CN110558974B - Electrocardiogram signal analysis method based on extreme value energy decomposition method - Google Patents

Electrocardiogram signal analysis method based on extreme value energy decomposition method Download PDF

Info

Publication number
CN110558974B
CN110558974B CN201910841968.XA CN201910841968A CN110558974B CN 110558974 B CN110558974 B CN 110558974B CN 201910841968 A CN201910841968 A CN 201910841968A CN 110558974 B CN110558974 B CN 110558974B
Authority
CN
China
Prior art keywords
extreme
signal
value
energy
component
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201910841968.XA
Other languages
Chinese (zh)
Other versions
CN110558974A (en
Inventor
周作建
宁新宝
姜晓东
王�华
王斌斌
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu Huakang Information Technology Co ltd
Original Assignee
Jiangsu Huakang Information Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jiangsu Huakang Information Technology Co ltd filed Critical Jiangsu Huakang Information Technology Co ltd
Priority to CN201910841968.XA priority Critical patent/CN110558974B/en
Priority to PCT/CN2019/121414 priority patent/WO2021042589A1/en
Publication of CN110558974A publication Critical patent/CN110558974A/en
Application granted granted Critical
Publication of CN110558974B publication Critical patent/CN110558974B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
    • A61B5/024—Detecting, measuring or recording pulse rate or heart rate
    • A61B5/0245—Detecting, measuring or recording pulse rate or heart rate by using sensing means generating electric signals, i.e. ECG signals
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7271—Specific aspects of physiological measurement analysis

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Molecular Biology (AREA)
  • Animal Behavior & Ethology (AREA)
  • Pathology (AREA)
  • Physics & Mathematics (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Medical Informatics (AREA)
  • Signal Processing (AREA)
  • Surgery (AREA)
  • Biophysics (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Cardiology (AREA)
  • Physiology (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Psychiatry (AREA)
  • Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)

Abstract

The invention discloses an electrocardiogram signal analysis method based on an extreme value energy decomposition method, which comprises the steps of obtaining an ECG signal x (t) of an unknown state at a given time and a given sampling frequency; performing denoising preprocessing on the ECG signal x (t); taking the denoised ECG signal x (t) as an original signal, decomposing the original signal x (t) into n extreme value modal function components and a margin, decomposing the original signal x (t) into n extreme value modal function components representing components of different frequency bands of the original signal, and judging whether the ECG signal is an abnormal electrocardiogram signal or not according to the n extreme value modal function components. The invention adopts an extreme value energy decomposition method to analyze an ECG signal, decomposes an original signal into a plurality of components, namely an extreme value component function, and calculates the energy of each component to obtain the energy distribution of the component.

Description

Electrocardiogram signal analysis method based on extreme value energy decomposition method
Technical Field
The invention relates to electrocardiogram signal analysis, in particular to an electrocardiogram signal analysis method based on an extreme value energy decomposition method.
Background
Physiological signals are generated by the interaction of multiple systems of a living body, and the time and the intensity of the action of different systems are different, so that the physiological signals have complexity in time and space. ECG (electrocardiograph) signals reflect the electrical activity process of heart beats, and have important guiding significance for basic heart functions and disease diagnosis. The ECG signal comprises several different wave groups of P wave, QRS wave, T wave and the like, each wave group comprises different frequency components, and the energy distribution proportion of the different frequency components in the ECG signal is different. Researches show that 99% of energy of an ECG signal is concentrated in a range of 0-40 Hz, and the frequency distribution range and the energy ratio have certain rules for different wave groups. The ECG signal energy distribution is changed due to heart diseases, and the study on the ECG energy distribution has important clinical guiding significance for revealing the function change caused by the heart diseases.
In the energy (spectrum) research of the ECG, the traditional classical methods such as spectrum analysis, time domain analysis and the like are many, however, the ECG signal is a non-stationary nonlinear signal, and the traditional frequency domain analysis method is more suitable for analyzing a stationary signal and only can give global frequency information. In addition, when the frequency domain is segmented, the traditional method is divided according to a plurality of fixed frequency points, and the fluctuation characteristics of different signals and the difference between the signals are not considered.
Therefore, it is desired to solve the above problems.
Disclosure of Invention
The purpose of the invention is as follows: the invention aims to provide an electrocardiogram signal analysis method based on an extreme value energy decomposition method, which can intuitively reflect the real rule of electrocardiogram energy distribution and signal fluctuation by adopting less data.
The technical scheme is as follows: in order to achieve the above purpose, the invention discloses an electrocardiogram signal analysis method based on an extreme value energy decomposition method, which comprises the following steps:
(1) acquiring an ECG signal x (t) of an unknown state at a given time and a given sampling frequency;
(2) carrying out denoising pretreatment on the ECG signal x (t); the specific method of pretreatment is as follows: filtering the ECG signal by a 40Hz zero-phase FIR low-pass filter to eliminate high-frequency noise, and then removing baseline drift by a median filter;
(3) taking the denoised ECG signal x (t) as an original signal, calculating all local extreme points of the original signal, and then connecting all the extreme points and all the minimum points of the original signal by adopting a spline curve to respectively form an upper envelope line emaxAnd a lower envelope eminObtaining an envelope mean value signal m (t) ═ e of the upper and lower envelope linesmax+emin)/2；
(4) Subtracting the envelope mean signal m (t) from the original signal x (t) to obtain h (t) ═ x (t) -m (t); then judging whether h (t) meets the judgment condition of the extreme value modal function, if not, returning h (t) to the step (3) as the original signal until hk(t) if the condition for determining the extremum mode function is satisfied, c is recorded1(t)＝hk(t) as a first extreme modal function component;
(5) subtracting the first extreme mode function component c from the original signal x (t)1(t) obtaining the residue r1(t)＝x(t)-c1(t) then judging hk(t) whether a stopping criterion is met, and if not, r1(t) as a new original sequence x (t), returning to steps (3) and (4) until hk(t) satisfying the stopping criterion, obtaining the 2 nd, 3 rd, … th, n extreme value modal function components and the margin rn(t) the original signal x (t) is then decomposed into n extreme modal function components and a residual, i.e.
Figure BDA0002193998280000021
(6) Opposite polar mode function component ci(t), i is 1, 2, …, n, and the center frequency of each extreme modal function component is obtained by performing spectrum analysis;
(7) n extreme value modal function components obtained by decomposing the original signal x (t) represent components of different frequency bands of the original signal, and then the energy of each component is calculated
Ei＝∫|ci(t)|2dt，i＝1，2，…，n
Normalizing each energy value to obtain a normalized energy distribution vector
pi＝Ei/E，i＝1，2，…，n
Wherein the content of the first and second substances,
Figure BDA0002193998280000022
first component p1Representing the energy of the highest frequency band, and representing the ratio of the energy distribution of the signal in the highest frequency bandExample, the last component pnRepresenting the proportion of the energy distribution of the signal in the lowest frequency band range; drawing a normalized energy distribution graph according to the normalized energy distribution vector, wherein the abscissa represents component levels, the ordinate represents normalized energy distribution vector values, the curve represents an average value, and the error bar represents a standard deviation;
(8) normalized energy distribution vector p of ECG signal of unknown stateiNormalized energy distribution vector p with standard ECG signaliCarrying out significance detection to obtain a probability value P on each extreme value modal function componentiDetermining a probability value P of the first extreme mode function component1And a probability value P on the second extremum mode function component2If the signal is smaller than the probability standard value, returning to the step (1) to acquire the signal again;
(9) in the presence of a catalyst satisfying P1And P2If the energy distribution vector p of the first extreme mode function component is smaller than the standard value of the probability1Energy distribution vector p smaller than second extreme modal function component2Then the ECG signal is determined to be an abnormal ECG signal.
Wherein, the minimum data amount N of the original signal x (t) is 2n+1And n is the number of the decomposed extreme mode function components.
Preferably, the condition for determining the extreme mode function in step (4) is: (a) in the whole data sequence, the number of the extreme points is equal to or different from the number of the zero-crossing points by one; (b) and the upper and lower envelopes are symmetric with respect to the time axis at any time.
Further, h in said step (5)k(t) the formula for satisfying the stopping criterion is:
Figure BDA0002193998280000031
and representing a screening threshold, and taking the value between 0.2 and 0.3.
Further, the standard value of the probability in the step (8) and the step (9) is 0.05.
And (3) the repetition frequency of the 6 th extreme value modal function component waveform obtained in the step (5) is the human heart rate, namely the cardiac cycle.
Has the advantages that: compared with the prior art, the invention has the following remarkable advantages: the invention adopts an Extreme Energy Decomposition (EED) method to analyze an ECG signal, decomposes an original signal into a plurality of components, namely an extreme component function, calculates the Energy of each component to obtain the Energy distribution of the component, and can judge whether the signal is normal or not according to the comparison of Energy values of different levels of components; the invention can decompose the signal into different time-level signals from high frequency to low frequency according to the fluctuation rule of the biomedical signal; and for the ECG signal, the repetition frequency can be determined from a special level thereof, namely the heart rate value (comprising an instantaneous heart rate value and an average heart rate value) of the ECG signal, and the method for obtaining the heart rate value is more accurate than the traditional method; the data length obtained by the extremum decomposition on all levels is the same, so that the data length cannot be reduced, and the extremum decomposition method can be used for short-time data analysis, namely, an accurate result can be obtained by analysis with less data quantity; the EED is not susceptible to noise for different levels of component energy analysis.
Drawings
FIG. 1 is a diagram of an original signal in the present invention;
FIG. 2 is a diagram illustrating the envelope of the original signal according to the present invention;
FIG. 3 is a schematic diagram of a subtracted envelope mean signal of an original signal in accordance with the present invention;
FIG. 4 is a schematic diagram of obtaining a first extreme modal function component according to the present invention;
FIG. 5 is a schematic flow chart of the extreme energy decomposition method of the present invention;
FIG. 6 is an EED decomposition diagram of ECG signal in embodiment 1 of the present invention;
FIG. 7 is a spectrum chart of an ECG signal in embodiment 1 of the present invention;
FIG. 8 is an EED decomposition diagram of ECG signal in embodiment 2 of the present invention;
FIG. 9 is a time-frequency diagram of ECG of a healthy person in the practice 2 of the present invention;
FIG. 10 is a time-frequency plot of the ECG of a CHF patient in accordance with practice 2 of the present invention.
Detailed Description
The technical scheme of the invention is further explained by combining the attached drawings.
As shown in fig. 1, fig. 2, fig. 3, fig. 4 and fig. 5, the electrocardiogram signal analyzing method based on the extreme value energy decomposition method of the present invention comprises the following steps:
(1) acquiring an ECG signal x (t) of an unknown state at a given time and a given sampling frequency; the minimum data amount N of the original signal x (t) is 2n+1Wherein n is the number of resolved extremum modal function components;
(2) carrying out denoising pretreatment on the ECG signal x (t); the specific method of pretreatment is as follows: filtering the ECG signal by a 40Hz zero-phase FIR low-pass filter to eliminate high-frequency noise, and then removing baseline drift by a median filter;
(3) taking the denoised ECG signal x (t) as an original signal, calculating all local extreme points of the original signal, and then connecting all the extreme points and all the minimum points of the original signal by adopting a spline curve to respectively form an upper envelope line emaxAnd a lower envelope eminObtaining an envelope mean value signal m (t) ═ e of the upper and lower envelope linesmax+emin)/2；
(4) Subtracting the envelope mean signal m (t) from the original signal x (t) to obtain h (t) ═ x (t) -m (t); then judging whether h (t) meets the judgment condition of the extreme value modal function, if not, returning h (t) to the step (3) as the original signal until hk(t) if the condition for determining the extremum mode function is satisfied, c is recorded1(t)＝hk(t) as a first extreme modal function component; the determination condition of the extreme value modal function is as follows: (a) in the whole data sequence, the number of the extreme points is equal to or different from the number of the zero-crossing points by one; (b) at any time, the upper envelope line and the lower envelope line are symmetrical relative to a time axis;
(5) subtracting the first extreme mode function component c from the original signal x (t)1(t) obtaining the residue r1(t)＝x(t)-c1(t) then judging hk(t) whether stop is satisfiedStopping criterion, if not satisfied, will r1(t) as a new original sequence x (t), returning to steps (3) and (4) until hk(t) satisfying the stopping criterion, obtaining the 2 nd, 3 rd, … th, n extreme value modal function components and the margin rn(t) the original signal x (t) is then decomposed into n extreme modal function components and a residual, i.e.
Figure BDA0002193998280000041
Wherein h isk(t) the formula for satisfying the stopping criterion is:
Figure BDA0002193998280000042
representing a screening threshold, and taking the value between 0.2 and 0.3; the extremum modal decomposition that satisfies the stopping criterion then satisfies the following two conditions: (a) finally obtained extreme value modal function component cn(t) or the remainder rn(t) is less than a predetermined threshold; (b) residual signal rn(t) becomes a monotonous signal from which an extreme mode function signal cannot be extracted again;
(6) opposite polar mode function component ci(t), i is 1, 2, …, n, and the center frequency of each extreme modal function component is obtained by performing spectrum analysis;
(7) n extreme value modal function components obtained by decomposing the original signal x (t) represent components of different frequency bands of the original signal, and then the energy of each component is calculated
Ei＝∫|ci(t)|2dt，i＝1，2，…，n
Normalizing each energy value to obtain a normalized energy distribution vector
pi＝Ei/E，i＝1，2，…，n
Wherein the content of the first and second substances,
Figure BDA0002193998280000051
first component p1Representing the energy of the highest frequency band, representing the proportion of the energy distribution of the signal in the range of the highest frequency band, the last minuteQuantity pnRepresenting the proportion of the energy distribution of the signal in the lowest frequency band range; drawing a normalized energy distribution graph according to the normalized energy distribution vector, wherein the abscissa represents component levels, the ordinate represents normalized energy distribution vector values, the curve represents an average value, and the error bar represents a standard deviation;
(8) normalized energy distribution vector p of ECG signal of unknown stateiNormalized energy distribution vector p with standard ECG signaliCarrying out significance detection to obtain a probability value P on each extreme value modal function componentiDetermining a probability value P of the first extreme mode function component1And a probability value P on the second extremum mode function component2If the signal is smaller than the probability standard value, returning to the step (1) to acquire the signal again; wherein the probability standard value is selected to be 0.05;
(9) in the presence of a catalyst satisfying P1And P2If the energy distribution vector p of the first extreme mode function component is smaller than the standard value of the probability1Energy distribution vector p smaller than second extreme modal function component2Then the ECG signal is determined to be an abnormal ECG signal.
The repetition frequency of the 6 th extreme value modal function component waveform obtained by the invention is the heart rate of the human body, namely the cardiac cycle.
The Extreme Energy Decomposition (EED) method adopted by the invention is a method based on the concept of an extreme mode function, wherein the extreme mode function is a signal with single frequency which simultaneously meets the following two conditions:
(a) the number of extreme points (including maxima and minima) and the number of zero-crossing points must be equal or differ by at most one throughout the data sequence;
(b) at any moment, the average value of an upper envelope line formed by the local maximum value points and a lower envelope line formed by the local minimum value points is zero, namely the local upper envelope line and the local lower envelope line are locally symmetrical relative to a time axis;
the above two conditions, condition (a) is similar to the requirement of the gaussian normal stationary process for the traditional narrow band, and condition (b) ensures that the instantaneous frequency calculated by the extreme value modal function has physical significance.
The standard selection of the extreme value modal function decomposition termination is moderate, and the conditions are too strict, so that the last extreme value modal function components lose significance; conditions are too loose, which can result in loss of useful components; in practical application, the number of layers of the extremum modal function components to be decomposed can be set according to requirements, and when the number of layers of decomposition is met, the calculation is terminated.
Example 1
The invention discloses an electrocardiogram signal analysis method based on an extreme value energy decomposition method, which comprises the following steps:
(1) acquiring an ECG signal x (t) at a given time and a given sampling frequency from a healthy person normal sinus database nsrdb of physionet; the data length is 8s, and the minimum data amount required by the original signal x (t) is N-2n+1＝29Wherein n is the number of the decomposed extreme mode function components, and n is 8;
(2) carrying out denoising pretreatment on the ECG signal x (t); the specific method of pretreatment is as follows: because the ECG energy is mainly concentrated in 0-40 Hz, the ECG signal is filtered by a 40Hz zero-phase FIR low-pass filter to eliminate high-frequency noise, and then baseline drift is removed by a median filter;
(3) taking the denoised ECG signal x (t) as an original signal, calculating all local extreme points of the original signal, and then connecting all the extreme points and all the minimum points of the original signal by adopting a spline curve to respectively form an upper envelope line emaxAnd a lower envelope eminObtaining an envelope mean value signal m (t) ═ e of the upper and lower envelope linesmax+emin)/2；
(4) Subtracting the envelope mean signal m (t) from the original signal x (t) to obtain h (t) ═ x (t) -m (t); then judging whether h (t) meets the judgment condition of the extreme value modal function, if not, returning h (t) to the step (3) as the original signal until hk(t) if the condition for determining the extremum mode function is satisfied, c is recorded1(t)＝hk(t) as a first extreme modal function component; decision strip of extreme mode functionThe parts are as follows: (a) in the whole data sequence, the number of the extreme points is equal to or different from the number of the zero-crossing points by one; (b) at any time, the upper envelope line and the lower envelope line are symmetrical relative to a time axis;
(5) subtracting the first extreme mode function component c from the original signal x (t)1(t) obtaining the residue r1(t)＝x(t)-c1(t) adding r1(t) as a new original sequence x (t), returning to the steps (3) and (4) to obtain the 2 nd, 3 rd, … th and 8 th extreme value modal function components and the margin r8(t) the original signal x (t) is then decomposed into 8 extreme modal function components and a residual, i.e.
Figure BDA0002193998280000061
As shown in FIG. 6, the ECG signal can be divided into three wave groups of P, QRS and T waves according to the characteristics of the waveform; c can be obtained by decomposing 8 extreme value modal function components1(t) and c2(t) the component of the decomposition of the QRS complex with the highest frequency representing the highest frequency in the ECG signal, c3(t) decomposition components representing superposition of high frequency QRS complex and P wave in ECG signal, c4(T) decomposition components of the superposition of high-frequency QRS complex, P-wave and T-wave in ECG signal, c5(T) decomposition components representing superposition of QRS complex, P wave and low frequency part of T wave in ECG signal, c6(t) cardiac cycles representing the rhythm of the heart beat, c7(t) and c8(t) represents the heart's circadian rhythm on a larger time scale representative of the long-term rhythm of the heart; observing the amplitude of the signal, wherein the component with the highest frequency has higher amplitude and the highest energy; the component with the lowest frequency has lower amplitude and lower energy;
(6) opposite polar mode function component ci(t), i is 1, 2, …, 8, and performing spectrum analysis to obtain the center frequency of each extreme modal function component, as shown in table 1; wherein pair c1Performing spectrum analysis to obtain a spectrogram shown in FIG. 7, and obtaining c1The center frequency of (A) is about 20Hz, and the main frequency is concentrated in the range of 15-25 Hz; as shown in the existing research, the frequency spectrum range of the P wave is 0-18 Hz (+ -3 Hz),the energy is mainly concentrated at 5-12 Hz; the QRS wave has a frequency spectrum range of 0-37 Hz (+ -5 Hz), and energy is mainly concentrated in 6-18 Hz; the frequency spectrum range of the T wave is 0-8 Hz (+ -2 Hz), and the energy is mainly concentrated in 0-8 Hz. As can be seen from comparison of Table 1, the frequency band of QRS complex mainly includes c1、c2Two components, P-waves containing predominantly c3，c4Two components, T-waves containing predominantly c4～c8The component part of (a). It should be noted that the inclusion is not intended to mean that each component is determined by a specific ECG complex (P, QRS, T wave) or that each ECG complex is included in a specific component, and that the above complex-energy relationship is a primary correspondence, not all. E.g. c representing the low frequency part5～c8The components are determined by the superposition of the low frequency portions of the individual ECG complexes, rather than by a particular ECG complex. The extreme value modal function component and the extreme value modal function components of all layers of the invention both show that the extreme value modal function component of the ECG can represent a certain ECG wave group fluctuation condition and reflect the fluctuation rule of the ECG on different layers; compared with the traditional frequency domain analysis method, the EED method can directly observe the fluctuation condition of the ECG on each level, and is very intuitive. The repetition frequency of the obtained 6 th extreme value modal function component waveform is the human heart rate, namely the cardiac cycle.
TABLE 1 center frequencies of respective extreme modal function components
Figure BDA0002193998280000071
Example 2
The EED analysis method is used to analyze the energy distribution of the ECG at different levels in healthy persons and CHF patients.
An electrocardiogram signal analysis method based on an extreme value energy decomposition method for healthy people comprises the following steps:
(1) acquiring an ECG signal x (t) with the data length of 10s and the sampling frequency of 128Hz from a normal sinus database nsrdb of physionet; wherein the nsrdb database contains 18 healthy persons (age 34.3 + -8.4), the most needed raw signal x (t)Small data size N-2n+1＝29Wherein n is the number of the decomposed extreme mode function components, and n is 8;
(2) carrying out denoising pretreatment on the ECG signal x (t); the specific method of pretreatment is as follows: because the ECG energy is mainly concentrated in 0-40 Hz, the ECG signal is filtered by a 40Hz zero-phase FIR low-pass filter to eliminate high-frequency noise, and then baseline drift is removed by a median filter;
(3) taking the denoised ECG signal x (t) as an original signal, calculating all local extreme points of the original signal, and then connecting all the extreme points and all the minimum points of the original signal by adopting a spline curve to respectively form an upper envelope line emaxAnd a lower envelope eminObtaining an envelope mean value signal m (t) ═ e of the upper and lower envelope linesmax+emin)/2；
(4) Subtracting the envelope mean signal m (t) from the original signal x (t) to obtain h (t) ═ x (t) -m (t); then judging whether h (t) meets the judgment condition of the extreme value modal function, if not, returning h (t) to the step (3) as the original signal until hk(t) if the condition for determining the extremum mode function is satisfied, c is recorded1(t)＝hk(t) as a first extreme modal function component; the determination condition of the extreme value modal function is as follows: (a) in the whole data sequence, the number of the extreme points is equal to or different from the number of the zero-crossing points by one; (b) at any time, the upper envelope line and the lower envelope line are symmetrical relative to a time axis;
(5) subtracting the first extreme mode function component c from the original signal x (t)1(t) obtaining the residue r1(t)＝x(t)-c1(t) adding r1(t) as a new original sequence x (t), returning to the steps (3) and (4) to obtain the 2 nd, 3 rd, … th and 8 th extreme value modal function components and the margin r8(t) the original signal x (t) is then decomposed into 8 extreme modal function components and a residual, i.e.
Figure BDA0002193998280000081
(6) Opposite polar mode function component ci(t) Performing spectrum analysis to obtain the center frequency of each extreme value modal function component to obtain a frequency domain analysis result graph;
(7) 8 extreme value modal function components obtained by decomposing the original signal x (t) represent components of different frequency bands of the original signal, and then the energy of each component is calculated
Ei＝∫|ci(t)|2dt，i＝1，2，…，8
Normalizing each energy value to obtain a normalized energy distribution vector
pi＝Ei/E，i＝1，2，…，8
Wherein the content of the first and second substances,
Figure BDA0002193998280000082
first component p1Representing the energy of the highest frequency band, representing the proportion of the energy distribution of the signal in the highest frequency band, the last component pnRepresenting the proportion of the energy distribution of the signal in the lowest frequency range.
An electrocardiogram signal analysis method based on an extreme value energy decomposition method for CHF patients comprises the following steps:
(1) acquiring an ECG signal x (t) with a data length of 10s and a sampling frequency of 128Hz from a CHF database chfdb, wherein the chfdb database comprises 15 CHF patients (age 58.8 +/-9.1), and the minimum data quantity N (t) required by a raw signal x (t) is 2n+1＝29Wherein n is the number of the decomposed extreme mode function components, and n is 8;
the method comprises the following steps of (2) carrying out denoising pretreatment on an ECG signal x (t), wherein the specific method of the pretreatment is that the ECG signal is filtered by a 40Hz zero-phase FIR low-pass filter to eliminate high-frequency noise because the ECG energy is mainly concentrated in 0-40 Hz, and then baseline drift is removed by a median filter;
(3) taking the denoised ECG signal x (t) as an original signal, calculating all local extreme points of the original signal, and then connecting all the extreme points and all the minimum points of the original signal by adopting a spline curve to respectively form an upper envelope line emaxAnd a lower envelope eminIs obtained atEnvelope mean value signal m (t) of lower envelope (e)max+emin)/2；
(4) Subtracting the envelope mean signal m (t) from the original signal x (t) to obtain h (t) ═ x (t) -m (t); then judging whether h (t) meets the judgment condition of the extreme value modal function, if not, returning h (t) to the step (3) as the original signal until hk(t) if the condition for determining the extremum mode function is satisfied, c is recorded1(t)＝hk(t) as a first extreme modal function component; the determination condition of the extreme value modal function is as follows: (a) in the whole data sequence, the number of the extreme points is equal to or different from the number of the zero-crossing points by one; (b) at any time, the upper envelope line and the lower envelope line are symmetrical relative to a time axis;
(5) subtracting the first extreme mode function component c from the original signal x (t)1(t) obtaining the residue r1(t)＝x(t)-c1(t) adding r1(t) as a new original sequence x (t), returning to the steps (3) and (4) to obtain the 2 nd, 3 rd, … th and 8 th extreme value modal function components and the margin r8(t) the original signal x (t) is then decomposed into 8 extreme modal function components and a residual, i.e.
Figure BDA0002193998280000091
(6) Opposite polar mode function component ci(t), i is 1, 2, …, 8, performing spectrum analysis to obtain the center frequency of each extreme value mode function component, and obtaining a frequency domain analysis result graph, as shown in fig. 9, the energy of the healthy person is distributed in the range of 0-40 Hz, the energy proportion of the high frequency part above 20Hz is high, as shown in fig. 10, the CHF energy is mainly concentrated below 20Hz, and the energy of the high frequency part is obviously reduced;
(7) 8 extreme value modal function components obtained by decomposing the original signal x (t) represent components of different frequency bands of the original signal, and then the energy of each component is calculated
Ei＝∫|ci(t)|2dt，i＝1，2，…，8
Normalizing each energy value to obtain a normalized energy distribution vector
pi＝Ei/E，i＝1，2，…，8
Wherein the content of the first and second substances,
Figure BDA0002193998280000101
first component p1Representing the energy of the highest frequency band, representing the proportion of the energy distribution of the signal in the highest frequency band, the last component pnRepresenting the proportion of the energy distribution of the signal in the lowest frequency band range; drawing a normalized energy distribution map according to the normalized energy distribution vectors of healthy people and CHF patients, wherein the abscissa represents the component level, the ordinate represents the normalized energy distribution vector value, the curve represents the average value, and the error bar represents the standard deviation;
as shown in fig. 8, at level 1, the healthy person is higher in energy than the CHF patient, and as the levels increase, the healthy person is gradually reduced in energy; CHF patients gradually increase in levels 1-3 and reach the highest value in level 3; when the level is greater than 3, the energy is gradually reduced along with the increase of the level; the energy of the healthy people is mainly concentrated on levels 1-4, namely in a high frequency range, which shows that the heart of the healthy people has stronger short-time regulating capability; the relatively low-level energy of a CHF patient indicates a reduced short-term accommodation capacity, while the main power is concentrated at the intermediate level, at level 6, which reflects the heart rhythm, and the higher level, which indicates that the CHF patient has a higher energy proportion for the accommodation of the heart rhythm.
(8) Normalized energy distribution vector p for ECG signals of a CHF patientiNormalized energy distribution vector p of ECG signal of healthy personiCarrying out significance detection to obtain a probability value P on each extreme value modal function componentiWhen probability value PiLess than 0.05 indicates a significant difference between the two, as shown in table 2, at levels 1, 2, 3, 5, 6, and a decrease in energy at the decomposition low level in CHF patients may indicate a decrease in the ability of the heart to modulate on the hourly scale due to the disease; while healthy people have higher energy at the decomposition low level, which shows that the heart of healthy people has better short-term regulation capability,the method has better adaptability to the changes of external environment and body environment;
TABLE 2 energy vector T test for healthy and CHF patients
Figure BDA0002193998280000102
(9) In the presence of a catalyst satisfying P1And P2Less than 0.05, the energy distribution vector p of the first extreme modal function component of the CHF patient1Energy distribution vector p smaller than second extreme modal function component2。

Claims (6)

1. An electrocardiogram signal analysis method based on an extreme value energy decomposition method is characterized by comprising the following steps:
(1) acquiring an ECG signal x (t) of an unknown state at a given time and a given sampling frequency;
(2) carrying out denoising pretreatment on the ECG signal x (t); the specific method of pretreatment is as follows: filtering the ECG signal by a 40Hz zero-phase FIR low-pass filter to eliminate high-frequency noise, and then removing baseline drift by a median filter;
(3) taking the denoised ECG signal x (t) as an original signal, calculating all local extreme points of the original signal, and then connecting all the extreme points and all the minimum points of the original signal by adopting a spline curve to respectively form an upper envelope line emaxAnd a lower envelope eminObtaining an envelope mean value signal m (t) ═ e of the upper and lower envelope linesmax+emin)/2；
(4) Subtracting the envelope mean signal m (t) from the original signal x (t) to obtain h (t) ═ x (t) -m (t); then judging whether h (t) meets the judgment condition of the extreme value modal function, if not, returning h (t) to the step (3) as the original signal until hk(t) if the condition for determining the extremum mode function is satisfied, c is recorded1(t)＝hk(t) as a first extreme modal function component;
(5) subtracting the first extreme mode function component c from the original signal x (t)1(t) obtaining the residue r1(t)＝x(t)-c1(t) then judging hk(t) whether a stopping criterion is met, and if not, r1(t) as a new original sequence x (t), returning to steps (3) and (4) until hk(t) satisfying the stopping criterion, obtaining the 2 nd, 3 rd, … th, n extreme value modal function components and the margin rn(t) the original signal x (t) is then decomposed into n extreme modal function components and a residual, i.e.
Figure FDA0002193998270000011
(6) Opposite polar mode function component ci(t), i is 1, 2, …, n, and the center frequency of each extreme modal function component is obtained by performing spectrum analysis;
(7) n extreme value modal function components obtained by decomposing the original signal x (t) represent components of different frequency bands of the original signal, and then the energy of each component is calculated
Ei＝∫|ci(t)|2dt，i＝1，2，…，n
Normalizing each energy value to obtain a normalized energy distribution vector
pi＝Ei/E，i＝1，2，...，n
Wherein the content of the first and second substances,
Figure FDA0002193998270000012
first component p1Representing the energy of the highest frequency band, representing the proportion of the energy distribution of the signal in the highest frequency band, the last component pnRepresenting the proportion of the energy distribution of the signal in the lowest frequency band range; drawing a normalized energy distribution graph according to the normalized energy distribution vector, wherein the abscissa represents component levels, the ordinate represents normalized energy distribution vector values, the curve represents an average value, and the error bar represents a standard deviation;
(8) normalized energy distribution vector p of ECG signal of unknown stateiNormalized energy distribution vector p with standard ECG signaliCarrying out significance detection to obtain each extreme value modal functionProbability value P on a number componentiDetermining a probability value P of the first extreme mode function component1And a probability value P on the second extremum mode function component2If the signal is smaller than the probability standard value, returning to the step (1) to acquire the signal again;
(9) in the presence of a catalyst satisfying P1And P2If the energy distribution vector p of the first extreme mode function component is smaller than the standard value of the probability1Energy distribution vector p smaller than second extreme modal function component2Then the ECG signal is determined to be an abnormal ECG signal.
2. An electrocardiogram signal analysis method based on the extreme energy decomposition method according to claim 1, characterized in that: the minimum data amount N of the original signal x (t) is 2n+1And n is the number of the decomposed extreme mode function components.
3. An electrocardiogram signal analysis method based on the extreme energy decomposition method according to claim 1, characterized in that: the determination conditions of the extreme mode function in the step (4) are as follows: (a) in the whole data sequence, the number of the extreme points is equal to or different from the number of the zero-crossing points by one; (b) and the upper and lower envelopes are symmetric with respect to the time axis at any time.
4. An electrocardiogram signal analysis method based on the extreme energy decomposition method according to claim 1, characterized in that: h in the step (5)k(t) the formula for satisfying the stopping criterion is:
Figure FDA0002193998270000021
and representing a screening threshold, and taking the value between 0.2 and 0.3.
5. An electrocardiogram signal analysis method based on the extreme energy decomposition method according to claim 1, characterized in that: and (3) the probability standard value in the step (8) and the step (9) is 0.05.
6. An electrocardiogram signal analysis method based on the extreme energy decomposition method according to claim 1, characterized in that: and (3) the repetition frequency of the 6 th extreme value modal function component waveform obtained in the step (5) is the heart rate of the human body, namely the cardiac cycle.
CN201910841968.XA 2019-09-06 2019-09-06 Electrocardiogram signal analysis method based on extreme value energy decomposition method Active CN110558974B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201910841968.XA CN110558974B (en) 2019-09-06 2019-09-06 Electrocardiogram signal analysis method based on extreme value energy decomposition method
PCT/CN2019/121414 WO2021042589A1 (en) 2019-09-06 2019-11-28 Extremum energy decomposition-based electrocardiogram signal analysis method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910841968.XA CN110558974B (en) 2019-09-06 2019-09-06 Electrocardiogram signal analysis method based on extreme value energy decomposition method

Publications (2)

Publication Number Publication Date
CN110558974A CN110558974A (en) 2019-12-13
CN110558974B true CN110558974B (en) 2020-11-03

Family

ID=68778164

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910841968.XA Active CN110558974B (en) 2019-09-06 2019-09-06 Electrocardiogram signal analysis method based on extreme value energy decomposition method

Country Status (2)

Country Link
CN (1) CN110558974B (en)
WO (1) WO2021042589A1 (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110881958B (en) * 2019-12-25 2022-12-06 福州数据技术研究院有限公司 Pulse signal non-physiological signal removing method for traditional Chinese medicine pulse diagnosis instrument
CN116992393B (en) * 2023-09-27 2023-12-22 联通(江苏)产业互联网有限公司 Safety production monitoring method based on industrial Internet of things

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102499670A (en) * 2011-11-23 2012-06-20 北京理工大学 Electrocardiogram baseline drifting correction method based on robust estimation and intrinsic mode function
CN104182625A (en) * 2014-08-15 2014-12-03 重庆邮电大学 Electrocardiosignal denoising method based on morphology and EMD (empirical mode decomposition) wavelet threshold value

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080269628A1 (en) * 2007-04-25 2008-10-30 Siemens Medical Solutions Usa, Inc. Denoising and Artifact Rejection for Cardiac Signal in a Sensis System
US7869864B2 (en) * 2007-07-09 2011-01-11 Dynacardia, Inc. Methods, systems and devices for detecting and diagnosing heart diseases and disorders
CN101496716A (en) * 2009-02-26 2009-08-05 周洪建 Measurement method for detecting sleep apnoea with ECG signal
CN106510619B (en) * 2016-10-11 2019-04-05 天津大学 ECG Signal Analysis method based on complex network and in the application being intelligently worn by
CN106344006B (en) * 2016-11-03 2018-06-26 太原理工大学 J wave detecting methods based on pole symmetric mode decomposition and support vector machines
CN108814579B (en) * 2018-04-16 2020-05-22 西安交通大学 Method for calculating heart rate variability through electrocardio and respiration combination based on EMD decomposition

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102499670A (en) * 2011-11-23 2012-06-20 北京理工大学 Electrocardiogram baseline drifting correction method based on robust estimation and intrinsic mode function
CN104182625A (en) * 2014-08-15 2014-12-03 重庆邮电大学 Electrocardiosignal denoising method based on morphology and EMD (empirical mode decomposition) wavelet threshold value

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
一种基于改进经验模态分解的癫痫脑电识别新方法;庞春颖等;《中国生物医学工程学报》;20131231;第32卷(第6期);第663-669页 *
基于Hilbert谱的心率变异信号时频分析方法;董红生等;《仪器仪表学报》;20110228;第32卷(第2期);第271-278页 *

Also Published As

Publication number Publication date
WO2021042589A1 (en) 2021-03-11
CN110558974A (en) 2019-12-13

Similar Documents

Publication Publication Date Title
CN110558973B (en) Computer equipment for executing electrocardiogram signal quantitative analysis method based on extreme value energy decomposition method
Beyramienanlou et al. Shannon’s energy based algorithm in ECG signal processing
Lin Frequency-domain features for ECG beat discrimination using grey relational analysis-based classifier
US7809433B2 (en) Method and system for limiting interference in electroencephalographic signals
CN110464337B (en) Heart rate variability signal analysis method based on extreme value energy decomposition method
Chakraborty et al. Determination of signal to noise ratio of electrocardiograms filtered by band pass and Savitzky-Golay filters
JP2009545345A (en) ECG signal processing method and apparatus
Malek et al. Automated detection of premature ventricular contraction in ECG signals using enhanced template matching algorithm
CN110558974B (en) Electrocardiogram signal analysis method based on extreme value energy decomposition method
CN110558959B (en) HRV signal analysis method for meditation training based on extreme value energy decomposition method
CN113384277B (en) Electrocardiogram data classification method and classification system
Yang et al. Removal of pulse waveform baseline drift using cubic spline interpolation
Wessel et al. Evaluation of renormalised entropy for risk stratification using heart rate variability data
Hegde et al. A review on ECG signal processing and HRV analysis
Yue et al. Time-frequency analysis of heart sounds in telemedicine consulting system for auscultation
Dembrani et al. Accurate detection of ECG signals in ECG monitoring systems by eliminating the motion artifacts and improving the signal quality using SSG filter with DBE
Karnewar et al. Preprocessing ECG signal by eliminating various noises using denoising methods
Shama et al. Hamming filter design for ECG signal detection and processing using co-simulation
Perlman et al. Cardiac arrhythmia classification in 12-lead ECG using synthetic atrial activity signal
Goyal et al. Study of HRV dynamics and comparison using wavelet analysis and Pan Tompkins algorithm
Patil et al. Discrimination between atrial fibrillation (AF) & normal sinus rhythm (NSR) using linear parameters
Shaik et al. Empirical mode decomposition for frequency analysis of heart rate variability
Georgieva-Tsaneva An Information System for Cardiological Data Studying
Georgieva-Tsaneva An Interactive Teaching System for Investigation of Heart Rate Variability
Bala HRV analysis of preterm infant ECG; tools, techniques and challenges

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant