CN107029354A - The impact signal detection method of heart defibrillator - Google Patents

Abstract

The impact signal detection method of heart defibrillator according to an embodiment of the invention may include steps of：Receive electrocardiosignal；First conversion is carried out to the electrocardiosignal of the receiving；Second is carried out to the electrocardiosignal by the described first conversion to convert；Utilize the signal of change characteristic value by the second conversion；And the characteristic value of the calculating is used as input value and impact signal is detected by Weighted Fuzzy membership function.

Description

Shock signal detection method of cardiac defibrillator

technical field

The present invention relates to a shock signal detection method of a cardiac defibrillator capable of shortening the time before defibrillation is performed by the cardiac defibrillator that applies a shock to a cardiac arrest patient.

Background technique

Many people die from sudden cardiac arrest (Sudden Cardiac Arrest, SCA), most of which are due to life-threatening cardiac arrhythmias. When profound cardiac arrhythmias occur on a life-threatening level, patient survival is closely linked to prompt response. As a result, the spread of automated external defibrillators (Automated External Defibrillators) has spread widely. The above-mentioned automatic external defibrillator is a device that automatically judges whether it is necessary to apply an electric shock and then applies an electric shock without professional medical personnel judging the electrocardiographic signal.

Compared with cardiac arrest patients in other countries such as the United States, the survival rate of cardiac arrest patients in Korea is very low because there is not enough environment that allows rapid emergency measures. Therefore, it is preferable that not only emergency personnel but also ordinary people use an automatic external defibrillator to automatically defibrillate a patient on the spot and then move the patient to a hospital.

The most common type of life-threatening arrhythmia is coarse ventricular fibrillation (coarse VF) or rapid ventricular tachycardia (rapid ventricular tachycardia, rapidVT). In the case of a cardiac arrest patient with the above-mentioned coarse VF or rapid VT, only defibrillation within 10 minutes can survive, and as time passes, the survival rate decreases by 7 to 10% per minute. External defibrillators provide rapid defibrillation to improve survival.

Therefore, in order to perform defibrillation, a step of detecting shock signals such as coarse VF and rapid VT is first required. Among the algorithms used to detect the above-mentioned shock signals so far, the Time-Delay algorithm published in 2007 detects coarse VF through ECG signals between 8 seconds. Moreover, the Parameter Set algorithm can detect coarse VF, rapid VT, NSR and N (other arrhythmias) through ECG signals within 10 seconds. The radial basis function (RBF) detection algorithm published in 2008 can detect VF/VT and NSR/N and strive for better results, but it is universal because of the pre-selection of ECG signals for experiments lower question.

As described above, research has been devoted to methods for shortening the time required to perform defibrillation. In practice, an algorithm for charging a capacitor before detecting an ECG signal in order to apply a shock signal has also been developed, but a method of shortening the ECG signal detection time itself and improving detection accuracy has not been disclosed.

Contents of the invention

An object of the present invention is to provide a shock signal detection method of a cardiac defibrillator capable of shortening the time until the defibrillator performs defibrillation, which applies a shock to a patient in cardiac arrest.

The shock signal detection method of a cardiac defibrillator according to an embodiment of the present invention may include the following steps: receiving an electrocardiographic signal; performing a first transformation on the accepted electrocardiographic signal; performing a second transformation on the signal; calculating a feature value using the second transformed signal; and using the calculated feature value as an input value and detecting a shock signal through a weighted fuzzy membership function.

The received signal may be a signal in digital form.

The first transform may be wavelet transform.

The second transformation may be a time delay transformation (Time Delay Transform).

The delay transformation can be realized by the formula "Y(x)=X(x+0.5)-X(x)", in which, the X(x) can be the first transformed signal, Y(x ) may be a second transformed signal.

The shock signal detection method of the cardiac defibrillator may further include a step of judging whether the cardiac arrest is by using the second converted signal.

The step of judging whether cardiac arrest may include a step of judging by using a sum of absolute values of the second transformed signal values.

The step of judging whether cardiac arrest may include a step of judging by using the distance between the second transformed signal values.

If the judgment result of whether the cardiac arrest is judged is negative, the step of calculating the characteristic value may be performed.

The step of calculating the characteristic value may include the step of calculating an average distance between values within a preset amplitude interval in the second transformed signal value.

The step of calculating the characteristic value may include the step of calculating the standard deviation of the slope of the straight line connecting the second transformed signal values.

The step of calculating the characteristic value may include calculating at least one of the number of peaks, the values before and after the peaks, and the standard deviation of the peaks included in the preset range of the second transformed signal.

The calculated eigenvalues may be seven.

The shock signal detection method according to one embodiment of the present invention has the effect that the time required for applying a defibrillation shock signal can be shortened by detecting a shock signal according to an electrocardiographic signal within 7 seconds.

Description of drawings

FIG. 1 is a flow chart illustrating a method for detecting an impact signal according to an embodiment of the present invention.

FIG. 2 is a graph illustrating an initial input signal in a shock signal detection method according to an embodiment of the present invention.

Fig. 3a is a graph showing a first transformed signal in a shock signal detection method according to an embodiment of the present invention.

Fig. 3b is a graph showing a second transformed signal in the shock signal detection method according to one embodiment of the present invention.

Figures 4a-4b are diagrams illustrating a method for detecting an asystole (Asystole) from the second transformed signal as shown in Figure 3b.

FIGS. 5 to 10 are diagrams illustrating schematic methods of calculating characteristic values for detecting shock signals when asystole is not detected according to FIGS. 4 a to 4 b.

detailed description

The specific structural descriptions or specific functional descriptions of the various embodiments of the present invention disclosed in this description or this application are provided only for the purpose of explaining the various embodiments of the present invention. Therefore, various embodiments according to the invention may be embodied in various different forms, and should not be construed as being limited to the various embodiments described in this specification or this application.

Embodiments according to the present invention can be modified in various ways and have various forms, and specific embodiments are illustrated in the drawings and described in detail in this specification or application. However, it is not intended to limit the present invention to specific examples, and it should be understood that all changes, equivalents, and substitutions included in the spirit and technical scope of the present invention are included.

In the present invention, terms such as first and/or second may be used to describe various components, but the components are not limited to the terms. The terms are only used to distinguish a certain component from other components. For example, within the protection scope of the concept according to the present invention, the first component can be named as the second component, and similarly, the second component can be named as the second component. a component.

When it is stated that a certain structural element is "connected" to another structural element, it can be understood as being directly connected to another structural element, or as having another structural element in between. Conversely, when it is stated that a certain structural element is "directly connected" to another structural element, it should be understood that there is no other structural element in between. On the other hand, other terms that describe the relationship between structural elements, namely "... "between", "right between", "adjacent to ..." and "directly adjacent to ..." etc. can also be understood in the same way.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, singular forms also include plural forms unless the context clearly dictates otherwise. It should also be understood that when the terms "comprise", "comprising", "have" and/or "having" are used herein, it means that said features, numbers, steps, The presence of operations, elements, components and/or combinations thereof does not preclude the presence or addition of one or more other features, numbers, steps, operations, elements, components and/or combinations thereof.

Unless otherwise defined, all terms including technical or scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs. Generally used terms that are the same as those defined in dictionaries should be interpreted as having meanings that are consistent with the contextual meanings of related technologies, and should not be interpreted as idealized or excessive forms if they are not clearly defined in this application meaning of transformation.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings to explain the present invention in detail. In the drawings, the same symbols are assigned to the same constituent elements.

FIG. 1 is a flow chart illustrating a method for detecting an impact signal according to an embodiment of the present invention.

Referring to FIG. 1 , the shock signal detection method according to one embodiment of the present invention can be implemented by a shock signal detection device, and can be executed by a receiving unit, a central control unit, etc. constituting the shock signal detection device.

First, an electrocardiographic signal is received from electrode pads attached to the patient's body (S101), and a first transformation is performed on the received electrocardiographic signal (S103). Perform a second transformation on the first transformed signal (S105). Whether the cardiac arrest (Asystole) is judged by using the second converted signal (S107), and when the judgment result is negative, a plurality of feature values are calculated (S109). As an example, the calculated eigenvalues may be seven. The above-mentioned feature value becomes an input value of a weighted fuzzy membership function obtained through pre-learning, and the shock signal can be detected from the weighted fuzzy membership function (S111).

Asystole is a state in which a defibrillation shock is not required because the heart has stopped.

In the case of the existing method, six eigenvalues are calculated by using the first converted signal to determine whether the cardiac arrest is present, but in the case of the present invention, the difference from the prior art is that, based on the After the first transformed signal is subjected to the second transformation, whether cardiac arrest is determined based on the second transformed signal, and seven eigenvalues are calculated. In addition, the method of determining whether or not the cardiac arrest is also different. In the past, whether cardiac arrest is judged by the average value of the peak points of the first transformed signal, but in the present invention, by outputting the sum of the absolute values of the peak points of the second transformed signal and the peak point The average value of the sum of the distances between them is used as the input value of the weighted fuzzy membership function with a preset learning process to determine whether cardiac arrest.

The calculation method of the above-mentioned plurality of feature values can be determined based on the learning result due to the weighted fuzzy membership function.

Fig. 2 is a graph showing the initial input signal in the shock signal detection method according to one embodiment of the present invention, and Fig. 3a is a signal showing a first transformation in the shock signal detection method according to one embodiment of the present invention Fig. 3b is a graph showing a second transformed signal in the shock signal detection method according to an embodiment of the present invention.

Referring to FIG. 2, the graph shown in FIG. 2 represents an example of an initial input signal before undergoing a first transformation, Haar Wavelet transformation. Fig. 3a is a diagram representing a first transformed signal. The first transformed signal is also subjected to a second transformation to become a signal having the form shown in the diagram of Fig. 3b. The second transformation can be realized by the following formula.

Mathematical formula 1

Y(x)=X(x+0.5)-X(x)

In the above formula, part Y(x) represents the signal after the second transformation, and part X(x) represents the signal after the first transformation. The above-mentioned second conversion is realized by using a time delay (Time Delay). As an example, the second transformation may be a time delay transformation (Time Delay Transform, TDT).

The first transform, namely wavelet transform, is a method that can make up for the shortcomings of the Fourier transform that provides global frequency characteristic information by simultaneously analyzing the frequency characteristics at a specific location in time in terms of signal processing. Discontinuous wavelet transform can convert time -Separation of frequency signals into non-continuous signals of different scales. The distribution of the peak points of the signals in the order d1, d2, d3 becomes increasingly simple, so that only the desired signal can be displayed. In the above wavelet transformation result, a second transformation is performed on the d3 value.

Figures 4a-4b are diagrams illustrating a method for detecting an asystole (Asystole) from the second transformed signal as shown in Figure 3b.

Referring to Fig. 4a and Fig. 4b, Fig. 4a shows the method of using the absolute value to judge whether the cardiac arrest (Asystole), and Fig. 4b shows the method of using the distance between the signal values to judge whether the cardiac arrest (Asystole). As shown in FIG. 4a and FIG. 4b, in order to determine whether the cardiac arrest is present, the signal after the second transformation can be used as the original signal (Original Signal). Whether cardiac arrest (Asystole) can be judged by the method shown in Fig. 4a and Fig. 4b. Its specific process is as above.

Whether the cardiac arrest is judged by the method shown in Figure 4a and Figure 4b, if the judgment result is yes, there is no need to apply shock, it is judged that the shock state cannot be applied, and the subsequent steps are ended, or, when a new command or signal is received In the case of input, the steps of determining whether cardiac arrest occurs through the first conversion and the second conversion may be repeated again.

5 to 10 show various exemplary methods of calculating a plurality of feature values for judging whether a shock can be applied (detecting a shock signal) when the judgment result of asystole is negative.

FIG. 5 shows an example of calculating eigenvalues using a phase space reconstruction (Phase Space Reconstruction, PSR) method. The above method is a method for analyzing dynamic waveforms or random signals based on phase space, which uses the value obtained by performing the second transformation on the d3 data of the first transformed signal, and represents the time t on the x-axis The amplitude unit (amplitude unit) value of the y-axis indicates the amplitude unit (amplitude unit) value at time t+0.5, so as to obtain a two-dimensional chart, and the characteristics are obtained by the number of boxes (boxes) represented at this time value. As shown in the upper left of Figure 5, in the case of a normal waveform (Normal Sinus Rhythm, NSR), it occupies a small space as shown in the lower left, but as shown in the upper right, in the case of an abnormal waveform (ventricular fibrillation: VF ) takes up a large space as shown on the right below. Based on the constituted phase space (Phase Space), the eigenvalue (d) can be obtained from the following equation.

Mathematical formula 2

FIG. 6 is a diagram illustrating a method of calculating a feature value by the number of peak points. From the second converted ECG signal, the number of peak points can be extracted. The left side of Fig. 6 shows the number of peak points in the case of normal waveform (Normal Sinus Rhythm, NSR), and the right side shows the number of peak points in the case of ventricular fibrillation (VF). difference. A point above the average value belonging to the entire point is used as a peak point. The number of the above-mentioned peak points is used as a feature value.

In the case of FIG. 7 , the feature value can be calculated by using two points on the front side and two points on the rear side of the peak point. Find the peak point in the signal after the second transformation to obtain two values (p1, p2) on the side before the peak point and two values (n1, n2) on the back side of the peak point, where the value of p1 can be used as a feature value.

FIG. 8 is a graph showing a method of using the standard deviation of points inside and outside a specific amplitude interval as a feature value, which uses the standard deviation of points inside and outside a preset amplitude interval. As shown in FIG. 8, it can be seen from a comparison between the case of the normal waveform and the case of ventricular fibrillation (VT) that point distributions differ according to the amplitude of the second converted signal.

FIG. 9 is a diagram showing a method of using the second converted signal and using the average value of the distances between points inside and outside a preset amplitude interval as a feature value. For example, when a certain range of the amplitude interval is preset, the average distance of points within the range can be calculated. The above range may be -50 to 50 or -200 to 200.

FIG. 10 is a diagram illustrating a method in which points and standard deviations of inclinations between points are used as feature values when points of the second transformed signal are connected into a straight line. Although the above-mentioned feature values have not been used in the past, in the present invention, the above-mentioned feature values are used as input values of the weighted fuzzy membership function, so that the shock signal detection time can be shortened and the detection accuracy can be improved.

The above schematic eigenvalues are used as the input features of the neural network based on the weighted fuzzy membership function, so that the weighted fuzzy membership function can output a judgment result on whether the signal is impacted. Neural Network with Weighted Fuzzy Membership Function (NEWFM) is a supervised learning fuzzy neural network for classification by utilizing the boundary value of the weighted fuzzy membership function learned from the input.

The structure of NEWFM consists of three levels: input, hyper-box, and class. The input layer is composed of n input nodes, and one feature value is input to each input node. The superbox hierarchy is composed of m hyperbox nodes, the first hyperbox node B1 is connected with a class node, and has n fuzzy sets.

Through the above method, the time required for judging whether to shock the signal is shortened, so that defibrillation can be quickly applied in the environment where defibrillation is required, thereby improving the survival rate of patients.

In addition, the above shock signal detection method is implemented by computer readable codes/instructions/programs. For example, the method can be implemented on a general-purpose digital computer operating the above-mentioned codes/instructions/programs by using a computer-readable recording medium. Examples of the above-mentioned computer-readable recording medium include magnetic storage media (such as optical disks, floppy disks, hard disks, magnetic tapes, etc.), optically readable media (such as CD-ROM or DVD), and storage media such as carrier wave (such as transmission through the Internet) forms. medium. Also, the embodiments of the present invention can be implemented by medium(s) embedding computer-readable code such that a plurality of computer systems connected by a network process and operate in a distributed manner. Furthermore, the functional programs, codes, and code segments realized by the method of the present invention can be easily deduced by programmers belonging to the technical field of the present invention.

The above description only relates to the description of a specific embodiment of the technical spirit of the present invention, and those skilled in the art of the present invention shall not deviate from the basic features of the present invention to make various modifications or changes. Therefore, the embodiments disclosed in the present invention are not intended to limit the technical spirit of the present invention but to describe the technical spirit, and the scope of the present invention should not be limited to the embodiments. The protection scope of the present invention should be determined by the claims, and all interpretations of the technical spirit within the equivalent scope should fall within the scope of the present invention.

Claims (13)

1. the impact signal detection method of a kind of heart defibrillator, it is characterised in that comprise the following steps：

Receive electrocardiosignal；

First conversion is carried out to the electrocardiosignal of the receiving；

Second is carried out to the electrocardiosignal by the described first conversion to convert；

Utilize the signal of change characteristic value by the second conversion；And

The characteristic value of the calculating is used as input value and impact signal is detected by Weighted Fuzzy membership function.

2. the impact signal detection method of heart defibrillator according to claim 1, it is characterised in that the receiving Signal is digital form signal.

3. the impact signal detection method of heart defibrillator according to claim 1, it is characterised in that described first becomes It is changed to wavelet transformation.

4. the impact signal detection method of heart defibrillator according to claim 1, it is characterised in that described second becomes It is changed to delay conversion.

5. the impact signal detection method of heart defibrillator according to claim 4, it is characterised in that the delay becomes Change and realized by formula " Y (x)=X (x+0.5)-X (x) ", in formula, the X (x) is the signal by the first conversion, and Y (x) is warp Cross the signal of the second conversion.

6. the impact signal detection method of heart defibrillator according to claim 1, it is characterised in that also include：

Judge whether asystolic step by using the signal by the second conversion.

7. the impact signal detection method of heart defibrillator according to claim 6, it is characterised in that the judgement is No asystolic step includes being judged by using the summation of the absolute value of the signal value by the second conversion The step of.

8. the impact signal detection method of heart defibrillator according to claim 6, it is characterised in that the judgement is No asystolic step is included by using the step judged by the distance between signal value of the second conversion Suddenly.

9. the impact signal detection method of heart defibrillator according to claim 6, it is characterised in that if it is determined that being The no asystolic judged result is no, the step of just carrying out the calculating characteristic value.

10. the impact signal detection method of heart defibrillator according to claim 1, it is characterised in that the calculating The step of characteristic value including calculate it is described by second conversion signal value in amplitude set in advance interval in value it Between average distance the step of.

11. the impact signal detection method of heart defibrillator according to claim 1, it is characterised in that the calculating The step of characteristic value, connects the step of the standard deviation of the gradient of the straight line of the signal value by the second conversion including calculating Suddenly.

12. the impact signal detection method of heart defibrillator according to claim 1, it is characterised in that the calculating The step of characteristic value including calculate the number of peak value of the signal by the second conversion, the front and rear value of peak value, be contained in it is pre- The step of at least one in the standard deviation of the peak value of the scope first set.

13. the impact signal detection method of heart defibrillator according to claim 1, it is characterised in that the calculating Characteristic value be seven.