KR101420345B1 - System and Method for Detecting Ｒ Wave of ECG Signal based on ＳＯＭ - Google Patents

Abstract

The present invention relates to an R-wave detection system for electrocardiogram signals based on a subtraction operation technique that enables accurate diagnosis of a healthcare system by performing R-wave detection of an ECG signal based on subtractive operation method (SOM) A pre-processing unit for pre-processing an ECG signal inputted through the ECG signal input unit to remove a variation of a baseline or a power line noise, a subtraction operation method, And a reverse search technique using the empirical threshold value and the RR interval in the sampling point is used to calculate the difference between the amplitude of the previous signal and the R signal to detect the R wave And a detection unit.

Description

TECHNICAL FIELD [0001] The present invention relates to a system and a method for detecting an R-wave of an ECG signal based on a subtraction operation technique,

The present invention relates to a healthcare system, and more particularly, to a subtraction operation method (SOM) that performs R wave detection of an electrocardiogram signal based on a subtractive operation method (SOM) to accurately diagnose a healthcare system with a minimum amount of calculation Based electrocardiogram signal and a method thereof.

In general, the analysis of ECG signals includes the steps of acquiring biological signals from the human body, removing noise, and detecting and diagnosing QRS waveforms. Based on analysis of ECG signals, It is widely used for diagnosis.

In the ECG analysis, the detection of the QRS region as the reference point should precede the QRS region. The QRS region is the most important characteristic curve of the ECG, representing the depolarization of the ventricle. T wave and ST segment are the criteria for clinical diagnosis.

In particular, the R-wave in the QRS domain can be regarded as the largest representative signal among the ECG signals. Since various feature points are detected on the basis of this point, much effort has been made in developing an algorithm for improving the detection performance of R-waves.

However, R wave detection is difficult to analyze due to various kinds of noise components. Particularly, the biggest problem in Holter electrocardiograph is the baseline fluctuation noise, which is caused by muscle contraction and rhythm of respiration in the area where electrodes are attached.

Further, detection of the T wave or P wave similar to the amplitude of the QRS domain to the R wave results in difficult detection.

In particular, real-time detection of R waves is required in a healthcare system that requires continuous monitoring of the health status of the general public. The ECG signal processing must take into account the efficiency of the hardware and software resources required for the algorithm, and the terminals used should use only as little resources as possible for signal processing for miniaturization and low power consumption.

In other words, it is necessary to design a suitable algorithm that can classify various arrhythmias by detecting accurate R waves with a minimum amount of computation.

Generally, the detection of R wave is divided into a preprocessing process for enhancing the QRS domain and a process for enhancing the R wave by applying a threshold value. In the preprocessing process, the QRS domain is emphasized using linear filtering and nonlinear signal processing methods.

As another detection method, a method of extracting a band corresponding to a frequency band of a QRS region using a wavelet transform has been used. So far, many studies have been done on the detection of R-wave of ECG signal, but Pan and Tompkins proposed the most prominent one.

However, the R wave detection technique of the prior art is based on the assumption that the measurement target is measured in a fixed state, and efficiency of hardware and software resources is not considered. Therefore, even though these techniques have a high detection rate of R wave, this method is not applicable to the healthcare environment that has recently been emphasized.

Especially, in a healthcare system that continuously monitors the health status of the general public, it is necessary to process ECG signals in real time. To do this, the terminals used must use only as little resources as possible for signal processing for miniaturization and low power consumption.

In other words, it is necessary to design a suitable algorithm that can classify various arrhythmias by detecting accurate R waves with a minimum amount of computation.

In order to solve the problem of the conventional healthcare system, the present invention performs R-wave detection of an electrocardiogram signal based on a Subtractive Operation Method (SOM), so that accurate diagnosis of a healthcare system The present invention provides a system and method for detecting an R wave of an electrocardiogram signal based on a subtraction operation technique.

The present invention relates to a method for reducing the amount of computation by calculating R-wave of an electrocardiogram signal by using a subtraction operation method (SOM) and calculating a difference between an amplitude value and a previous signal based on a current signal, The present invention provides a system and a method for detecting an R wave of an electrocardiogram signal based on a subtraction operation technique that can be used for classification of an arrhythmia of a subject.

The present invention relates to an R-wave detection system of an electrocardiogram signal based on a subtraction operation technique for detecting an accurate R-point at a minimum calculation amount in a state where a normal biological signal property to a specific person is already known, And a method thereof.

The present invention relates to a method and apparatus for real-time processing of ECG signals in a healthcare system in which health status is continuously monitored, and an R-wave of an electrocardiogram signal based on a subtraction operation technique using only as few resources as possible for signal processing for miniaturization and low- And to provide a detection system and method.

The present invention removes noise after preprocessing from the ECG signal and performs R-wave detection by applying an empirical threshold value and a reverse search technique to remove various noise included in the electrocardiogram signal to accurately detect the R wave And a system and a method for detecting an R wave of an electrocardiogram signal based on a subtraction operation technique.

The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned can be clearly understood by those skilled in the art from the following description.

According to another aspect of the present invention, there is provided a system for detecting an R wave of an electrocardiogram signal based on a subtraction operation technique, comprising: an ECG signal input unit for inputting an ECG signal; an ECG signal input through the ECG signal input unit, A pre-processing unit for eliminating variations in power line noise and power line noise, applying a subtracting operation technique to divide a difference signal into a positive signal and a negative signal, and applying a reverse search technique using an empirical threshold value and an RR interval in a sampling point, And an R wave detector for detecting an R wave only by calculating a difference between the amplitude of the previous signal and the previous signal.

Here, in the R wave detector, the sampling point is a value corresponding to 144 to 432 sampling points at a sampling frequency of 360 Hz, and the period of each RR interval is about 0.4 to 1.2.

According to another aspect of the present invention, there is provided a method of detecting an R wave of an electrocardiogram signal based on a subtraction operation technique, the method comprising the steps of pre-processing an input ECG signal by a morphological operation to remove baseline variation or power line noise, And detecting the R wave in the sampling point, and verifying the R wave detection by applying a reverse search technique using the RR interval.

Here, in the step of removing noise,

으로, 입력값(f(n))은 ECG 신호(f)와 구조요소(g(n))의 팽창과 침식을 위한 연산으로 최소화하고, 형태 연산을 통한 필터링 된 결과(Mf (n))는 입력값(f(n))과 ECG 신호 평균값(f _avg (n))의 차이를 통해 계산되는 것을 특징으로 한다.

, The input value f (n) is minimized to an operation for expansion and erosion of the ECG signal f and the structural element g (n) , and the filtered result Mf (n) Is calculated through the difference between the input value f (n) and the average value of the ECG signal ( f _avg (n) ).

The result of the form operation depends on the type of the structure element used in the operation, and the shape of the structure element is determined by the size, shape, and length of the structure element.

The step of dividing the difference signal into a positive signal and a negative signal and detecting an R wave in the sampling point includes: obtaining a difference signal X _d (n) between a current signal and a previous signal through a subtracting operation; ( X _D ) for peak detection if the value satisfies the condition of the threshold value and to 0 if the value does not satisfy the condition of the threshold value; and a step of converting the candidate signal to 0.14s Searching for the maximum value from left to right within the point (point). That is, it measures the time existing between the positions of the adjacent maximum values and finds an accurate maximum value.

The difference signal X _d (n)

와 같이 현재와 이전 시간의 진폭 값 차이를 통해 계산되며, 양과 음의 신호로 각각 나누어지는 것을 특징으로 한다.

Is calculated through the difference between the amplitude values of the current time and the previous time, and is divided into positive and negative signals, respectively.

And outputs the difference signal X _d (n ) as a candidate signal X _D

을 이용하여 변환하고, 경험적 문턱 값에 의해 0, 양의 피크 값(+X _d (n)), 음의 피크 값(-X _d (n))으로 각각 분할되고,

,

는 MIT-BIH 부정맥 데이터베이스 레코드의 양과 음의 진폭 값에 대한 평균인 것을 특징으로 한다.

X _d (n ) and a negative peak value ( -X _d (n) ), respectively, by the empirical threshold value,

,

Is an average of the positive and negative amplitude values of the MIT-BIH arrhythmia database record.

The step of verifying the R wave detection by applying the inverse search technique using the RR interval may include the steps of: determining whether a new R wave occurs again within a predetermined time period when an R wave is detected through a subtraction operation; And if it occurs again within the set time, it is assumed that an FP that misinterprets the T wave as an R wave is generated, and that it is detected within the allowable range of the average RR interval, and then the R wave is detected again.

The R wave detection system and method of the electrocardiogram signal based on the subtraction operation technique according to the present invention has the following effects.

First, the R wave of the electrocardiogram signal is detected by the subtraction operation method (SOM), and the calculation amount can be reduced by calculating only the difference of the amplitude value from the previous signal based on the current signal.

Second, accurate detection can be performed with minimal computation volume by detecting R-wave of ECG signal based on subtraction operation technique.

Third, after eliminating noise after preprocessing from ECG signal, R wave can be detected by applying empirical threshold and inverse search technique to remove various noise included in ECG signal to accurately detect R wave.

Fourth, because it does not require complicated signal processing, it can be used to classify various arrhythmias in real time in a healthcare system that performs long-term monitoring.

Fifth, it is possible to judge heart disease in real time by detecting an accurate R point with a minimum calculation amount in a state in which a normal bio-signal property to a specific person is already known.

Sixth, the ECG signal can be processed in the healthcare system in real time, and signal processing is possible for miniaturization and low power consumption of the terminals used.

Seventh, it is possible to forecast the disease of the patient in advance and contribute to the systematic health care field of the socioeconomic on the basis of reduction of economic burden and health promotion.

1 is a block diagram of an R-wave detection system of an electrocardiogram signal based on a subtraction operation technique according to the present invention
FIG. 2 is a block diagram showing a preprocessing process through a form operation
3 is a flowchart illustrating a method of detecting an R wave of an electrocardiogram signal based on a subtraction operation technique according to the present invention.
4 is a flowchart showing a method of detecting an R wave using the dynamic reverse search technique according to the present invention.
5 is a flowchart showing a result of baseline noise removal through a morphological filter;
6 and 7 are graphs showing the results of R wave detection by the subtraction operation technique
8 is a graph showing the results of R wave detection using the dynamic reverse search technique

Hereinafter, a preferred embodiment of a system and method for detecting an R wave of an electrocardiogram signal based on a subtraction operation technique according to the present invention will be described in detail.

The features and advantages of the R wave detection system and method of an electrocardiogram signal based on the subtraction operation technique according to the present invention will be apparent from the following detailed description of each embodiment.

FIG. 1 is a configuration diagram of an R-wave detection system of an electrocardiogram signal based on a subtraction operation technique according to the present invention, and FIG. 2 is a configuration diagram illustrating a preprocessing process by a form operation.

The present invention is to reduce the amount of computation by calculating the difference of the amplitude value from the previous signal based on the current signal by performing R-wave detection of the electrocardiogram signal by the subtraction operation method (SOM), thereby enabling accurate diagnosis in real time.

The present invention removes noise after preprocessing from the ECG signal and performs R-wave detection by applying an empirical threshold value and a reverse search technique to remove various noise included in the electrocardiogram signal to accurately detect the R wave It is.

1, the system for detecting an R wave of an electrocardiogram signal based on a subtraction operation technique according to the present invention comprises an ECG signal input unit 11 for inputting an ECG signal, an ECG signal input unit 11, A preprocessing unit 12 for removing a variation of a baseline and a power line noise and a subtracting operation method to divide the subtracted signal into a positive signal and a negative signal and performing a reverse search using the empirical threshold value and the RR interval in the sampling point And an R-wave detector 13 for detecting an R-wave by calculating only the difference between the amplitude of the previous signal and the previous signal based on the current signal.

The R wave detection process of the ECG signal based on the subtraction operation technique of the R wave detection system of the electrocardiogram signal based on the subtraction operation technique according to the present invention having the above configuration can be roughly divided into steps of removing noise after preprocessing from the ECG signal, And a step of verifying the efficiency of the R wave detection by applying a reverse search technique using an RR interval.

For accurate diagnosis, it is necessary to remove various noise included in ECG signal to detect R wave accurately. Especially, the preconditioning process is necessary to remove baseline fluctuation caused by breathing or power line noise.

The present invention applies a pre-processing method based on a morphological operation in order to reduce the deformation of the original signal and minimize the complexity of the calculation amount.

The morphological operations are divided into opening and closing operations according to the order of application of erosion and dilation operations. The elimination operation expands after erosion and the filling operation performs erosion after expansion. It proceeds in a way to remove the negative peak value.

The process of removing the ECG noise using the form operation applied in the present invention is shown in FIG.

As shown in FIG. 2, the process includes sequentially applying erosion and expansion operations using a structural element (SE), and finally obtaining a filtered result through a difference from the average.

Simplifies the operator to reduce processing complexity with a minimal amount of computation.

In order to detect the R wave, it is necessary to preserve only the QRS region of the ECG signal. For this purpose, the input value f (n ) is the sum of the expansion of the ECG signal f and the structure element g (n) Minimize the operation for erosion.

The filtered results finally through the operation form (Mf (n)) is calculated from the difference of the input value (f (n)) with the ECG signal mean value (f _avg (n)).

Here, the reason why the average value is selected as the signal for the final operation is that the average operation is easy to eliminate the DC offset. The result of the form operation depends on the type of the structure element used in the operation, and the shape of the structure element is determined by the size, shape, and length of the structure element.

In the present invention, a triangular waveform structure element is selected to select a structure element similar to a QRS region.

In the present invention, the algorithm for detecting the R wave using the subtraction operation technique is to reduce the calculation amount by calculating only the amplitude difference between the current signal and the previous signal based on the current signal.

The R wave is first detected using the empirical threshold and the subtracting method of the preprocessed signal through the form operation, and the reverse search method using the RR interval is used if it is not detected by this value.

The R wave detection process using the subtraction operation technique is shown in FIG.

3 is a flowchart illustrating a method of detecting an R wave of an electrocardiogram signal based on a subtraction operation technique according to the present invention.

First, a difference ( X _d (n) ) between a current signal and a previous signal is obtained through a subtracting operation (S301), threshold values T1 and T2 are calculated (S302), and if the value satisfies the condition of the threshold value (S303), it is converted into a candidate signal ( X _D ) for peak detection, or to 0 otherwise (S304)

Thereafter, the candidate signal is searched from left to right within the sampling point to find the maximum value (S306).

That is, it measures the time existing between the positions of the adjacent maximum values and finds an accurate maximum value.

The subtracted signal X _d (n) is calculated through the difference between the amplitude values of the current time and the previous time,

The subtracted signal should be converted into a candidate signal for R wave detection. The candidate signal X _D is 0, positive peak value (+ X _d (n) ) by empirical threshold value as shown in Equation (3) And the peak value ( -X _d (n) ).

,

는 MIT-BIH 부정맥 데이터베이스 레코드의 양과 음의 진폭 값에 대한 평균을 나타낸다.In Equation 3,

,

Represents the mean of the positive and negative amplitude values of the MIT-BIH arrhythmia database record.

T1 and T2 are thresholds for finding the maximum positive and negative values of the difference signal.

According to medical definition, the cycle of each RR interval is about 0.4 to 1.2. That is, it corresponds to 144 to 432 sampling points at a sampling frequency of 360 Hz.

,

는 각각 0.48과 0.47mV가 된다.For example, the RR interval interval for MIT-BIH record 103 is approximately 0.864 s (311 sampling points), where the total sum of the sampling points is 2084 * 311 for 30 minutes,

,

Are 0.48 and 0.47 mV, respectively.

The threshold values T1 and T2 are based on the respective experimental values, and the purpose of setting these values is to find the correct positive and negative maximum values.

However, if these two values are set too small, a non-zero value existing in the RR interval of the ECG signal is found so that the maximum value and the maximum value to be searched for will not be easily found. If the values are set too high, By removing the R wave, we can not find the exact amount and the maximum value of the sound.

And R wave detection using dynamic reverse search method using RR interval is as follows.

4 is a flowchart illustrating a method of detecting an R wave using the dynamic reverse search technique according to the present invention.

FP (False Positive) may occur when P wave or T wave is detected as R wave when R wave is detected, and TN (True Negative) which can not detect it may occur even if R wave actually exists.

In order to prevent such erroneous detection, the present invention uses a dynamic backward searching method (DBSM).

The process of searching again for the erroneously detected R wave using the inverse search technique is shown in FIG.

If a new R wave is generated again within a period of 200 ms or more within 360 ms (S402), it is assumed that an FP for causing the T wave to be an R wave is generated (S405) It is confirmed whether or not it is detected within 150% of the RR interval (S404) and the R wave is detected again (S406) so that the false detection can be minimized.

In this case, when the reverse search period is set to be longer than 200 ms and 360 ms, the absolute reflex period generally lasts for about 200 ms. After the relative reflex period lasts for about 150 ms, the excitation is slowly recovered. because of the contraction of the heart after latency

The average detection rate of R wave was 99.41%.

Especially, it is considered to be more efficient in the healthcare system which considers the efficiency of resources for miniaturization and low power, and warns after abnormal condition is detected.

The R wave detection system and method of the electrocardiogram signal based on the subtraction operation technique according to the present invention has a baseline noise canceling characteristic through a form operation.

In the R wave detection system of the electrocardiogram signal based on the subtraction operation technique according to the present invention, when the MIT / BIH 108 data with a sampling frequency of 360 Hz is selected and applied to a morphological filter, the baseline noise elimination characteristic .

As shown in FIG. 5 (a), the ECG signal is applied to the algorithm according to the present invention, and the signal with the baseline variation removed is shown in FIG. 5 (a) b.

It can be seen that not only the baseline variation curve was removed but also the DC component was also removed.

FIG. 6 shows a result of baseline noise removal according to a morphological filter by selecting a section between 1230 and 1240s in which the baseline variation occurs in the MIT / BIH 108 record.

The present invention removes baseline noise through a morphological operation and detects an R wave through a subtracting method for a noise canceled signal. The extraction of a positive peak and a negative peak for R wave detection is performed for the current and previous time The difference between the amplitude values is used.

In order to confirm whether the extraction of positive and negative peak values by the subtracting method is suitable for R wave detection, the efficiency is confirmed through the 100th and 203th records as follows.

FIGS. 6 and 7 are graphs showing R wave detection results through a subtraction operation technique.

The reason for the two records is that most of the 100th region represents the normal region, and the 203th record contains the arrhythmia signal, and the effectiveness of the R wave detection technique through the subtraction operation is compared This is because it is appropriate.

As shown in FIG. 6, in the case of the 100th record, positive and negative peak values exist at the same time intervals as shown in FIG. 6 (b) in a region where an R wave exists at a certain time interval. .

As shown in FIG. 7, in the case of the 208th record, although there is a section that degrades the R wave detection performance due to the inverse of the QRS region due to the early ventricular contraction bit, the R wave is actually present 7 (b), positive and negative peak values are present at the same time, and the result of the subtraction operation is accurately detected through the subtraction operation.

8 is a graph showing the results of R wave detection using the dynamic reverse search technique.

In the present invention, FP (False Positive) or TN (True Negative) is generated at the time of R wave detection through the subtraction operation, which may result in a lower detection rate. In the present invention, The efficiency of the R wave detection through the dynamic backward searching method is verified in FIG.

Although an erroneous detection interval exists in the interval of the 108th record, the dynamic backward search technique accurately detects the result.

The system and method for detecting an R wave of an electrocardiogram signal based on a subtraction operation technique according to the present invention detects an R wave using an empirical threshold value and a subtraction signal for detecting an R wave of an electrocardiogram signal, For the efficiency of detection, we use reverse search method using RR interval.

In order to verify the performance of the morphological filter, we have selected MIT / BIH 108 record with relatively high baseline variation, which shows that not only the baseline variation curve but also the DC component is removed.

In order to confirm the suitability of the R wave detection by the subtracting operation technique, the efficiency of the R signal was confirmed through the record of the normal signal of 100 and the abnormal signal of 203. As a result, .

Finally, in order to verify the performance of the dynamic reverse search technique, we have tested 108 records in which there are many baseline noise and arrhythmia intervals. The results show that the undetectable R wave is correctly detected by the reverse search technique .

The R wave detection system and method of the electrocardiogram signal based on the subtraction operation technique according to the present invention described above can detect the R wave of the electrocardiogram signal by the subtraction operation method (SOM) Which can reduce the computational complexity and enable accurate diagnosis in real time.

As described above, it will be understood that the present invention is implemented in a modified form without departing from the essential characteristics of the present invention.

It is therefore to be understood that the specified embodiments are to be considered in an illustrative rather than a restrictive sense and that the scope of the invention is indicated by the appended claims rather than by the foregoing description and that all such differences falling within the scope of equivalents thereof are intended to be embraced therein It should be interpreted.

11. ECG signal input section 12. Preprocessing section
13. R wave detector

Claims (9)

An ECG signal input unit for inputting an ECG signal;
A preprocessor for preliminarily processing an ECG signal input through the ECG signal input unit by type calculation to remove fluctuations of baseline and power line noise;
The subtraction operation method is applied to calculate the subtracted signal by calculating the difference between the amplitude values of the current time and the previous time. The obtained subtracted signal is divided into a positive signal and a negative signal, and the empirical threshold value and the RR interval And an R-wave detector for detecting the R-wave by calculating a difference between the amplitude of the previous signal and the previous signal based on the search signal. The apparatus according to claim 1, wherein, in the R wave detector,
Wherein the sampling point is a value corresponding to 144 to 432 sampling points at a sampling frequency of 360 Hz, and a period of each RR interval is 0.4 to 1.2. The R wave detection system of the electrocardiogram signal based on the subtraction operation technique. Preprocessing the input ECG signal by a form operation to remove baseline variation or power line noise;
Calculating a difference signal by calculating a difference between an amplitude value of a current time and a previous time by applying a subtraction operation technique, dividing the subtracted signal into a positive signal and a negative signal, and detecting an R wave in the sampling point;
And detecting R wave detection by applying an inverse search technique using an RR interval. 4. The method of claim 3,

to,
The input value f (n) is minimized by an operation for expansion and erosion of the ECG signal f and the structural element g (n) , and the filtered result Mf (n) (f (n)) with the ECG signal mean value (f _avg (n)) R-wave detecting method of the subtraction operation scheme based ECG signals, characterized in that, calculated by the difference. 5. The method of claim 4, wherein the result of the form operation is dependent on the type of the structure element used in the operation, and the type of the structure element is determined by the size, shape, and length of the structure element. A method for detecting an R wave of an electrocardiogram signal. 4. The method of claim 3, wherein the subtracting operation technique is applied to calculate a subtracted signal by calculating the difference between the current and previous amplitude values, dividing the subtracted signal into a positive signal and a negative signal, Comprising:
A difference signal X _d (n) between the current and previous signals is obtained through a subtracting operation and the threshold values T1 and T2 are calculated. If the value of the difference signal satisfies the condition of the threshold value, X _D ), and if not, converting to 0,
And searching for a maximum value by searching the candidate signal from the left to the right in the sampling point. 7. The apparatus of claim 6, wherein the subtracted signal X _d (n)

Wherein the difference between the current and previous amplitude values is divided into a positive and a negative signal, respectively. The method according to claim 6, wherein the difference signal ( X _d (n) ) is converted into a candidate signal ( X _D )

And then,
Is divided into 0 by the empirical threshold value, a positive peak value (+ X _d (n) ), and a negative peak value (- X _d (n)

,

Is an average of the positive and negative amplitude values of the MIT-BIH arrhythmia database record. 4. The method of claim 3, wherein the step of verifying the R wave detection by applying a reverse search technique using the RR interval comprises:
Determining whether a new R wave occurs again within a set time when an R wave is detected through a subtracting operation;
When the new R wave occurs again within the set time, it is assumed that the FP that misinterprets the T wave as the R wave is generated, and that it is detected whether it is within the allowable range of the average RR interval and then the R wave is detected again A method of detecting an R wave of an electrocardiogram signal based on a subtraction operation technique.

Images