG. 12). Thus, the processing of step S4 is performed by the maximum value detection unit 7 and the index value calculation unit 8. FIG.

The subtraction processing unit 51 inputs the calculated first index value x(i) to the FIFO buffer 60 at each sampling time (step S5 in FIG. 10). At the same time, the subtraction processing unit 81 inputs the calculated second index value y(i) to the FIFO buffer 90 at each sampling time (step S6 in FIG. 10).

The heartbeat time determination unit 10 acquires the first index value x(i)=x3 and the second index value y(i)=y3 stored in the first positions of the FIFO buffers 60 and 90 . In addition, the heartbeat time determining unit 10 stores the first index value x (1 ms) past the first index value x(i)=x3 stored in the second position of the FIFO buffer 60 by one sampling (1 ms). Get (i−1)=x2. Further, the heartbeat time determining unit 10 determines the first index value x (2 ms) past the first index value x(i)=x3 stored in the third position of the FIFO buffer 60 by two samplings (2 ms). Get (i-2)=x1.

The heartbeat time determining unit 10 determines that the first index value x(i)=x3 exceeds the threshold TH (YES in step S7 in FIG. 10) and the first index value x(i−1)=x2 one sampling before. when the first index value x(i)=x3 is increasing (YES in step S8 in FIG. 10), the second index value at the same time as the first index value x(i)=x3 If y(i)=y3 and the first index value x(i−2)=x1 two samplings before satisfy the condition that (x1−y3)>a (YES in step S9 in FIG. 10), the first The sampling time of the index value x(i)=x3 is set as the heartbeat time (step S10 in FIG. 10).

As described above, the time difference positive/negative inversion value Y(i) is calculated from the sampling data D(i), and the first index value x(i) is calculated from the time difference positive/negative inversion value Y(i). Therefore, the sampling time of the first index value x(i) is the sampling time of the time difference positive/negative inversion value Y(i) (the sampling time of the data D(i)). Is possible.
In this way, by repeatedly executing the processing of steps S1 to S10 for each sampling period, time-series data of heartbeat times can be obtained.

In this embodiment, the second index value y(i)=y3 at the first sampling time (t3 in the example of FIG. 3) when the first index value x(i)=x3 exceeds the threshold TH , a first index value x(i−2)=x1 at a second sampling time (t1 in the example of FIG. 3) having a predetermined relationship with the first sampling time satisfies the condition. However, the second sampling time may be a time that is a predetermined time past the first sampling time, and does not have to be two samplings (2 ms) earlier.

Storage unit 2, positive/negative inversion value calculation unit 3, maximum value detection units 4 and 7, index

value calculation units

5 and 8, index

value storage units

6 and 9, and heartbeat time determination unit 10 of the heartbeat detection device described in this embodiment can be realized by a computer having a CPU (Central Processing Unit), a storage device and an interface, and a program controlling these hardware resources. A configuration example of this computer is shown in FIG.

The computer includes a CPU 100, a storage device 101, and an interface device (I/F) 102. The I/F 102 is connected with the electrocardiograph 1 and the like. A program for implementing the heartbeat detection method of the present invention is stored in the storage device 101 . The CPU 100 executes the processing described in this embodiment according to the programs stored in the storage device 101 .

The present invention can be applied to technology for detecting the heartbeat of a living body.

DESCRIPTION OF SYMBOLS 1... electrocardiograph, 2... storage part, 3... positive/negative value calculation part, 4, 7... maximum value detection part, 5, 8... index value calculation part, 6, 9... index value storage part, 10... heartbeat time Determination section 40 to 42, 50, 60, 70 to 72, 80, 90... FIFO buffer, 43, 73... Detection processing section, 51, 81... Subtraction processing section.

Claims

a first step of calculating, at each sampling time, a positive/negative inversion value of a time difference of sampling data from a sampling data string of an electrocardiogram waveform of a living body;
sampling a first maximum value of the positive/negative inverted value in a first time range before the sampling time to be processed and the positive/negative inverted value in a second time range after the sampling time to be processed; a second step of detecting for each time;
a second maximum value of the positive/negative inverted value in a third time range before the sampling time to be processed and the positive/negative inverted value in a fourth time range after the sampling time to be processed; a third step of detecting at each sampling time;
a fourth step of calculating a subtraction value obtained by subtracting the first maximum value from the positive/negative inverted value of the sampling time to be processed as a first index value for each sampling time;
a fifth step of calculating a subtraction value obtained by subtracting the second maximum value from the positive/negative inverted value of the sampling time to be processed as a second index value for each sampling time;
The second index value at a first sampling time when the first index value exceeds a predetermined threshold, and the first index at a second sampling time having a predetermined relationship with the first sampling time. and a sixth step of setting the first sampling time as a heartbeat time when the value satisfies a predetermined condition.
The heart rate detection method of claim 1, wherein
The timing of the leading edge of the third time range with respect to the sampling time to be processed is the same as the timing of the leading edge of the first time range,
The timing of the trailing edge of the fourth time range with respect to the sampling time to be processed is the same as the timing of the trailing edge of the second time range,
The heartbeat detection method, wherein the interval between the third time range and the fourth time range is shorter than the interval between the first time range and the second time range.
The heartbeat detection method according to claim 1 or 2,
The sixth step increases the first index value at the first sampling time with respect to the first index value before the first sampling time, and when the second index value at the second sampling time and the first index value at the second sampling time satisfy a predetermined condition, the first sampling time is set to the heartbeat time. Heart rate detection method.
The heartbeat detection method according to any one of claims 1 to 3,
A heartbeat detection method, wherein the interval between the third time range and the fourth time range is in the range of 20 ms to 30 ms.
a positive/negative inversion value calculation unit configured to calculate, at each sampling time, a positive/negative inversion value of the time difference of the sampling data from the sampling data string of the electrocardiogram waveform of the living body;
sampling a first maximum value of the positive/negative inverted value in a first time range before the sampling time to be processed and the positive/negative inverted value in a second time range after the sampling time to be processed; a first maximum value detection unit configured to detect for each time;
a second maximum value of the positive/negative inverted value in a third time range before the sampling time to be processed and the positive/negative inverted value in a fourth time range after the sampling time to be processed; a second maximum value detection unit configured to detect at each sampling time;
a first index value calculation unit configured to calculate a subtraction value obtained by subtracting the first maximum value from the positive/negative inverted value of the sampling time to be processed as a first index value at each sampling time;
a second index value calculation unit configured to calculate a subtraction value obtained by subtracting the second maximum value from the positive/negative inverted value of the sampling time to be processed as a second index value at each sampling time;
The second index value at a first sampling time when the first index value exceeds a predetermined threshold, and the first index at a second sampling time having a predetermined relationship with the first sampling time. a heartbeat time determination unit configured to determine the first sampling time as a heartbeat time when the value of the heartbeat satisfies a predetermined condition.
The heartbeat detection device of claim 5, wherein
The timing of the leading edge of the third time range with respect to the sampling time to be processed is the same as the timing of the leading edge of the first time range,
The timing of the trailing edge of the fourth time range with respect to the sampling time to be processed is the same as the timing of the trailing edge of the second time range,
The heartbeat detection device, wherein the interval between the third time range and the fourth time range is shorter than the interval between the first time range and the second time range.
The heartbeat detection device according to claim 5 or 6,
The heartbeat time determining unit increases the first index value at the first sampling time with respect to the first index value before the first sampling time, and and the first index value at the second sampling time satisfies a predetermined condition, the first sampling time is set as the heartbeat time. .
A heartbeat detection device according to any one of claims 5 to 7,
The heartbeat detection device, wherein the interval between the third time range and the fourth time range is in the range of 20 ms to 30 ms.