CN110742599B - Electrocardiosignal feature extraction and classification method and system - Google Patents
Electrocardiosignal feature extraction and classification method and system Download PDFInfo
- Publication number
- CN110742599B CN110742599B CN201911060613.3A CN201911060613A CN110742599B CN 110742599 B CN110742599 B CN 110742599B CN 201911060613 A CN201911060613 A CN 201911060613A CN 110742599 B CN110742599 B CN 110742599B
- Authority
- CN
- China
- Prior art keywords
- wave
- signal
- electrocardiosignal
- positioning
- electrocardiosignals
- 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
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
- A61B5/24—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
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
-
- 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
- A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
-
- 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
- A61B5/7235—Details of waveform analysis
-
- 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
- A61B5/7235—Details of waveform analysis
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
Abstract
The application relates to an electrocardiosignal feature extraction and classification method and system, and the method mainly comprises the following steps: (1) sequentially positioning R waves, Q waves, S waves, T waves and P waves of the electrocardiosignals according to an adaptive threshold value, (2) slidingly intercepting the electrocardiosignals for 1 second by taking a division point between cardiac cycles as a starting point, wherein the window length of a sliding window is 1 second, and the step length is one cardiac cycle, (3) solving time domain characteristics and frequency domain characteristics of the electrocardiosignals for 1 second, solving morphological characteristics of complete cardiac cycle signals included in the signals for 1 second, (4) utilizing a rough set to carry out attribute reduction, and screening effective characteristics as input of a discrimination network; the length of the sliding window and the setting of the variable step length ensure that the complete cardiac cycle signal cannot be truncated, and the time-frequency domain characteristics and the morphological characteristics are simultaneously obtained in one-time signal detection by combining the fixed length and single cycle modes, so that the dimensionality of the characteristic set is increased; and the rough set is utilized to carry out attribute reduction on the feature set, so that the efficiency and the precision of classification and judgment are improved.
Description
Technical Field
The invention belongs to the field of electrocardiosignal processing, and particularly relates to an electrocardiosignal feature extraction and classification method and system.
Background
Electrocardiosignals (ECG) are widely used in clinical practices such as cardiovascular disease diagnosis and curative effect evaluation, and particularly used for diagnosing arrhythmia, postoperative heart conditions and the like, so that accurate extraction and effective classification of characteristics such as time domain, frequency domain, morphology and the like in the ECG signals are particularly important.
At present, the characteristic extraction method of the ECG signal is mostly divided into two modes of fixed-length interception signal and single cardiac cycle interception signal. The fixed-length interception signal is used for independently extracting the ECG with fixed length, the step length of a sliding window during signal extraction is also fixed, the mode can be used for extracting the time-frequency domain characteristics in the signal, but the signal of a complete cardiac cycle is usually intercepted, so that the signal of a single cardiac cycle is partially lost, the characteristic extraction is incomplete, and the fixed-length interception mode cannot acquire the morphological characteristics of the signal waveform; the signal acquisition mode of single cardiac cycle can acquire the signal of the complete cardiac cycle, can extract the morphological characteristics of the signal waveform in the cycle, but can not acquire the time-frequency domain characteristics, and both the two modes have limitations.
In addition, at present, a neural network and a Support Vector Machine (SVM) are generally used for classifying signal features so as to judge the type of an electrocardiosignal, and most of the existing classification methods input all the features into a classifier for judgment, wherein the features include irrelevant features, so that the classification efficiency is low and the judgment precision is reduced.
Disclosure of Invention
Based on the above, the invention aims to provide an electrocardiosignal feature extraction and classification method, wherein a sliding window uses a variable step length to ensure that a signal of a complete cardiac cycle cannot be cut off, a fixed-length signal and a signal of the complete cardiac cycle can be simultaneously obtained to simultaneously obtain a time-frequency domain feature and a morphological feature, and a rough set is used for carrying out advanced screening on effective features so as to overcome the defects of the prior art.
The invention provides an electrocardiosignal feature extraction and classification method, which comprises the following steps:
acquiring electrocardiosignals, and sequentially positioning R waves, Q waves, S waves, T waves and P waves of the electrocardiosignals according to a preset self-adaptive threshold;
calculating the midpoint of a time axis between the T wave and the P wave of the adjacent next cardiac cycle signal, taking the midpoint as a dividing point between different cardiac cycles, and taking the dividing point as a starting point to intercept the 1-second cardiac signal by using a sliding window, wherein the window length of the sliding window is 1 second, and the step length is one cardiac cycle;
solving time domain characteristics and frequency domain characteristics aiming at the intercepted 1 second electrocardiosignals;
determining the positions of two adjacent segmentation points in the 1 second electrocardiosignal, and solving morphological characteristics of the electrocardiosignal between the two adjacent segmentation points;
and performing attribute reduction on the obtained feature set by using the rough set, screening features irrelevant to the type of the electrocardiosignal to be judged, outputting effective features, and performing electrocardiosignal type judgment by using signal data corresponding to the effective features as the input of a neural network and an SVM (support vector machine).
Further, in order to filter out high frequency interference and baseline wander and preserve a significant portion of the cardiac signal, acquiring the cardiac signal comprises:
and carrying out 20Hz low-pass filtering and 0.5Hz high-pass filtering on the original electrocardiosignals of the human body to obtain the filtered electrocardiosignals.
Further, the positioning of the R-wave comprises:
and detecting the QRS complex of the electrocardiosignal by adopting a Pan-Tompkins method, and positioning the peak of which the amplitude is higher than the R wave threshold value in the QRS complex as the R wave.
Further, in order to increase the signal-to-noise ratio of the signal and enhance the R wave height, and reduce the possibility of misidentifying the T wave as the R wave, the detecting the QRS complex further comprises:
and filtering the electrocardiosignal by using a band-pass filter, and performing square operation on the filtered electrocardiosignal.
Further, the positioning of the Q-wave and the S-wave comprises:
and taking the R wave as a center, solving two coordinate points which are 15ms away from the R wave, respectively obtaining wave troughs between the two coordinate points and the R wave, and respectively positioning the wave troughs into the Q wave and the S wave according to a Q wave threshold and an S wave threshold.
Further, the positioning of the T-wave comprises:
and defining a first signal segment by taking the S wave as a starting point and the Q wave of the electrocardiosignal of the next cardiac cycle adjacent to the S wave as an end point, setting a T wave threshold value aiming at the first signal segment, and positioning the wave crest of the first signal segment higher than the T wave threshold value as the T wave.
Further, the positioning of the P-wave comprises:
and defining a second signal segment by taking the T wave as a starting point and the end point of the first signal segment as an end point, setting a P wave threshold value aiming at the second signal segment, and positioning a wave crest which is higher than the P wave threshold value in the second signal segment as a P wave.
Further, the adaptive threshold is updated by the formula
Thren＝Ampn-1×0.125+Thren-1×0.875，
Thre in the formulanSignal threshold, Amp, representing the nth detectionn-1Represents the n-1 measured signal amplitude, Thren-1Representing the signal threshold for the (n-1) th detection.
Further, the time domain features include at least one of an average slope, a median slope, an amplitude range, a signal integral, an average peak-to-peak amplitude, a signal root mean square.
Further, the Frequency domain characteristics include at least one of Amplitude Spectral Area (AMSA), Power Spectral Analysis (PSA), maximum Power (Max Power, MP), Peak Power Frequency (PF), Center Frequency (CF), Energy (EG).
Further, the morphological feature includes at least one of P, Q, R, S, T wave height, P-R interval, Q-T interval, S-T interval, QRS complex width, P wave width.
The invention also provides an electrocardiosignal feature extraction and classification system, which comprises:
the electrocardiosignal acquisition unit is used for acquiring electrocardiosignals to be processed;
the signal positioning unit is used for positioning P, Q, R, S, T waves of the electrocardiosignals according to the adaptive threshold;
the signal intercepting unit is used for intercepting the positioned electrocardiosignals for 1 second by utilizing the sliding window with variable step length;
the characteristic extraction unit is used for solving frequency domain, time domain and morphological characteristics of the intercepted 1 second electrocardiosignal;
and the rough set reduction unit is used for screening all the characteristics of the electrocardiosignals to be distinguished, outputting effective characteristics related to the type to be distinguished, and outputting the electrocardiosignal data corresponding to the effective characteristics to a next-stage distinguishing network.
Furthermore, the electrocardiosignal acquisition unit also comprises a low-pass and high-pass filtering module.
Furthermore, the signal positioning unit further comprises a QRS complex detection module and a threshold updating module.
According to the technical scheme, the invention has the following advantages:
the invention relates to an electrocardiosignal feature extraction and classification method and system, wherein the window length of a sliding window is set to be 1 second, the step length is one cardiac cycle, so that the intercepted signal can not intercept one complete cardiac cycle, and the time-frequency domain feature and the morphological feature can be simultaneously obtained in one-time signal detection by combining the fixed-length and single-cycle extraction mode, so that the dimension of the feature set is increased, and the feature extraction precision is improved; the cardiac cycle is used as the variable step length of the sliding window, so that the electrocardiosignals of the complete cardiac cycle are ensured not to be cut off, and the signal processing and analysis are facilitated; aiming at the defect that all characteristics of signals are used as input in the existing classification method and the operation redundancy efficiency is low, the invention utilizes a rough set to carry out attribute reduction on a characteristic set, screens out effective characteristics and eliminates unnecessary characteristics, and improves the efficiency and the accuracy of classification and judgment.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is a block diagram of an electrocardiosignal feature extraction and classification system according to an embodiment of the present invention
FIG. 2 is a flow chart of a method for extracting and classifying cardiac electrical signal features according to another embodiment of the present invention
FIG. 3 is a waveform diagram of an electrocardiographic signal obtained in accordance with another embodiment of the present invention
FIG. 4 is a schematic diagram of the positioning of the electrocardiosignal R wave according to another embodiment of the present invention
FIG. 5 is a schematic diagram of the positioning of Q wave and S wave of a cardiac electrical signal according to another embodiment of the present invention
FIG. 6 is a schematic diagram of the location of the electrocardiosignal P, Q, R, S, T according to another embodiment of the present invention
FIG. 7 is a schematic diagram of the location of a division point of a cardiac cycle according to another embodiment of the present invention
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, the present embodiment provides an electrocardiographic feature extraction and classification system for extracting features related to arrhythmia to improve the accuracy and efficiency of arrhythmia classification, which includes an electrocardiographic signal acquisition unit 110, a signal positioning unit 120, a signal interception unit 130, a feature extraction unit 140, and a rough set reduction unit 150;
the electrocardiosignal acquisition unit 110 further comprises a low-pass and high-pass filtering module 111;
the signal locating unit 120 further comprises a QRS complex detection module 121 and a threshold update module 122.
The system has the following steps when in work:
the electrocardiosignal acquisition unit 110 acquires an original electrocardiosignal of a human body, and performs 20Hz low-pass filtering and 0.5Hz high-pass filtering on the original electrocardiosignal by using the filtering module 111 to filter high-frequency interference and baseline drift and keep an effective part of the electrocardiosignal.
The filtered electrocardiographic signal is transmitted to the signal positioning unit 120 to perform positioning of P, Q, R, S, T waves, wherein when detecting R waves, the QRS complex detection module 121 needs to detect QRS complexes, and then positioning is performed according to a threshold. All the adaptive thresholds are initialized in the threshold updating module 122, and the thresholds are continuously updated according to a built-in updating algorithm to dynamically adjust the baseline of the thresholds, so that the positioning accuracy is higher.
It should be noted that the threshold updating module may also be externally disposed on the signal positioning unit, as long as the signal positioning unit can receive the real-time positioning threshold, and the position of the threshold updating module does not form a limitation to the present invention.
The electrocardiographic signal with the positioning mark is transmitted to a signal intercepting unit 130, which intercepts the signal by using a sliding window, the window length of the sliding window is 1 second, and the step length is one cardiac cycle; the feature extraction unit 140 first finds out time domain and frequency domain features from the intercepted 1 second signal, then detects at least two cardiac cycle division points in the 1 second signal, the signal between the division points is the signal of a complete cardiac cycle, and finds out morphological features from the partial signal.
The rough set reduction unit 150 performs attribute reduction on the feature set obtained above, including the time domain, the frequency domain and the morphological feature, to screen out effective features related to arrhythmia, and the feature set corresponding to the electrocardiographic signal is used as the input of the next stage discrimination neural network.
It should be noted that the system provided by the present invention is not limited to extracting arrhythmia-related features, and any embodiment of signal feature extraction that can be implemented by the method of the present invention falls within the scope of the present invention.
Referring to fig. 2 to 7, the present invention provides a method for extracting and classifying electrocardiographic signal features, which can simultaneously extract time-frequency domain and morphological features of electrocardiographic signals, increase the dimensionality of feature sets, and improve classification efficiency and precision.
Obtaining an original electrocardiosignal of a human body, performing low-pass filtering of 20Hz and high-pass filtering of 0.5Hz on the original electrocardiosignal, filtering high-frequency interference and baseline drift, and reserving an effective part of the electrocardiosignal to obtain the filtered electrocardiosignal.
In this embodiment, the sampling rate of the electrocardiographic signal is set to 360 sampling points within 1 second, and the waveform of the acquired portion of the electrocardiographic signal is as shown in fig. 3.
And respectively positioning the P wave, the Q wave, the R wave, the S wave and the T wave of the signal according to a preset self-adaptive threshold.
Detecting QRS complex of electrocardiosignal by Pan-Tompkins method, locating the peak with amplitude higher than R wave threshold in QRS complex as R wave, as shown in FIG. 4, marking the location result of R wave, and updating R wave threshold by using the formula
QRS_Thren＝R_Peakn-1×0.125+QRS_Thren-1×0.875，
Wherein QRS _ ThrenR-wave threshold for nth detection, R _ Peakn-1QRS _ Thre, the amplitude of the R wave measured at the n-1 st timen-1Is the R wave threshold value of the n-1 detection, and n is a positive integer.
In order to increase the signal-to-noise ratio of the signal, enhance the height of the R wave, and reduce the possibility of misidentifying the T wave as the R wave, the QRS group is detected by performing bandpass filtering and squaring on the electrocardiosignal.
Taking the obtained R wave as a center, obtaining two coordinate points which are 15ms away from the R wave, respectively obtaining wave troughs between the two coordinate points and the R wave, respectively positioning the Q wave and the S wave according to a Q wave threshold and an S wave threshold, wherein the Q wave threshold is updated to be Q _ Thren＝Q_Ampn-1×0.125+Q_Thren-1X 0.875, the update formula of S wave is S _ Thren＝S_Ampn-1×0.125+S_Thren-1X 0.875, wherein Q _ ThrenQ-wave threshold for nth detection, Q _ Ampn-1Q _ Thre, the Q wave amplitude measured for the n-1 th timen-1Threshold value of Q wave for n-1 detection, S _ ThrenS-wave threshold for nth detection, S _ Ampn-1S _ Thre, the S wave amplitude measured for the n-1 th timen-1Is the n-1 thAnd n is a positive integer.
The results of the positioning of the Q-wave and S-wave are shown in FIG. 5.
Defining a first signal segment by taking the obtained S wave as a starting point and the Q wave of the electrocardiosignal of the next cardiac cycle adjacent to the S wave as an end point, setting a T wave threshold value aiming at the first signal segment, positioning the wave crest of the first signal segment higher than the T wave threshold value as a T wave, and updating the T wave threshold value by a T _ Thre formulan＝T_Ampn-1×0.125+T_Thren-1×0.875，T_ThrenT wave threshold for nth detection, T _ Ampn-1T _ Thre, the T wave amplitude measured for the n-1 th timen-1Is the threshold value of the T wave detected at the n-1 th time.
Setting a P wave threshold value aiming at a second signal segment by taking the obtained T wave as a starting point and the end point of the first signal segment, namely the Q wave of the electrocardiosignal of the next adjacent cardiac cycle as an end point, positioning the wave crest of the second signal segment higher than the P wave threshold value as a P wave, and adopting a P wave threshold value updating formula as P _ Thren＝P_Ampn-1×0.125+P_Thren-1×0.875，P_ThrenP-wave threshold for nth detection, P _ Ampn-1P _ Thre, the P wave amplitude measured for the n-1 th timen-1Is the P wave threshold value of the n-1 detection. The results of the positioning marks for the P-wave, Q-wave, R-wave, S-wave, and T-wave are shown in fig. 6.
Calculating the midpoint of a time axis between the T wave and the P wave of the adjacent next cardiac cycle signal, taking the midpoint as a dividing point between different cardiac cycles, taking the dividing point as a starting point, and intercepting the 1-second cardiac signal by using a sliding window, wherein the window length of the sliding window is 1 second, and the step length is one cardiac cycle; the marks of the division points are shown in fig. 7, the signal between two adjacent division points represents an electrocardiosignal of a complete cardiac cycle, and under normal conditions, the time of a complete cardiac cycle is shorter than 1 second, so that the signal of 1 second necessarily comprises at least one complete cardiac cycle signal, and the morphological characteristics of the signal can be obtained by using the electrocardiosignal of the complete cardiac cycle.
And solving time domain characteristics and frequency domain characteristics aiming at the intercepted 1 second electrocardiosignals, determining the positions of two adjacent segmentation points in the 1 second electrocardiosignals, and solving morphological characteristics of the electrocardiosignals between the two adjacent segmentation points.
The time domain features include at least one of an average slope, a median slope, an amplitude range, a signal integral, an average peak-to-peak amplitude, a signal root mean square.
The Frequency domain features include at least one of Amplitude Spectral Area (AMSA), Power Spectral Analysis (PSA), maximum Power (Max Power, MP), Peak Power Frequency (PF), Center Frequency (CF), Energy (EG).
The morphological feature includes at least one of P, Q, R, S, T wave height, P-R interval, Q-T interval, S-T interval, QRS complex width, P wave width.
And performing attribute reduction on the obtained feature set comprising the time-frequency domain and the morphological features by using a rough set, screening features irrelevant to the type of the electrocardiosignal to be judged, outputting effective features, and taking signal data corresponding to the effective features as the input of a neural network and an SVM (support vector machine) to judge the type of the electrocardiosignal.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. An electrocardiosignal feature extraction and classification method is characterized by comprising the following steps:
acquiring an electrocardiosignal, and sequentially positioning R wave, Q wave, S wave, T wave and P wave of the electrocardiosignal according to a preset self-adaptive threshold;
calculating the midpoint of a time axis between the T wave and the P wave of the adjacent next cardiac cycle signal, taking the midpoint as a dividing point between different cardiac cycles, taking the dividing point as a starting point, and intercepting the 1-second cardiac signal by using a sliding window, wherein the window length of the sliding window is 1 second, and the step length is one cardiac cycle;
solving time domain characteristics and frequency domain characteristics of the 1 second electrocardiosignal;
determining the positions of two adjacent segmentation points in the 1 second electrocardiosignal, and solving morphological characteristics of the electrocardiosignal between the two adjacent segmentation points;
and performing attribute reduction on the feature set comprising the time domain feature, the frequency domain feature and the morphological feature by using a rough set, screening features irrelevant to the type of the electrocardiosignal to be judged, outputting effective features, and performing signal type judgment by using signal data corresponding to the effective features as input of a judgment network.
2. The method for extracting and classifying features of an electrocardiographic signal according to claim 1, wherein the obtaining the electrocardiographic signal comprises:
and carrying out 20Hz low-pass filtering and 0.5Hz high-pass filtering on the original electrocardiosignals of the human body to obtain filtered electrocardiosignals.
3. The method for extracting and classifying features of an electrocardiosignal according to claim 1, wherein the positioning of the R wave comprises:
detecting a QRS complex of the electrocardiosignal by adopting a Pan-Tompkins method, and positioning a peak with the amplitude value higher than an R wave threshold value in the QRS complex as an R wave.
4. The method for extracting and classifying features of an electrocardiograph signal according to claim 3, wherein the detecting the QRS complex of the electrocardiograph signal further comprises:
and filtering the electrocardiosignals by using a band-pass filter, and performing square operation on the filtered electrocardiosignals.
5. The method for extracting and classifying features of an electrocardiographic signal according to claim 1, wherein the positioning of the Q-wave and the S-wave comprises:
and taking the R wave as a center, solving two coordinate points which are 15ms away from the R wave, respectively obtaining wave troughs between the two coordinate points and the R wave, and respectively positioning the wave troughs into the Q wave and the S wave according to a Q wave threshold and an S wave threshold.
6. The method for extracting and classifying features of an electrocardiographic signal according to claim 1, wherein the positioning of the T wave comprises:
and defining a first signal segment by taking the S wave as a starting point and the Q wave of the electrocardiosignal of the next cardiac cycle adjacent to the S wave as an end point, setting a T wave threshold value aiming at the first signal segment, and positioning the wave crest of the first signal segment higher than the T wave threshold value as the T wave.
7. The method for extracting and classifying features of an electrocardiographic signal according to claim 6, wherein the positioning of the P-wave comprises:
and defining a second signal segment by taking the T wave as a starting point and the end point of the first signal segment as an end point, setting a P wave threshold value aiming at the second signal segment, and positioning the wave crest of the second signal segment, which is higher than the P wave threshold value, as a P wave.
8. The method for extracting and classifying cardiac signal features according to claim 1, wherein the adaptive threshold is updated by the formula
Thren＝Ampn-1×0.125+Thren-1×0.875，
Thre in the formulanSignal threshold, Amp, representing the nth detectionn-1Represents the signal amplitude, Thre, measured at the (n-1) th timen-1Representing the signal threshold for the (n-1) th detection.
9. An electrocardiosignal feature extraction and classification system is characterized by comprising:
the electrocardiosignal acquisition unit is used for acquiring electrocardiosignals to be processed;
the signal positioning unit is used for positioning P, Q, R, S, T waves of the electrocardiosignals according to the adaptive threshold;
the signal intercepting unit is used for intercepting the positioned electrocardiosignals for 1 second by utilizing the sliding window;
the characteristic extraction unit is used for solving frequency domain, time domain and morphological characteristics of the intercepted 1 second electrocardiosignal;
and the rough set reduction unit is used for screening all the characteristics of the electrocardiosignals to be distinguished, outputting effective characteristics related to the type to be distinguished, and outputting the electrocardiosignal data corresponding to the effective characteristics to a next-stage distinguishing network.
10. The system for extracting and classifying features of an electrocardiographic signal according to claim 9, wherein the signal locating unit further comprises a QRS complex detecting module and a threshold updating module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911060613.3A CN110742599B (en) | 2019-11-01 | 2019-11-01 | Electrocardiosignal feature extraction and classification method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911060613.3A CN110742599B (en) | 2019-11-01 | 2019-11-01 | Electrocardiosignal feature extraction and classification method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110742599A CN110742599A (en) | 2020-02-04 |
| CN110742599B true CN110742599B (en) | 2022-05-10 |
Family
ID=69281857
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911060613.3A Active CN110742599B (en) | 2019-11-01 | 2019-11-01 | Electrocardiosignal feature extraction and classification method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110742599B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113384277B (en) * | 2020-02-26 | 2022-09-20 | 京东方科技集团股份有限公司 | Electrocardiogram data classification method and classification system |
| CN111657918B (en) * | 2020-06-12 | 2021-09-24 | 电子科技大学 | Falling detection method and system combining electrocardio and inertial sensing data |
| CN112426160A (en) * | 2020-11-30 | 2021-03-02 | 贵州省人民医院 | Electrocardiosignal type identification method and device |
| CN115211866B (en) * | 2022-09-07 | 2022-12-27 | 西南民族大学 | Arrhythmia classification method and system and electronic equipment |
| CN115881276B (en) * | 2023-02-22 | 2023-06-09 | 合肥工业大学 | Time-frequency double-bar code characteristic image generation method of electrocardiosignal and storage medium |
| CN116509414B (en) * | 2023-04-14 | 2023-12-15 | 中国科学院大学 | Electrocardiosignal denoising classification system and method |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101919695A (en) * | 2010-08-06 | 2010-12-22 | 李楚雅 | Electrocardiosignal QRS complex detection method based on wavelet transform |
| CN102397064A (en) * | 2011-12-14 | 2012-04-04 | 中国航天员科研训练中心 | Continuous blood pressure measuring device |
| CN103549950A (en) * | 2013-11-19 | 2014-02-05 | 上海理工大学 | Improved difference threshold detection algorithm for mobile ECG (electrocardiogram) monitoring |
| CN104173043A (en) * | 2014-09-04 | 2014-12-03 | 东莞理工学院 | Electrocardiogram (ECG) data analysis method suitable for mobile platform |
| CN105877739A (en) * | 2016-02-25 | 2016-08-24 | 姜坤 | Clinical examination method of electrocardio intelligent analyzing system |
| CN108056773A (en) * | 2017-12-11 | 2018-05-22 | 重庆邮电大学 | Based on the algorithms of QRS complexes detection in electrocardiogram signal for improving variation mode decomposition |
| CN108294745A (en) * | 2018-03-07 | 2018-07-20 | 武汉大学 | P waves, T wave start-stop point detecting methods and system in multi-lead ECG signal |
| WO2018160890A1 (en) * | 2017-03-01 | 2018-09-07 | University Of Washington | Efficient fetal-maternal ecg signal separation from two maternal abdominal leads via diffusion-based channel selection |
| CN108888259A (en) * | 2018-05-21 | 2018-11-27 | 南京大学 | A kind of real-time QRS wave detection method of electrocardiosignal |
| CN109171708A (en) * | 2018-10-25 | 2019-01-11 | 广东工业大学 | One kind can defibrillation rhythm of the heart identification device |
| CN109330584A (en) * | 2018-11-08 | 2019-02-15 | 山东大学 | Electrocardiosignal personal identification method and system based on dictionary learning and rarefaction representation |
| CN109674468A (en) * | 2019-01-30 | 2019-04-26 | 大连理工大学 | It is a kind of singly to lead brain electricity sleep mode automatically method by stages |
| CN109691994A (en) * | 2019-01-31 | 2019-04-30 | 英菲泰克(天津)科技有限公司 | A kind of rhythm of the heart analysis method based on electrocardiogram |
| CN109770892A (en) * | 2019-02-01 | 2019-05-21 | 中国科学院电子学研究所 | A kind of sleep stage method based on electrocardiosignal |
| CN109998523A (en) * | 2019-03-27 | 2019-07-12 | 苏州平稳芯跳医疗科技有限公司 | It is a kind of singly to lead electrocardiosignal classification method and singly lead electrocardiosignal categorizing system |
| CN110327032A (en) * | 2019-03-27 | 2019-10-15 | 苏州平稳芯跳医疗科技有限公司 | It is a kind of singly to lead the accurate recognizer of electrocardiosignal PQRST wave joint |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9008762B2 (en) * | 2009-11-03 | 2015-04-14 | Vivaquant Llc | Method and apparatus for identifying cardiac risk |
| US10194821B2 (en) * | 2014-10-29 | 2019-02-05 | Khalifa University of Science and Technology | Medical device having automated ECG feature extraction |
| US9962102B2 (en) * | 2015-02-18 | 2018-05-08 | Medtronic, Inc. | Method and apparatus for atrial arrhythmia episode detection |
| CN106333671A (en) * | 2016-09-21 | 2017-01-18 | 广东工业大学 | Electrocardiogram detection system |
-
2019
- 2019-11-01 CN CN201911060613.3A patent/CN110742599B/en active Active
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101919695A (en) * | 2010-08-06 | 2010-12-22 | 李楚雅 | Electrocardiosignal QRS complex detection method based on wavelet transform |
| CN102397064A (en) * | 2011-12-14 | 2012-04-04 | 中国航天员科研训练中心 | Continuous blood pressure measuring device |
| CN103549950A (en) * | 2013-11-19 | 2014-02-05 | 上海理工大学 | Improved difference threshold detection algorithm for mobile ECG (electrocardiogram) monitoring |
| CN104173043A (en) * | 2014-09-04 | 2014-12-03 | 东莞理工学院 | Electrocardiogram (ECG) data analysis method suitable for mobile platform |
| CN105877739A (en) * | 2016-02-25 | 2016-08-24 | 姜坤 | Clinical examination method of electrocardio intelligent analyzing system |
| WO2018160890A1 (en) * | 2017-03-01 | 2018-09-07 | University Of Washington | Efficient fetal-maternal ecg signal separation from two maternal abdominal leads via diffusion-based channel selection |
| CN108056773A (en) * | 2017-12-11 | 2018-05-22 | 重庆邮电大学 | Based on the algorithms of QRS complexes detection in electrocardiogram signal for improving variation mode decomposition |
| CN108294745A (en) * | 2018-03-07 | 2018-07-20 | 武汉大学 | P waves, T wave start-stop point detecting methods and system in multi-lead ECG signal |
| CN108888259A (en) * | 2018-05-21 | 2018-11-27 | 南京大学 | A kind of real-time QRS wave detection method of electrocardiosignal |
| CN109171708A (en) * | 2018-10-25 | 2019-01-11 | 广东工业大学 | One kind can defibrillation rhythm of the heart identification device |
| CN109330584A (en) * | 2018-11-08 | 2019-02-15 | 山东大学 | Electrocardiosignal personal identification method and system based on dictionary learning and rarefaction representation |
| CN109674468A (en) * | 2019-01-30 | 2019-04-26 | 大连理工大学 | It is a kind of singly to lead brain electricity sleep mode automatically method by stages |
| CN109691994A (en) * | 2019-01-31 | 2019-04-30 | 英菲泰克(天津)科技有限公司 | A kind of rhythm of the heart analysis method based on electrocardiogram |
| CN109770892A (en) * | 2019-02-01 | 2019-05-21 | 中国科学院电子学研究所 | A kind of sleep stage method based on electrocardiosignal |
| CN109998523A (en) * | 2019-03-27 | 2019-07-12 | 苏州平稳芯跳医疗科技有限公司 | It is a kind of singly to lead electrocardiosignal classification method and singly lead electrocardiosignal categorizing system |
| CN110327032A (en) * | 2019-03-27 | 2019-10-15 | 苏州平稳芯跳医疗科技有限公司 | It is a kind of singly to lead the accurate recognizer of electrocardiosignal PQRST wave joint |
Non-Patent Citations (3)
| Title |
|---|
| "A Human Electrocardiogram Signal Processing Algorithm for Real-Time Control and Monitoring Applications";Karimipour等;《2nd RSI/ISM International Conference on Robotics and Mechatronics》;20141017;第428-433页 * |
| "多导心电算法及应用系统的研制";孙纪光;《中国优秀硕士学位论文全文数据库信息科技辑》;20170715;第I136-70页 * |
| "粗糙集理论辅助现代医疗诊断研究综述";高静;《科技与创新》;20190615;第25-27页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110742599A (en) | 2020-02-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110742599B (en) | Electrocardiosignal feature extraction and classification method and system | |
| CN107041743B (en) | Real-time R wave detection method for electrocardiosignals | |
| CN109907752B (en) | Electrocardiogram diagnosis and monitoring system for removing motion artifact interference and electrocardio characteristic detection | |
| CN101449973B (en) | Judgment index generation method and device for cardiac interference signal identification | |
| CN110013247B (en) | Detection, differentiation and positioning method for electrocardiogram P wave and T wave | |
| CN101828918B (en) | Electrocardiosignal R peak detection method based on waveform characteristic matching | |
| WO2021143401A1 (en) | Method and device for detecting p wave and t wave in electrocardiogram signal | |
| CN105030228A (en) | Method for determining P wave position of electrocardiosignal and device for realizing method | |
| US11253204B2 (en) | Method for assessing electrocardiogram signal quality | |
| US20150342489A1 (en) | Qrs complex identification in electrocardiogram signals | |
| CN111419212A (en) | Method and device for processing electrocardiogram data, storage medium and computer equipment | |
| CN107320096B (en) | Electrocardio R wave positioning method | |
| CN109009087B (en) | Rapid detection method for electrocardiosignal R wave | |
| KR101510522B1 (en) | System for classification electrocardiogram signals | |
| Bayasi et al. | 65-nm ASIC implementation of QRS detector based on Pan and Tompkins algorithm | |
| CN111166325A (en) | Electrocardiosignal QRS complex wave detection method and system based on IPCMM algorithm | |
| Bayasi et al. | A 65-nm low power ECG feature extraction system | |
| CN112438735B (en) | Electrocardiogram P wave detection method, analysis device and storage medium | |
| CN105050493A (en) | Apparatus and method for determining the occurrence of a QRS complex in ECG data | |
| Sovilj et al. | Real time P-wave detector based on wavelet analysis | |
| TWI696191B (en) | Algorithm of qrs detection capable of reducing noise effects | |
| CN114098754A (en) | Atrial fibrillation signal preprocessing method, detection system, equipment and storage medium | |
| CN113397557A (en) | Real-time low-power-consumption heart rhythm monitoring method | |
| TWM574469U (en) | QRS wave instant detection device | |
| CN114159068A (en) | Method and device for detecting atrial flutter of electrocardiosignal |
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 |