KR101245445B1 - R-peak detection method in ECG Signal and apparatus thereof - Google Patents

Abstract

The present invention relates to a method and apparatus for R-peak detection in an ECG signal. According to the present invention, there is provided a method of analyzing an ECG signal, comprising the steps of: preprocessing an ECG signal; setting a first threshold value in the ECG signal using the largest peak value of signals within a first time range; Peak value; detecting a first to an N-th R-peak value corresponding to an upper peak value within the detection range; Peak value in an ECG signal comprising the step of setting a second threshold using an average of R-peak values of the N-th peak and an R-peak value of N + 1 using the second threshold, A peak detection method and apparatus therefor are provided.
According to the method and apparatus for detecting an R-peak in an ECG signal according to the present invention, noise is effectively removed from an original signal using an SVD filter, and then an R-peak value The detection efficiency can be increased. In addition, there is an advantage in that the detection error rate can be reduced because the R-peak value that has not been detected previously can be reliably detected again by performing a separate search back process.

Description

(R-peak detection method in ECG signal and apparatus thereof)

The present invention relates to a method and apparatus for detecting an R-peak in an ECG signal, and more particularly to a method for detecting an R-peak in an ECG signal for detecting an R-peak in an ECG (Electrocardiogram) .

The ECG signal is formed by a continuous circulation of electrical activity emanating from the heart muscle. The time and type of occurrence of this ECG signal reflect the cardiac condition of the patient. Of the components of the ECG signal, QRS is the most prominent signal. Among them, the interval between the R-peak values (RRI) can be used to predict the cardiac cycle, and the accurate prediction of the cardiac cycle results from the detection accuracy of the R-peak value. Therefore, it is necessary to develop a reliable QRS detection algorithm.

Many studies have been conducted to accurately detect QRS. In a typical study, a digital filter is mainly used for preprocessing an ECG signal, which causes a distortion of the original signal. The distorted signal modifies the shape of the R-peak so that it can be recognized as a noise at the time of detection of the R-peak, and as a result, it becomes difficult to accurately detect the R-peak. In addition, such an incorrect detection result leads to a loss of data, thereby deteriorating the detection accuracy. Therefore, the noise can be reduced without distortion and a method for more accurately detecting the R-peak is needed.

An object of the present invention is to provide a method and apparatus for detecting an R-peak in an ECG signal that can increase the accuracy of detecting an R-peak value from an ECG original signal.

The present invention relates to a method for detecting an ECG signal, comprising the steps of: preprocessing an ECG signal; setting a first threshold value in the ECG signal using the largest peak value of signals within a first time range; - detecting a first to an N-th R-peak value corresponding to an upper peak value within the detection range; and setting a detection range of the latest N-1 Peak value in an ECG signal comprising setting a second threshold using an average of R-peak values, and detecting an R-peak value of N + 1 using the second threshold value. Detection method.

Here, in the step of preprocessing the ECG signal, the ECG signal is an A matrix of 16 × 16 form, where A = UΣV ^T , U is a left singular matrix, V is a right singular matrix, A singular value decomposition filter is used to remove noise from the ECG signal, and the SVD filter can be defined by the following equation.

Here,? _I denotes a diagonal matrix component value of the i-th column in the A matrix.

The step of setting the detection range of the R-peak value using the first threshold value may include setting a first ECG value that is greater than the first threshold value among the signals within the first time range as a starting point, Setting an ECG value smaller than the first threshold value as an end point from the starting point as an end point; and if an ECG value greater than the first threshold value exists within a second time range from the end point, And updating the end point by setting a first ECG value smaller than the first threshold value as a new end point from a larger ECG value.

Also, the first threshold value may be 25 to 35% of the largest peak value of the signals within the first time range, the first time range may be within 1500 ms, and the second time range may be within 360 ms .

Also, the setting of the second threshold value may set the value of 25 to 35% of the average of the latest N-1 R-peak values to the second threshold value.

The detecting of the (N + 1) th R-peak value may further include detecting a peak value exceeding the second threshold value after the N-th R-peak value as the (N + 1) can do.

If the Nth RRI that is a time interval between the Nth and (N + 1) th R-Peak values is 140 to 150% or more of the small RRI of the previous N-1 RRI and the N-2 RRI, The R-peak value can be re-detected within the Nth RRI.

At this time, the step of re-detecting the R-peak value may include the steps of setting a third threshold value using the N-th R-peak value, and determining the third threshold value within a second time range from the third threshold value. If a peak value larger than the threshold value exists, a peak value exceeding the third threshold value within a range between a first end point smaller than the third threshold value and the (N + 1) Peak value as the R-peak value.

A first threshold value setting unit configured to set a first threshold value using the largest peak value of signals within a first time range in the ECG signal; A detection range setting unit for setting a detection range of the R-peak value using the first threshold value; and a first detection unit for detecting first to Nth R-peak values corresponding to upper peak values within the detection range, A second threshold value setting unit configured to set a second threshold value using an average of the latest N-1 R-peak values among the R-peak values, and a second threshold value setting unit configured to set the And a second detection section for detecting an R-peak value.

According to the method and apparatus for detecting an R-peak in an ECG signal according to the present invention, noise is effectively removed from an original signal using an SVD filter, and then an R-peak value The detection efficiency can be increased. In addition, there is an advantage in that the detection error rate can be reduced because the R-peak value that has not been detected previously can be reliably detected again by performing a separate search back process.

1 shows a general ECG signal form.
2 is a configuration diagram of an R-peak detection apparatus in an ECG signal according to an embodiment of the present invention.
FIG. 3 is a flowchart of an R-peak detection method in the ECG signal using FIG.
FIG. 4 shows an example of an ECG signal to which the preprocessing step of FIG. 1 is applied.
FIG. 5 is an explanatory diagram for R-peak value detection according to FIG. 2. FIG.
FIG. 6 is an explanatory diagram for re-detection of the R-peak value in FIG. 2. FIG.
FIG. 7 shows a result of preprocessing the MIT-BIH data 100 with an SVD filter according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention.

The present invention provides a method and apparatus for detecting an R-peak value from an ECG signal. 1 shows a general ECG signal form. The R point in the ECG signal is the R-peak point representing the contraction of the atrium, and QRST is derived from the contraction of the ventricle.

The RRI interval between R and R is R-R interval. The detection of such RRI can predict the cardiac cycle, and the accurate cardiac cycle prediction is due to the detection accuracy of the R-peak value. Therefore, the present invention provides a method for increasing the detection accuracy of the R-peak value in the ECG signal. To this end, the ECG signal is pre-processed to remove noise components, and R-peak values are detected, and R-peak values that are not detected in the process are re-detected by a search back method to improve the reliability of detection.

2 is a configuration diagram of an R-peak detection apparatus in an ECG signal according to an embodiment of the present invention. The apparatus 100 includes a preprocessing unit 110, a first threshold value setting unit 120, a detection range setting unit 130, a first detection unit 140, a second threshold value setting unit 150, (160), and re-detection unit (170).

FIG. 3 is a flowchart of an R-peak detection method in the ECG signal using FIG. Hereinafter, the R-peak detection method will be described in detail with reference to FIG. 1 to FIG.

First, the preprocessing unit 110 preprocesses the ECG signal (S310). At this time, a singular value decomposition (SVD) filter is used to represent the signal in the time / frequency domain.

The ECG signal may be expressed as A = UΣV ^T , which is an A matrix of 16 × 16 form. The SVD filter decomposes the ECG signal into a left singular matrix (U), a diagonal matrix (?), And a right singular matrix (V) to remove the noise value.

Here, Σ is a matrix in which the components other than the diagonal components are 0 and the diagonal components are arranged in descending order. A method of removing noise from the ECG signal using the SVD filter is described in Equation (1).

Here,? _I denotes a diagonal matrix component value of the i-th column. That is, when Equation (1) is applied to the diagonal matrix (?), The components of columns 1 to 2 of the diagonal matrix are used as they are, and the components of the columns 3 to 5 are applied with an exponential function to improve the quality of the signal. Then, the ECG signal components of the 6th to 16th columns are removed by applying a value of 0 as noise.

That is, α _i is obtained by multiplying the i-th component of the noise signal by a specific value. According to Equation 1, the components in columns 1 and 2 are used as they are, while the components in columns 3 and 5 are used to improve signal quality and remove other components .

Then, the right-handed matrix uses only up to two columns and removes the remaining components.

Thereafter, the left singular matrix, the diagonal matrix from which unnecessary elements have been removed, and the right singular matrix using only up to two columns are reunited to represent the original signal, and the signal can be recovered as a noise-canceled signal.

FIG. 4 shows an example of an ECG signal to which the preprocessing step of FIG. 1 is applied. This confirms that the noise is effectively removed as a result of applying the SVD filter to the ECG signal to which the noise is added.

After the SVD filter is applied, a noise of 5 Hz or less is removed by using a first-order Butterworth HPF to remove the baseline variation. The ECG signal has important information between 5 and 15 Hz, so that even if the signal value below 5 Hz is ignored, it does not affect the detection of the R-peak.

After removing the baseline variation, a 5-point moving average is applied to reduce residual noise and soften the signal. Finally, squaring the signal to emphasize the signal.

After the above-described preprocessing, R-peak values are detected from the ECG signal. Please refer to the following for the detailed procedure.

FIG. 5 is an explanatory diagram for R-peak value detection according to FIG. 2. FIG. First, the first threshold value setting unit 120 sets a first threshold value THV1 (THV1) using the largest peak value of the signals within the first time range (ex, 1500 ms) in the ECG signal (S320).

For example, the first threshold value THV1 uses 25 to 35% (ex, 30%) of the largest peak value among the signals within the range of 1500 ms in the ECG signal. If the largest peak value is 1.5, the first threshold value THV1 is 0.45 corresponding to 30% of 1.5.

The process of setting the first threshold value may be expressed as Equation (2).

Then, the detection range setting unit 130 sets the detection range of the R-peak value using the first threshold value THV1 (S330). Step S330 is as follows.

First, the first time range (ex, 1500ms) to the first threshold larger than the first ECG values (THV1) the starting point of the signal is within; is set to be _{(S S s [n])} . In FIG. 5, for convenience of description, the starting point S is indicated as a point overlapping the first threshold value.

Next, a small first ECG value than the first threshold, the endpoint from the starting point (S); and sets as _{(E1 S e [n])} .

The calculation process of the starting point and the ending point can be summarized in Equation (3).

If an ECG value (S 'point) larger than the first threshold value (THV1) exists within the second time range (ex, 360 ms) from the end point (E1) (E; E2) smaller than the first threshold value (THV1) from the value (S 'point) to the new end point to update the end point. Accordingly, the end point E is updated from E1 to E2 while maintaining the start point S as it is. If this process is repeated for an updated endpoint within 1500ms, the final endpoint is updated to E6 and determined.

For the sake of convenience, the above description omits the end point detection process for the T-wave (or P-wave) portion. Therefore, if T wave is larger than the threshold value THV1 as shown in FIG. 5, a new end point after E; E1 can be detected as the end portion Te of T wave. In this case, the new end point E2, which appears after Te again, is updated to the end point. Repeat this process within 1500 ms and the final end point is determined as E6.

According to the above-described step S330, the detection range of the R-peak value is determined from the starting point S to the last updated end point E (E6). The updating process of the end point can be summarized by the following equation (4).

Since the R-peak value detection range setting step S330 is performed within the range of 1500 ms, the detection range of the finally determined R-peak value is substantially smaller than the 1500 ms range.

Thereafter, the first detector 140 detects first through Nth (ex, N = 6) R-peak values corresponding to upper peak values within the detection range at step S340.

5, there are six R-peak values corresponding to the upper peak value within the detection range. Hereinafter, the case where N = 6 will be described as an embodiment. Needless to say, the present invention is not necessarily limited thereto.

Next, the second threshold value setting unit 150 sets the average of the latest N-1, that is, the latest five R-peak values (R2 to R6) out of the sequentially detected six R-peak values R1 to R6 The second threshold value THV2 is set (S350).

At this time, a value of 25 to 35% (ex, 30%) with respect to the average of the R-peak values of the last five is set as the second threshold value THV2. That is, when the average of the peak values of R2 to R6 is 1.3, the second threshold value THV2 is 0.39 corresponding to 30% of 1.3.

The process of setting the second threshold value may be expressed by the following equation (5).

Here, Rpeak _AVG represents the average of the five R-peak values (R2 through R6) in the past. As described above, the next R-peak value can be detected by setting 30% of the average value of the five R-peak values to a new threshold value (second threshold value).

That is, the second detector 160 detects the (N + 1) th R-peak value, that is, the seventh R-peak value, using the second threshold THV2 (S360). More specifically, a peak value exceeding the second threshold value THV after the sixth R-peak value R6 is detected as the seventh R-peak value R7.

The process of step S360 may be applied to each subsequent R-peak value detection. That is, the threshold value can be continuously updated every time the next R-peak value is detected, and the corresponding R-peak signal can be detected. For example, at the time of detecting the eighth R-peak value, a 30% value of the average of the latest R 5 to R 7 peak values may be used as a new threshold value for detecting the eighth R-peak value.

However, an R-peak value that has not been detected during the detection process as described above may exist. Therefore, referring to FIG. 6, a search back method for detecting an undetected R-peak value is described. FIG. 6 is an explanatory diagram for re-detection of the R-peak value in FIG. 2. FIG.

That is, the sixth RRI, which is the RRI between the sixth R-peak value R6 and the seventh R-peak value R7, is smaller than the first RRI between the first RRI and the fourth RRI, Peak value is not detected between R6 and R7, and the re-detection unit 170 re-detects the R-peak value within the sixth RRI (S370).

The step S370 will be described in detail as follows. First, the third threshold value THV3 is set using the sixth R-peak value R6. At this time, the third threshold value may be set to a value of 40% of the sixth R-peak value R6. This 40% value is intended to prevent noise other than T-wave from being included in the detection.

If there is a peak value larger than the third threshold value THV3 within the second time range (ex, 360 ms) from the third threshold value THV3, it is regarded as a T-wave rather than an R-peak value , And then, after the second time range, a peak (THV3) exceeding the third threshold value (THV3) within the range between the first peak value (R7) and the first end point (E) And further detects the largest peak value among the values as the R-peak value (R). That is, when the signal value larger than the third threshold value THV3 is at the point of 360 ms or longer from the previous R-peak value, the largest signal value within the range (less than the seventh peak value) is detected as the R-peak value will be.

The signaling period varies from person to person and is not constantly maintained by diseases such as rapid heart rate or PVC. If the peak value is not detected within 150% of the smallest RRI among the two RRIs, the peak value is newly detected within the interval. This is because the peak value can not be detected when the RRI interval and the interval that have not been detected recently are larger than usual due to bass and PVC. According to the re-detection process of step S370, the error rate for recognizing the R-peak value as a T-wave or the P-wave can be reduced, and the R-peak value can be detected, thereby reducing the overall detection error rate.

Experimental results using the embodiment of the present invention will be described below. The equipment used was Intel Pentium Dual CPU E2140 (1.6GHz) with 2GB of RAM and running on Windows XP. The MIT-BIH database was used to evaluate the embodiments of the present invention and programmed with MATLAB 7.10. The MIT-BIH data consists of 48 data sampled at 360 Hz, measuring the patient's ECG signal for about 30 minutes. In this experiment, we apply resampling to the algorithm at 200Hz.

FIG. 7 shows a result of preprocessing the MIT-BIH data 100 with an SVD filter according to an embodiment of the present invention. In the detail graph in Fig. 7, since the scales on the horizontal axis are different from each other, reference is made to the part of the circle which is a common part.

FIG. 7A shows an ECG original signal, and FIG. 7B shows a result of removing the first-order noise by applying an SVD filter to FIG. 7A. (c) is the result of eliminating the baseline variation through the Butterworth HPF, (d) is the result of smoothing the residual noise by applying a 5-point moving average to the signal from the SVD filter , (e) is a result of emphasizing the components of the signal through the squaring process. When the SVD filter time was measured, it took about 3.5 ms to filter 256 data including about two cycles in one cycle based on 200 Hz sampling. (f) is a result of detecting the R-peak value from the data of (e).

Sensitivity (SEN), positive predictability (PRE), and detection error (DE) were used for the reliability evaluation of the method according to the embodiment of the present invention. sensitivity is the ability of the algorithm to detect true bits, positive predictivity is true, the ability of the algorithm to distinguish flase bits, and detection error is the accuracy of the algorithm.

Equation (6) is used for the calculation of each ability.

Here, TP means a correctly detected bit, FN means a bit which can not be detected, and FP means a bit which is erroneously detected.

A total of 48 data were tested and the results are summarized in Table 1.

Total beats FP FN FP + FN SEN (%) PRE (%) DE (%) 100 2273 0 0 0 100.00 100.00 0.00 101 1865 2 One 3 99.95 99.89 0.16 102 2187 0 0 0 100.00 100.00 0.00 103 2084 0 2 2 99.90 100.00 0.10 104 2229 8 38 46 98.30 99.64 2.06 105 2572 67 4 71 98.84 97.40 2.76 106 2027 13 3 16 99.85 99.36 0.79 107 2137 38 17 55 99.20 98.22 2.57 108 1774 42 7 49 99.61 97.63 2.76 109 2532 0 0 0 100.00 100.00 0.00 111 2124 0 2 2 99.91 100.00 0.09 112 2539 0 0 0 100.00 100.00 0.00 113 1795 0 0 0 100.00 100.00 0.00 114 1879 0 3 3 99.84 100.00 0.16 115 1953 0 0 0 100.00 100.00 0.00 116 2412 0 One One 99.96 100.00 0.04 117 1535 0 0 0 100.00 100.00 0.00 118 2288 0 0 0 100.00 100.00 0.00 119 1987 0 0 0 100.00 100.00 0.00 121 1863 3 0 3 100.00 99.84 0.16 122 2476 0 0 0 100.00 100.00 0.00 123 1518 0 3 3 99.80 100.00 0.20 124 1619 0 3 3 99.81 100.00 0.19 200 2601 One 0 One 100.00 99.96 0.04 201 2000 45 14 59 99.30 97.75 2.95 202 2136 0 55 55 97.43 100.00 2.57 203 2980 179 100 279 96.64 93.99 9.36 205 2656 2 18 20 99.32 99.92 0.75 207 1848 11 7 18 99.62 99.40 0.97 208 2955 2 16 18 99.46 99.93 0.61 209 3005 0 7 7 99.77 100.00 0.23 210 2650 48 15 63 99.43 98.19 2.38 212 2748 0 0 0 100.00 100.00 0.00 213 3251 0 4 4 99.88 100.00 0.12 214 2262 One 7 8 99.69 99.96 0.35 215 3363 0 5 5 99.85 100.00 0.15 217 2208 54 5 59 99.77 97.55 2.67 219 2154 0 2 2 99.91 100.00 0.09 220 2047 0 0 0 100.00 100.00 0.00 221 2427 6 0 6 100.00 99.75 0.25 222 2483 0 4 4 99.84 100.00 0.16 223 2605 4 25 29 99.04 99.85 1.11 228 2053 47 29 76 98.59 97.71 3.70 230 2256 0 0 0 100.00 100.00 0.00 231 1573 0 2 2 99.87 100.00 0.13 232 1780 5 0 5 100.00 99.72 0.28 233 3079 0 6 6 99.81 100.00 0.19 234 2753 One 0 One 100.00 99.96 0.04 48 patients 109541 579 405 984 99.63 99.47 0.89

Referring to Table 1, 405 FNs, 579 FPs, SEN = 99.63%, PRE = 99.47%, and DE = 0.89% were obtained. Using the above method, it is possible to effectively detect the R-peak value by improving the quality of the signal in data in which a lot of noise exists.

As a result, according to the present invention, detection efficiency can be improved by effectively removing noise from an original signal using an SVD filter and then detecting an R-peak value using a different threshold within a predetermined time range. Also, since the R-peak value that has not been detected previously can be reliably detected again by performing a separate search back process, there is an advantage that the detection error rate can be reduced overall.

While the present invention has been described with reference to exemplary embodiments, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims. Accordingly, the true scope of the present invention should be determined by the technical idea of the appended claims.

100: R-peak detecting device 110:
120: first threshold value setting unit 130: detection range setting unit
140: first detection unit 150: second threshold value setting unit
160: second detection unit 170:

Claims (14)

Pre-processing the ECG signal;
Setting a first threshold value in the ECG signal using a largest peak value of signals within a first time range;
Setting a detection range of the R-peak value using the first threshold value;
Detecting first to N-th R-peak values corresponding to upper peak values within the detection range;
Setting a second threshold value using an average of the latest N-1 R-peak values among the R-peak values; And
And detecting the (N + 1) th R-peak value using the second threshold value,
Wherein the step of setting the detection range of the R-peak value using the first threshold value comprises:
Setting a first ECG value that is greater than the first threshold value among the signals within the first time range as a start point;
Setting a first ECG value less than the first threshold value as an end point from the starting point; And
Setting a first ECG value smaller than the first threshold value as a new end point from an ECG value larger than the first threshold value when an ECG value larger than the first threshold value exists within a second time range from the end point, Gt; R-peak < / RTI > in an ECG signal. The method according to claim 1,
In the step of pre-processing the ECG signal,
The ECG signal is an A matrix of 16x16 form, where A = U? V ^T , U is a left singular matrix, V is a right singular matrix,? Is a diagonal matrix, Lt; RTI ID = 0.0 >
Wherein the SVD filter is an R-peak detection method in an ECG signal defined by the following equation:

Here,? _I denotes a diagonal matrix component value of the i-th column in the A matrix. delete The method according to claim 1,
Wherein the first threshold value is 25 to 35% of the largest peak value of the signals within the first time range,
The first time range is within 1500 ms,
Wherein the second time range is within 360 ms. The method according to claim 1,
Wherein the step of setting the second threshold value comprises:
And setting a value of 25 to 35% with respect to an average of the latest N-1 R-peak values to the second threshold value. The method according to claim 1 or 5,
The step of detecting the (N + 1) th R-peak value comprises:
Peak value in the ECG signal is detected as the (N + 1) th R-peak value after the N-th R-peak value exceeds the second threshold value. The method according to claim 1,
The Nth RRI, which is the time interval between the Nth and (N + 1) th R-Peak values, is 140 to 150% or more of the small RRI of the previous N-1 RRI and the N-2 RRI, Further comprising redetecting the R-peak value within the RRI. The method of claim 7,
The step of redetecting the R-peak value comprises:
Setting a third threshold value using the Nth R-peak value; And
If the peak value is larger than the third threshold value within the second time range from the third threshold value, the first N + 1 peak value from the first end point that is smaller than the third threshold value after the second time range Peak value of the peak value exceeding the third threshold within a range between the first peak value and the second peak value is detected as an R-peak value. A preprocessor for preprocessing the ECG signal;
A first threshold value setting unit configured to set a first threshold value using the largest peak value of the signals within the first time range in the ECG signal;
A detection range setting unit configured to set a detection range of the R-peak value using the first threshold value;
A first detector for detecting first to Nth R-peak values corresponding to upper peak values within the detection range;
A second threshold value setting unit configured to set a second threshold value using an average of the latest N-1 R-peak values among the R-peak values; And
And a second detection unit for detecting an (N + 1) th R-peak value using the second threshold value,
Wherein the detection range setting unit comprises:
Setting a first ECG value that is greater than the first threshold value among the signals within the first time range as a start point,
Setting a first ECG value smaller than the first threshold value as an end point from the start point,
Setting a first ECG value smaller than the first threshold value as a new end point from an ECG value larger than the first threshold value when an ECG value larger than the first threshold value exists within a second time range from the end point, Gt; R-peak < / RTI > The method of claim 9,
The preprocessor removes noise from the ECG signal using a SVD (Singular Value Decomposition) filter,
The ECG signal is an A matrix of 16 × 16 form, where A = UΣV ^T , U is a left singular matrix, V is a right singular matrix, Σ is a diagonal matrix,
Wherein the SVD filter is an R-peak detection device in an ECG signal defined by the following equation:

Here,? _I denotes a diagonal matrix component value of the i-th column in the A matrix. delete The method of claim 9,
Wherein the second detection unit comprises:
And a peak value exceeding the second threshold value after the Nth R-peak value is detected as the (N + 1) th R-peak value. The method of claim 9,
The Nth RRI, which is the time interval between the Nth and (N + 1) th R-Peak values, is 140 to 150% or more of the small RRI of the previous N-1 RRI and the N-2 RRI, And a re-detection unit for re-detecting the R-peak value within the RRI. 14. The method of claim 13,
The re-
Setting a third threshold value using the N-th R-peak value,
If the peak value is larger than the third threshold value within the second time range from the third threshold value, the first N + 1 peak value from the first end point that is smaller than the third threshold value after the second time range Peak value among the peak values exceeding the third threshold within a range between the first peak value and the second peak value, as an R-peak value.

Images