CN107358196B - Heart beat type classification method and device and electrocardiograph - Google Patents
Heart beat type classification method and device and electrocardiograph Download PDFInfo
- Publication number
- CN107358196B CN107358196B CN201710564424.4A CN201710564424A CN107358196B CN 107358196 B CN107358196 B CN 107358196B CN 201710564424 A CN201710564424 A CN 201710564424A CN 107358196 B CN107358196 B CN 107358196B
- Authority
- CN
- China
- Prior art keywords
- heart beat
- electrocardiosignals
- waveform
- heartbeat
- template
- 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.)
- Expired - Fee Related
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
- G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
Abstract
The invention discloses a method and a device for classifying heartbeat types and an electrocardiograph, and belongs to the field of medical instruments. The method comprises the following steps: the heart beat waveform characteristic parameters are extracted from the electrocardiosignals, meanwhile, the matching degree characteristic parameters of the heart beats and the template heart beats are extracted, and the decision tree classifier is used for analyzing, so that the type analysis result of the measured heart beats is obtained. The characteristic vector extracted by the method can better show the difference of electrocardiosignals, makes up the defect of singly using a waveform analysis method or a template analysis method, and also comprises the electrocardiograph with the method. The method for extracting the characteristic vector is simple and easy to realize, has high operation speed and high classification accuracy, and is suitable for a wearable automatic analysis and diagnosis system for real-time electrocardiographic monitoring.
Description
Technical Field
The invention relates to the technical field of medical instruments, in particular to a heart beat type classification method and device and an electrocardiograph.
Background
Cardiovascular disease remains a leading cause of death worldwide. The electrocardiogram signal of human body records the process of depolarization and repolarization of heart cells, reflects the process of generation and conduction of electrical excitation of the heart, can objectively reflect the physiological conditions of all parts of the heart to a certain extent, and provides important basis for diagnosis of heart diseases and evaluation of heart functions. In modern medicine, automatic detection and analysis of an electrocardiographic signal (ECG) are of great importance for analysis and diagnosis of cardiovascular diseases, and are the core components of a plurality of electrocardiographic monitoring devices with automatic analysis and diagnosis functions. Therefore, automatic detection and analysis methods based on electrocardiograms have been widely studied.
The existing heart beat type classification methods have certain difference, analysis is carried out according to characteristic waveforms in electrocardiograms, and the main selected characteristic waveform parameters are as follows: the QRS complex wave width, the waveform amplitude, the waveform interval, the waveform form direction and the like, the characteristic waveform parameters are compared with a preset fixed judgment threshold, and the corresponding heart beat type is judged according to the comparison result with the threshold, such as: sinus heartbeat, ventricular premature beat, atrial heartbeat and the like, however, in the method, a fixed threshold is selected as a judgment basis, the selection of the threshold has high dependence on clinical experience and analytical experience, the universality is insufficient, the effective selection of the threshold is difficult, and the heartbeat type with an unobvious difference cannot be effectively judged.
Disclosure of Invention
The invention aims to overcome the defects, and provides a heart beat type classification method which can improve the accuracy of classification, is favorable for judging heart beat types and provides scientific diagnosis basis for clinic, a second aspect of the invention provides a heart beat type analysis method device which is provided with a device for bearing the heart beat type analysis method, a storage and a processor, and an effective carrier is provided for the method, so that the heart beat type can be effectively analyzed more accurately, a third aspect of the invention provides an electrocardiograph, and the electrocardiograph with the heart beat type classification method can ensure that a patient can use the electrocardiograph to detect the electrocardiosignal of the patient at any time, can monitor the change of the electrocardiosignal at any time when sitting at home, can go to a hospital for treatment in time if the electrocardiosignal is abnormal, and has the functions of preventing heart stems, preventing heart beats, and analyzing heart beat types, Coronary heart disease, heart failure, etc.
The embodiment of the invention provides a heart beat type classification method and device and an electrocardiograph. The following presents a simplified summary in order to provide a basic understanding of some aspects of the disclosed embodiments. This summary is not an extensive overview and is intended to neither identify key/critical elements nor delineate the scope of such embodiments. Its sole purpose is to present some concepts in a simplified form as a prelude to the more detailed description that is presented later.
According to a first aspect of the present invention, there is provided a method of classifying a heart beat type, comprising the steps of:
preprocessing input electrocardiosignals to obtain filtered electrocardiosignals, wherein the electrocardiosignals comprise known electrocardiosignals and electrocardiosignals to be detected;
carrying out heart beat positioning on the filtered electrocardiosignals to obtain positioned electrocardiosignals, and extracting the waveform characteristic parameters of heart beats of the positioned electrocardiosignals;
normalizing the positioning electrocardiosignals to obtain normalized electrocardiosignals, and extracting heartbeat data of the normalized electrocardiosignals;
matching the heart beat data of the normalized electrocardiosignal with the template heart beat data, and extracting matching degree characteristic parameters of the heart beat data and the template heart beat data;
inputting the extracted waveform characteristic parameters and matching degree characteristic parameters of heart beats of the positioning electrocardiosignals into a decision tree classifier as characteristic vectors of the heart beats, and training the decision tree classifier to obtain a trained decision tree classifier;
and inputting the waveform characteristic parameters of the heart beat of the normalized electrocardiosignal corresponding to the electrocardiosignal to be detected and the corresponding matching degree characteristic parameters into the trained decision tree classifier by taking the characteristic vectors of the heart beat, and outputting the classification result of the heart beat type.
By the method, especially, the matching degree characteristic parameters are increased, so that the classification method is more accurate.
Preferably, said known types of cardiac electrical signals comprise: normal sinus heartbeat electrocardiosignals, supraventricular premature electrocardiosignals, ventricular premature electrocardiosignals and other unclassified heartbeat electrocardiosignals.
Preferably, the input electrocardiosignals are preprocessed to obtain filtered electrocardiosignals, wherein the preprocessing is preprocessing in a filtering mode, and myoelectric interference in the electrocardiosignals is removed in a low-pass filtering mode; and removing baseline drift and power frequency interference by high-pass filtering.
Preferably, the heart beat positioning of the filtered electrocardiosignal to obtain the positioned electrocardiosignal comprises positioning the center of the waveform position by searching the peak value of the heart beat waveform, and obtaining the positioned electrocardiosignal by taking the positioned heart beat position as the center.
Preferably, the waveform characteristic parameters for extracting heart beat of the positioned electrocardiosignal comprise wave width, R-R interval and waveform morphology.
Preferably, the width, the R-R interval and the waveform form of the extracted wave comprise:
detecting a starting point and an end point of a wave by an algorithm according to the heart beat position of the positioning electrocardiosignal by adopting a method of combining an equipotential line and a slope threshold, and calculating the width of the wave by calculating the difference value between the end point and the starting point;
obtaining an R-R interval between two continuous heart beat waveforms of the positioned electrocardiosignal after calculating and positioning;
the waveform form refers to the main wave direction of the heart beat of the positioned electrocardiosignal, and the parameters of the waveform form are obtained by extracting the main wave direction of the heart beat of the positioned electrocardiosignal.
Preferably, the method for selecting the template heart beat comprises the following steps:
selecting 3 or more than 3 continuous waves from the positioning electrocardiosignals, wherein the widths of the continuous waves are smaller than a certain threshold value;
and extracting R-R interval values of 2 heart beats for comparison, and selecting one heart beat as a template heart beat when the comparison value is smaller than a certain threshold value.
Preferably, the normalization processing is performed on the positioning electrocardiographic signal, and the normalization processing includes:
taking the heart beat waveform of each positioning electrocardiosignal as a reference, and respectively intercepting numerical values with the same length from front to back in the X-axis direction by taking the center of the template heart beat waveform as a reference;
taking the heart beat waveform of each characteristic electrocardiosignal and the same potential of the heart beat of the template as a reference, and taking a potential comparison value which is behind the same potential as the reference in the Y-axis direction;
the amplitude of each characteristic cardiac electrical signal heartbeat waveform is adjusted to the same order of magnitude.
Preferably, the heartbeat data of the normalized electrocardiosignal is matched with the template heartbeat data, and the matching degree characteristic parameters of the normalized electrocardiosignal and the template heartbeat data are extracted, wherein the matching degree characteristic parameters comprise standard deviation, Euclidean distance and cross-correlation value between the heartbeat data of the standard electrocardiosignal and the template heartbeat data.
Preferably, the template heart beat is continuously dynamically updated, the template heart beat selected in the current period is compared with the subsequent heart beats of the same type with the wave width smaller than a certain threshold value selected from the normalized electrocardiosignals, the R-R interval values of the template heart beat and the subsequent heart beats are compared, and when the change between the R-R interval values and the R-R interval values is larger than or equal to 40%, the template heart beat is dynamically updated.
The second aspect of the present invention also employs an apparatus for an electrocardiographic signal classification method, the apparatus comprising:
the device comprises a preprocessing unit, a signal processing unit and a signal processing unit, wherein the preprocessing unit is used for preprocessing input electrocardiosignals to obtain filtered electrocardiosignals, and the electrocardiosignals comprise known electrocardiosignals and electrocardiosignals to be detected;
the characteristic extraction unit is used for carrying out heart beat positioning on the filtered electrocardiosignals to obtain positioned electrocardiosignals and extracting the waveform characteristic parameters of heart beats of the positioned electrocardiosignals;
the normalization unit is used for performing normalization processing on the positioning electrocardiosignals to obtain normalized electrocardiosignals and extracting heartbeat data of the normalized electrocardiosignals;
the matching degree unit is used for matching the heart beat data of the normalized electrocardiosignal with the template heart beat data and extracting matching degree characteristic parameters of the heart beat data and the template heart beat data;
the training unit is used for inputting the extracted waveform characteristic parameters and matching degree characteristic parameters of heart beats of the positioning electrocardiosignals into the decision tree classifier as characteristic vectors of the heart beats, training the decision tree classifier and obtaining the trained decision tree classifier;
and the classification unit is used for inputting the waveform characteristic parameter of the heart beat of the normalized electrocardiosignal corresponding to the electrocardiosignal to be detected and the matching degree characteristic parameter corresponding to the electrocardiosignal to be detected as the characteristic vector of the heart beat into the trained decision tree classifier and outputting a classification storage medium of the classification result of the heart beat type.
The device enables the measuring result to be more accurate.
Preferably, the preprocessing unit is configured to preprocess an input electrocardiographic signal to obtain a filtered electrocardiographic signal, where the electrocardiographic signal includes an electrocardiographic signal of a known type and an electrocardiographic signal to be detected, and preprocess the filtered electrocardiographic signal in a filtering manner, and remove electromyographic interference in the electrocardiographic signal in a low-pass filtering manner; removing baseline drift and power frequency interference through a high-pass filtering mode, wherein the known electrocardiosignals comprise: normal sinus heartbeat electrocardiosignals, supraventricular premature electrocardiosignals, ventricular premature electrocardiosignals and other unclassified heartbeat electrocardiosignals.
Preferably, the feature extraction unit is configured to perform heartbeat location on the filtered electrocardiographic signal to obtain a located electrocardiographic signal, locate a center of a waveform position by finding a heartbeat waveform peak, obtain a located electrocardiographic signal by taking the located heartbeat position as a center, and extract waveform feature parameters of a heartbeat of the located electrocardiographic signal, where the waveform feature parameters include a width of a wave, an R-R interval, and a waveform form, and includes:
detecting a starting point and an end point of a wave by an algorithm according to the heart beat position of the positioning electrocardiosignal by adopting a method of combining an equipotential line and a slope threshold, and calculating the width of the wave by calculating the difference value between the end point and the starting point;
obtaining an R-R interval between two continuous heart beat waveforms of the positioned electrocardiosignal after calculating and positioning;
the waveform form refers to the main wave direction of the heart beat of the positioned electrocardiosignal, and the parameters of the waveform form are obtained by extracting the main wave direction of the heart beat of the positioned electrocardiosignal.
Preferably, the system is characterized by further comprising a template heart beat unit for selecting a template heart beat, wherein the template heart beat unit comprises:
selecting 3 or more than 3 continuous waves from the positioning electrocardiosignals, wherein the widths of the continuous waves are smaller than a certain threshold value;
and extracting R-R interval values of 2 heart beats for comparison, and selecting one heart beat as a template heart beat when the comparison value is smaller than a certain threshold value.
Preferably, the normalization unit is configured to perform normalization processing on the positioning electrocardiographic signal to obtain a normalized electrocardiographic signal, and extract heartbeat data of the normalized electrocardiographic signal; the normalization process includes:
taking the heart beat waveform of each positioning electrocardiosignal as a reference, and respectively intercepting numerical values with the same length from front to back in the X-axis direction by taking the center of the template heart beat waveform as a reference;
taking the heart beat waveform of each characteristic electrocardiosignal and the same potential of the heart beat of the template as a reference, and taking a potential comparison value which is behind the same potential as the reference in the Y-axis direction;
the amplitude of each characteristic cardiac electrical signal heartbeat waveform is adjusted to the same order of magnitude.
Preferably, the matching degree unit is configured to match heartbeat data of the normalized cardiac signal with template heartbeat data, and extract matching degree characteristic parameters of the normalized cardiac signal and the template heartbeat data; the matching degree characteristic parameters comprise standard deviation, Euclidean distance and cross-correlation value between heart beat data of standard electrocardiosignals and template heart beat data.
The third aspect of the invention also uses an electrocardiograph, which comprises the heart beat type classification method and the device.
The technical scheme provided by the embodiment of the invention has the following beneficial effects:
the method combines the waveform characteristics and the template matching degree characteristics of the electrocardiosignals, and the combination of the characteristics can more comprehensively and more fully reflect the difference of the electrocardiosignals of different heart beat types, so that the electrocardiosignals of different heart beat types can be more effectively detected, and meanwhile, the decision tree classifier is used for classifying sinus heart beats, supraventricular heart beats and ventricular heart beats, so that the generated decision tree classifier model has simple rules, is easy to understand and realize, and improves the classification speed and accuracy.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
FIG. 1 is a first flow diagram illustrating a method for classifying heart beat types according to an exemplary embodiment;
FIG. 2 is a flow diagram illustrating a method for classifying heart beat types in accordance with an exemplary embodiment;
FIG. 3 is a schematic diagram of an apparatus illustrating a method of classifying cardiac electrical signals according to an exemplary embodiment;
FIG. 4 is a flow diagram illustrating a classification method of a decision tree classifier in a classification method of heart beat types according to an exemplary embodiment.
Detailed Description
The following description and the drawings sufficiently illustrate specific embodiments of the invention to enable those skilled in the art to practice them. The examples merely typify possible variations. Individual components and functions are optional unless explicitly required, and the sequence of operations may vary. Portions and features of some embodiments may be included in or substituted for those of others. The scope of embodiments of the invention encompasses the full ambit of the claims, as well as all available equivalents of the claims. Embodiments may be referred to herein, individually or collectively, by the term "invention" merely for convenience and without intending to voluntarily limit the scope of this application to any single invention or inventive concept if more than one is in fact disclosed. Herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed. The embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the structures, products and the like disclosed by the embodiments, the description is relatively simple because the structures, the products and the like correspond to the parts disclosed by the embodiments, and the relevant parts can be just described by referring to the method part.
As shown in the flow chart of fig. 1, the invention discloses a heart beat type classification method, which comprises the following steps:
s101: preprocessing input electrocardiosignals, wherein the electrocardiosignals comprise known electrocardiosignals and electrocardiosignals to be detected;
s102: carrying out heart beat positioning on the preprocessed electrocardiosignals to obtain characteristic electrocardiosignals, and extracting waveform characteristic parameters of heart beats of the characteristic electrocardiosignals;
s103: normalizing the characteristic electrocardiosignals to obtain standard electrocardiosignals, and extracting heart beats of the standard electrocardiosignals;
s104: matching heart beats of the standard electrocardiosignals with template heart beats, and extracting matching degree characteristic parameters of the heart beats and the template heart beats;
s105: inputting waveform characteristic parameters and matching degree characteristic parameters of heart beats of the extracted characteristic electrocardiosignals into a decision tree classifier as characteristic vectors of the heart beats, and training the decision tree classifier to obtain a trained decision tree classifier;
s106: and inputting the waveform characteristic parameters of the heart beat of the characteristic electrocardiosignal corresponding to the electrocardiosignal to be detected and the corresponding matching degree characteristic parameters serving as the characteristic vectors of the heart beat into the trained decision tree classifier, and outputting the classification result of the heart beat type.
The invention realizes that heart beats are divided into 4 types according to different excitation origins of electrocardio: normal sinus heartbeat (N or NSR), supraventricular Premature beat or abnormal (S), ventricular Premature beat (V), unclassified heartbeat (Q), i.e. the heartbeat types are divided into four different types, i.e. N type, S type, V type and Q type, and the known type of electrocardiosignals input in the above-mentioned method are the different types of electrocardiosignals in the above-mentioned 4.
As shown in fig. 2, according to the above method for classifying heartbeat types, the preprocessing of the input known cardiac electric signal is performed by filtering, and the step S101 specifically includes:
s1011: myoelectric interference in the electrocardiosignal is removed through low-pass filtering;
s1012: and removing baseline drift through high-pass filtering and removing power frequency interference through power frequency filtering.
The heart beat positioning is carried out on the preprocessed electrocardiosignals, the center of the waveform position is positioned by searching the peak value of the heart beat waveform, and the data segment of the QRS complex wave is extracted by taking the positioned heart beat position as the center, so that the heart beats can be more accurately positioned, the extracted data is more accurate, and the characteristic electrocardiosignals are obtained.
Extracting waveform characteristic parameters of heart beats of the characteristic electrocardiosignals, and extracting three parameters of the width, the R-R interval and the waveform form of a QRS complex as the characteristic parameters of a waveform; and extracting waveform characteristics from the QRS complex data segment.
The method for extracting and calculating the waveform characteristic parameters of heart beats of the characteristic electrocardiosignals comprises the following steps:
1. width of extracted wave
Because both normal and supraventricular beats have a normally-conducted QRS wave morphology, and the QRS waveform of a ventricular beat is widely malformed, the width of the QRS complex can be used as a judgment index.
The starting point and the end point of the QRS complex wave are detected through an algorithm, and the difference value is the width of the QRS.
The algorithm for measuring the width of the QRS complex comprises: the method of combining the equipotential line and the slope threshold value is adopted, so that the starting point and the starting point of the QRS can be accurately positioned.
2. Extraction of R-R intervals
Since supraventricular premature beats or abnormalities and ventricular premature beats occur most early in time compared to normal sinus beats, the R-R interval between beats may also be used as a criterion.
From the located heart beat positions, the R-R interval between every two consecutive heart beats can be found.
3. Extracting waveform morphology
Because the electrocardiographic waveforms have different forms, it is most obvious that the main wave has positive and negative components. When the dominant wave shape of one waveform and the dominant wave shape of another waveform in the same sequence are obviously different, the two waveforms can be judged to be of two types. The width of the QRS complex is characterized by the waveform of the currently detected heart beat, the R-R interval is characterized by the cycle transition relationship between adjacent heart beats, and the waveform morphology is characterized by the difference between the current heart beat and other types of heart beats.
Further, in order to extract matching parameters between heart beats of the characteristic electrocardiosignal and template heart beats, other differences between waveforms of the heart beats and the template heart beats need to be eliminated to the maximum extent, so that normalization processing is carried out on each heart beat sequence, and the normalization processing comprises the following steps:
1. aligning data on the X-axis
The heart beat position of the characteristic electrocardiosignal positioned by each heart beat is taken as an alignment standard on an X axis, and data with the same length are intercepted in front and at the back.
2. Base line of unified Y axis
Due to external disturbances or other factors, the equipotential lines of the heart beat for each characteristic cardiac signal are not on the same horizontal line, which greatly aggravates the difference between the waveforms. When the Y-axis data is normalized, the data in the heartbeat of each characteristic electrocardiosignal is a comparison value based on the equipotential line.
3. Order of magnitude of uniform QRS complex amplitude
When the electrocardio signal is collected, the amplitude of the waveform can fluctuate greatly due to the interference of respiration or other factors, so that the waveform can have a large difference, the essence of the difference is not the morphological difference of the electrocardio waveform, and the classification of heart beats is not facilitated. When amplitude unification is performed, the amplitude of the heart beat of each characteristic cardiac electrical signal is unified to the same order of magnitude.
Carrying out normalization processing on the characteristic electrocardiosignals to obtain standard electrocardiosignals, and extracting heart beats of the standard electrocardiosignals;
in order to calculate the matching degree characteristic parameters between heartbeats of standard electrocardiosignals and template heartbeats, the template heartbeats need to be selected firstly, and the selection of the template heartbeats comprises the following steps:
selecting 3 or more than 3 continuous heart beats with QRS width less than 100 ms-130 ms from standard electrocardio signals, comparing 2R-R interval values when detecting that the heart beat type is one of the 4 known types, and selecting one heart beat as a template heart beat when the change between the two is less than 40%.
The template heart beat is not always unchanged, and the template heart beat needs to be updated in real time in order to improve the accuracy of the classification of the heart beat type, and the updating method of the template heart beat comprises the following steps: when the 2R-R interval values are compared, and the change between the two values is more than or equal to 40%, the template heart rate is dynamically updated.
Further, the main parameters of the matching degree characteristic parameters and the calculation method thereof comprise the following steps:
1. standard deviation of
The standard deviation represents the degree of dispersion of a data set, where the standard deviation of the difference between the measured heart beat and the template heart beat data, i.e., the degree of dispersion of the difference between the two heart beats, is calculated, thus, the difference between the two waveforms is quantified.
Template matching is performed by the following calculation formula, and the standard deviation of the difference between the measurement heartbeat and the template heartbeat is calculated. The formula is as follows:
wherein x isiIs data of a test heart beat, yiIs data of a template heart beat, and n is the length of the heart beat data.
2. Euclidean distance
Euclidean distance is the most common distance metric, measuring the absolute distance between points in space, with greater distances indicating greater differences between individuals. The calculation formula of the euclidean distance is as follows:
wherein x isiIs data of a test heart beat, yiIs data of a template heart beat, and n is the length of the heart beat data.
3. Cross correlation coefficient
The cross-correlation coefficient is a statistical index used to measure the closeness of the correlation between two variables. The value range of the cross correlation coefficient r is [ -1,1], r >0 represents positive correlation, r <0 represents negative correlation, and | r | represents the degree of correlation between variables. Specifically, r ═ 1 is referred to as a complete positive correlation, r ═ 1 is referred to as a complete negative correlation, and r ═ 0 is referred to as an uncorrelated. Generally, | r | is greater than 0.8, the two variables are considered to have strong linear correlation.
Wherein x isiIs data of a test heart beat, yiIs data of a template heart beat, and n is the length of the heart beat data.
Inputting waveform characteristic parameters and matching degree characteristic parameters of heart beats of the extracted characteristic electrocardiosignals into a decision tree classifier as characteristic vectors of the heart beats, and training the decision tree classifier to obtain a trained decision tree classifier;
the decision tree classifier shown in fig. 4 is a tree-like decision diagram showing a mapping between object feature attributes and object classes, each non-leaf node representing a test on a feature attribute, each branch representing the output of the feature attribute over a range of values, and each leaf node storing a heart beat classification class. The process of using the decision tree to carry out decision classification is to test corresponding characteristic attributes in items to be classified from a root node, select an output branch according to the value of the characteristic attributes until a leaf node is reached, and use the category stored by the leaf node as a heart beat classification result.
And inputting the waveform characteristic parameters of the heart beat of the characteristic electrocardiosignal corresponding to the electrocardiosignal to be detected and the corresponding matching degree characteristic parameters into the trained decision tree classifier as the characteristic vectors of the heart beat, and outputting the classification result of the heart beat type, namely judging that the heart beat to be detected is one of 4 types of N type, S type, V type and Q type.
The decision tree classifier is trained through multiple groups of data, the decision tree classifier with accurate judgment is trained, the electrocardiosignal to be detected is input into the trained decision tree classifier, and the heart beat type can be judged through the program.
According to a second aspect of the embodiments of the present invention, there is provided an apparatus for an electrocardiograph signal classification method, as shown in fig. 3, the apparatus includes:
the preprocessing unit 201 is configured to preprocess an input electrocardiographic signal to obtain a filtered electrocardiographic signal, where the electrocardiographic signal includes an electrocardiographic signal of a known type and an electrocardiographic signal to be detected;
a feature extraction unit 202 for performing heart beat positioning on the filtered electrocardiosignals to obtain positioned electrocardiosignals
A normalization unit 203, configured to perform normalization processing on the positioning electrocardiographic signal to obtain a normalized electrocardiographic signal, and extract heartbeat data of the normalized electrocardiographic signal;
a matching degree unit 204, configured to match the heartbeat data of the normalized cardiac electrical signal with the template heartbeat data, and extract matching degree characteristic parameters of the two;
the training unit 205 is configured to input the extracted waveform characteristic parameters of heart beats of the cardiac electrical signals and the matching degree characteristic parameters into a decision tree classifier as feature vectors of heart beats, train the decision tree classifier, and obtain a trained decision tree classifier;
and the classification unit 206 is a classification storage medium, and is configured to input the waveform characteristic parameter of the heartbeat of the normalized electrocardiograph signal corresponding to the electrocardiograph signal to be detected and the corresponding matching degree characteristic parameter as a heartbeat characteristic vector into the trained decision tree classifier, and output a classification result of the heartbeat type.
Further, the preprocessing unit 201 is configured to preprocess the input electrocardiographic signal to obtain a filtered electrocardiographic signal, where the electrocardiographic signal includes a known type of electrocardiographic signal and an electrocardiographic signal to be detected, preprocess the filtered electrocardiographic signal in a filtering manner, and remove myoelectric interference in the electrocardiographic signal in a low-pass filtering manner; removing baseline drift and power frequency interference through a high-pass filtering mode, wherein the known electrocardiosignals comprise: normal sinus heartbeat electrocardiosignals, supraventricular premature electrocardiosignals, ventricular premature electrocardiosignals and other unclassified heartbeat electrocardiosignals.
Further, the feature extraction unit 202 is configured to perform heartbeat location on the filtered electrocardiographic signal to obtain a located electrocardiographic signal, locate a center of a waveform position by finding a heartbeat waveform peak, obtain a located electrocardiographic signal by taking the located heartbeat position as a center, and extract waveform feature parameters of a heartbeat of the located electrocardiographic signal, where the waveform feature parameters include a width of a wave, an R-R interval, and a waveform form, and includes:
detecting a starting point and an end point of a wave by an algorithm according to the heart beat position of the positioning electrocardiosignal by adopting a method of combining an equipotential line and a slope threshold, and calculating the width of the wave by calculating the difference value between the end point and the starting point;
obtaining an R-R interval between two continuous heart beat waveforms of the positioned electrocardiosignal after calculating and positioning;
the waveform form refers to the main wave direction of the heart beat of the positioned electrocardiosignal, and the parameters of the waveform form are obtained by extracting the main wave direction of the heart beat of the positioned electrocardiosignal.
On the basis, the system further comprises a template heart beat unit 203, which is used for selecting a template heart beat and comprises: selecting 3 or more than 3 continuous waves from the positioning electrocardiosignals, wherein the widths of the continuous waves are smaller than a certain threshold value;
and extracting R-R interval values of 2 heart beats for comparison, and selecting one heart beat as a template heart beat when the comparison value is smaller than a certain threshold value.
Further, the normalization unit 204 is configured to perform normalization processing on the positioning electrocardiographic signal to obtain a normalized electrocardiographic signal, and extract heartbeat data of the normalized electrocardiographic signal; the normalization process includes:
taking the heart beat waveform of each positioning electrocardiosignal as a reference, and respectively intercepting numerical values with the same length from front to back in the X-axis direction by taking the center of the template heart beat waveform as a reference;
taking the heart beat waveform of each characteristic electrocardiosignal and the same potential of the heart beat of the template as a reference, and taking a potential comparison value which is behind the same potential as the reference in the Y-axis direction;
the amplitude of each characteristic cardiac electrical signal heartbeat waveform is adjusted to the same order of magnitude.
Further, the matching degree unit 205 is configured to match the heartbeat data of the normalized cardiac electrical signal with the template heartbeat data, and extract matching degree feature parameters of the heartbeat data and the template heartbeat data; the matching degree characteristic parameters comprise standard deviation, Euclidean distance and cross-correlation value between heart beat data of standard electrocardiosignals and template heart beat data.
According to a third aspect of the embodiments of the present invention, there is provided an electrocardiograph including the above-described heart beat type classification method.
Further, on the basis, the electrocardiograph further comprises the device of the electrocardiosignal classification method.
It is to be understood that the present invention is not limited to the procedures and structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the invention is limited only by the appended claims.
Claims (17)
1. A method of classifying a heart beat type, comprising the steps of:
preprocessing input electrocardiosignals to obtain filtered electrocardiosignals, wherein the electrocardiosignals comprise known electrocardiosignals and electrocardiosignals to be detected;
carrying out heart beat positioning on the filtered electrocardiosignals to obtain positioned electrocardiosignals, and extracting the waveform characteristic parameters of heart beats of the positioned electrocardiosignals;
normalizing the positioning electrocardiosignals to obtain normalized electrocardiosignals, and extracting heartbeat data of the normalized electrocardiosignals;
matching the heart beat data of the normalized electrocardiosignal with the template heart beat data, and extracting matching degree characteristic parameters of the heart beat data and the template heart beat data;
inputting the extracted waveform characteristic parameters and matching degree characteristic parameters of heart beats of the positioning electrocardiosignals into a decision tree classifier as characteristic vectors of the heart beats, and training the decision tree classifier to obtain a trained decision tree classifier;
and inputting the waveform characteristic parameters of the heart beat of the normalized electrocardiosignal corresponding to the electrocardiosignal to be detected and the corresponding matching degree characteristic parameters into the trained decision tree classifier by taking the characteristic vectors of the heart beat, and outputting the classification result of the heart beat type.
2. The method for classifying heart beat types according to claim 1, wherein the known types of cardiac electrical signals comprise: normal sinus heartbeat electrocardiosignals, supraventricular premature electrocardiosignals, ventricular premature electrocardiosignals and other unclassified heartbeat electrocardiosignals.
3. The heart beat type classification method according to claim 1, wherein the input electrocardiosignals are preprocessed to obtain filtered electrocardiosignals, the preprocessing is performed in a filtering mode, and myoelectric interference in the electrocardiosignals is removed in a low-pass filtering mode; and removing baseline drift and power frequency interference by high-pass filtering.
4. The method for classifying heartbeat types as claimed in claim 1, wherein the heart beat locating of the filtered electrocardiosignal to obtain the located electrocardiosignal comprises locating the center of the waveform position by searching the peak value of the heart beat waveform and obtaining the located electrocardiosignal by taking the located heart beat position as the center.
5. The method for classifying heart beat types according to claim 1 or 4, wherein the waveform characteristic parameters for extracting heart beats for locating electrocardiosignals comprise wave width, R-R interval and waveform morphology.
6. The method for classifying heartbeat types as in claim 5, wherein said extracting wave width, R-R interval, waveform morphology comprises:
detecting a starting point and an end point of a wave by an algorithm according to the heart beat position of the positioning electrocardiosignal by adopting a method of combining an equipotential line and a slope threshold, and calculating the width of the wave by calculating the difference value between the end point and the starting point;
obtaining an R-R interval between two continuous heart beat waveforms of the positioned electrocardiosignal after calculating and positioning;
the waveform form refers to the main wave direction of the heart beat of the positioned electrocardiosignal, and the parameters of the waveform form are obtained by extracting the main wave direction of the heart beat of the positioned electrocardiosignal.
7. The method for classifying heartbeat types according to claim 1, wherein the method for selecting the template heartbeat comprises:
selecting 3 or more than 3 continuous waves from the positioning electrocardiosignals, wherein the widths of the continuous waves are smaller than a certain threshold value;
and extracting R-R interval values of 2 heart beats for comparison, and selecting one heart beat as a template heart beat when the comparison value is smaller than a certain threshold value.
8. The method for classifying a heart beat type according to claim 7, wherein the localization cardiac signal is normalized, and the normalization process includes:
taking the heart beat waveform of each positioning electrocardiosignal as a reference, and respectively intercepting numerical values with the same length from front to back in the X-axis direction by taking the center of the template heart beat waveform as a reference;
taking the heart beat waveform of each characteristic electrocardiosignal and the same potential of the heart beat of the template as a reference, and taking a potential comparison value which is behind the same potential as the reference in the Y-axis direction;
the amplitude of each characteristic cardiac electrical signal heartbeat waveform is adjusted to the same order of magnitude.
9. The method for classifying heartbeat types according to claim 1, wherein heartbeat data of the normalized electrocardiosignal is matched with template heartbeat data, and a matching degree characteristic parameter of the normalized electrocardiosignal and the template heartbeat data is extracted, wherein the matching degree characteristic parameter comprises standard deviation, Euclidean distance and cross-correlation coefficient values between the heartbeat data of the standard electrocardiosignal and the template heartbeat data.
10. The method for classifying heartbeat types as in claim 7, wherein the template heartbeats are dynamically updated continuously, the selected template heartbeat in the current period is compared with a subsequent heartbeat of the same type with the wave width of the selected wave from the normalized electrocardiosignals smaller than a certain threshold value, the R-R interval values of the template heartbeat and the subsequent heartbeat of the same type are compared, and the template heartbeat is dynamically updated when the change between the template heartbeat and the subsequent heartbeat of the same type is larger than or equal to 40%.
11. An apparatus for a method of classifying an electrocardiographic signal, the apparatus comprising:
the device comprises a preprocessing unit, a signal processing unit and a signal processing unit, wherein the preprocessing unit is used for preprocessing input electrocardiosignals to obtain filtered electrocardiosignals, and the electrocardiosignals comprise known electrocardiosignals and electrocardiosignals to be detected;
the characteristic extraction unit is used for carrying out heart beat positioning on the filtered electrocardiosignals to obtain positioned electrocardiosignals and extracting the waveform characteristic parameters of heart beats of the positioned electrocardiosignals;
the normalization unit is used for performing normalization processing on the positioning electrocardiosignals to obtain normalized electrocardiosignals and extracting heartbeat data of the normalized electrocardiosignals;
the matching degree unit is used for matching the heart beat data of the normalized electrocardiosignal with the template heart beat data and extracting matching degree characteristic parameters of the heart beat data and the template heart beat data;
the training unit is used for inputting the extracted waveform characteristic parameters and matching degree characteristic parameters of heart beats of the positioning electrocardiosignals into the decision tree classifier as characteristic vectors of the heart beats, training the decision tree classifier and obtaining the trained decision tree classifier;
and the classification unit is used for inputting the waveform characteristic parameter of the heart beat of the normalized electrocardiosignal corresponding to the electrocardiosignal to be detected and the matching degree characteristic parameter corresponding to the electrocardiosignal to be detected as the characteristic vector of the heart beat into the trained decision tree classifier and outputting a classification storage medium of the classification result of the heart beat type.
12. The apparatus for classifying an electrocardiographic signal according to claim 11, wherein the preprocessing unit is configured to preprocess an input electrocardiographic signal to obtain a filtered electrocardiographic signal, the electrocardiographic signal includes an electrocardiographic signal of a known type and an electrocardiographic signal to be detected, the preprocessing is performed by filtering, and electromyographic interference in the electrocardiographic signal is removed by low-pass filtering; removing baseline drift and power frequency interference through a high-pass filtering mode, wherein the known electrocardiosignals comprise: normal sinus heartbeat electrocardiosignals, supraventricular premature electrocardiosignals, ventricular premature electrocardiosignals and other unclassified heartbeat electrocardiosignals.
13. The apparatus for classifying an electrocardiographic signal according to claim 11, wherein the feature extraction unit is configured to perform heartbeat location on the filtered electrocardiographic signal to obtain a located electrocardiographic signal, locate a center of a waveform position by finding a heartbeat waveform peak, obtain a located electrocardiographic signal by taking the located heartbeat position as a center, and extract waveform feature parameters of heartbeats of the located electrocardiographic signal, the waveform feature parameters including a width of a wave, an R-R interval, and a waveform shape, and the apparatus includes:
detecting a starting point and an end point of a wave by an algorithm according to the heart beat position of the positioning electrocardiosignal by adopting a method of combining an equipotential line and a slope threshold, and calculating the width of the wave by calculating the difference value between the end point and the starting point;
obtaining an R-R interval between two continuous heart beat waveforms of the positioned electrocardiosignal after calculating and positioning;
the waveform form refers to the main wave direction of the heart beat of the positioned electrocardiosignal, and the parameters of the waveform form are obtained by extracting the main wave direction of the heart beat of the positioned electrocardiosignal.
14. The apparatus for classifying an electrocardiographic signal according to claim 11, further comprising a template heart beat unit for selecting a template heart beat, comprising:
selecting 3 or more than 3 continuous waves from the positioning electrocardiosignals, wherein the widths of the continuous waves are smaller than a certain threshold value;
and extracting R-R interval values of 2 heart beats for comparison, and selecting one heart beat as a template heart beat when the comparison value is smaller than a certain threshold value.
15. The apparatus for classifying an electrocardiographic signal according to claim 11, wherein the normalization unit is configured to perform normalization processing on the positioned electrocardiographic signal to obtain a normalized electrocardiographic signal, and extract heartbeat data of the normalized electrocardiographic signal; the normalization process includes:
taking the heart beat waveform of each positioning electrocardiosignal as a reference, and respectively intercepting numerical values with the same length from front to back in the X-axis direction by taking the center of the template heart beat waveform as a reference;
taking the heart beat waveform of each characteristic electrocardiosignal and the same potential of the heart beat of the template as a reference, and taking a potential comparison value which is behind the same potential as the reference in the Y-axis direction;
the amplitude of each characteristic cardiac electrical signal heartbeat waveform is adjusted to the same order of magnitude.
16. The apparatus of the electrocardiographic signal classifying method according to claim 11, wherein the matching degree unit is configured to match heartbeat data of the normalized electrocardiographic signal with template heartbeat data, and extract matching degree feature parameters of the heartbeat data and the template heartbeat data; the matching degree characteristic parameters comprise standard deviation, Euclidean distance and cross-correlation value between heart beat data of standard electrocardiosignals and template heart beat data.
17. An electrocardiograph comprising an apparatus capable of implementing the heart beat type classifying method according to any one of claims 1 to 10 and further comprising the electrocardiograph according to any one of claims 11 to 16.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710564424.4A CN107358196B (en) | 2017-07-12 | 2017-07-12 | Heart beat type classification method and device and electrocardiograph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710564424.4A CN107358196B (en) | 2017-07-12 | 2017-07-12 | Heart beat type classification method and device and electrocardiograph |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107358196A CN107358196A (en) | 2017-11-17 |
| CN107358196B true CN107358196B (en) | 2020-11-10 |
Family
ID=60292566
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710564424.4A Expired - Fee Related CN107358196B (en) | 2017-07-12 | 2017-07-12 | Heart beat type classification method and device and electrocardiograph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107358196B (en) |
Families Citing this family (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107837082B (en) | 2017-11-27 | 2020-04-24 | 乐普（北京）医疗器械股份有限公司 | Automatic electrocardiogram analysis method and device based on artificial intelligence self-learning |
| CN107714023B (en) * | 2017-11-27 | 2020-09-01 | 上海优加利健康管理有限公司 | Static electrocardiogram analysis method and device based on artificial intelligence self-learning |
| CN107736888A (en) * | 2017-11-27 | 2018-02-27 | 中山大学 | A kind of method and system that real-time cardiac monitoring is carried out using wearable ecg equipment |
| CN107951485B (en) | 2017-11-27 | 2019-06-11 | 深圳市凯沃尔电子有限公司 | Ambulatory ECG analysis method and apparatus based on artificial intelligence self study |
| CN107981858B (en) * | 2017-11-27 | 2020-12-01 | 上海优加利健康管理有限公司 | Automatic electrocardiogram heart beat identification and classification method based on artificial intelligence |
| CN107832737B (en) | 2017-11-27 | 2021-02-05 | 上海优加利健康管理有限公司 | Artificial intelligence-based electrocardiogram interference identification method |
| CN109948396B (en) * | 2017-12-20 | 2021-07-23 | 深圳市理邦精密仪器股份有限公司 | Heart beat classification method, heart beat classification device and electronic equipment |
| CN108491769A (en) * | 2018-03-08 | 2018-09-04 | 四川大学 | Atrial fibrillation sorting technique based on phase between RR and multiple characteristic values |
| CN109077719A (en) * | 2018-07-05 | 2018-12-25 | 广州视源电子科技股份有限公司 | Signal recognition method, device, equipment and storage medium |
| CN108903935B (en) * | 2018-07-11 | 2021-06-25 | 上海夏先机电科技发展有限公司 | Ventricular premature beat identification method and system and electronic equipment |
| CN108937915B (en) * | 2018-07-24 | 2021-02-19 | 杭州质子科技有限公司 | Method for identifying premature beat in dynamic electrocardiogram |
| CN109431488B (en) * | 2018-09-28 | 2021-03-19 | 上海优加利健康管理有限公司 | Method for identifying typical data in electrocardiogram dynamic real-time analysis data |
| CN109303559B (en) * | 2018-11-01 | 2021-05-14 | 杭州质子科技有限公司 | Dynamic electrocardiogram and heartbeat classification method based on gradient boosting decision tree |
| CN109303561A (en) * | 2018-11-01 | 2019-02-05 | 杭州质子科技有限公司 | It is a kind of to clap the recognition methods clapped with the abnormal heart based on the artifact heart of misclassification and supervised learning |
| CN109871742B (en) * | 2018-12-29 | 2021-05-14 | 安徽心之声医疗科技有限公司 | Electrocardiosignal positioning method based on attention circulation neural network |
| CN111493856B (en) * | 2019-01-31 | 2023-05-09 | 深圳市理邦精密仪器股份有限公司 | Correction method and device for heart beat classification, electrocardiogram checking system and storage medium |
| CN109691995A (en) * | 2019-01-31 | 2019-04-30 | 英菲泰克(天津)科技有限公司 | The arrhythmia cordis real-time data analysis system and method for single lead electrocardiogram |
| CN111053551B (en) * | 2019-12-27 | 2021-09-03 | 深圳邦健生物医疗设备股份有限公司 | RR interval electrocardio data distribution display method, device, computer equipment and medium |
| CN111150387B (en) * | 2020-01-15 | 2021-06-08 | 深圳市邦健科技有限公司 | Electrocardiogram template matching method and device, computer equipment and storage medium |
| CN113712567B (en) * | 2020-05-12 | 2023-09-01 | 深圳市科瑞康实业有限公司 | Method and device for generating heart beat interval difference value data sequence coefficient |
| CN111616696A (en) * | 2020-05-20 | 2020-09-04 | 联想(北京)有限公司 | Electrocardiosignal detection method and device and storage medium |
| CN113712525A (en) * | 2020-05-21 | 2021-11-30 | 深圳市理邦精密仪器股份有限公司 | Physiological parameter processing method and device and medical equipment |
| CN113768511B (en) * | 2020-06-04 | 2023-09-22 | 深圳市理邦精密仪器股份有限公司 | Physiological parameter detection method and electronic equipment |
| CN111772619B (en) * | 2020-06-19 | 2022-09-20 | 厦门纳龙健康科技股份有限公司 | Heart beat identification method based on deep learning, terminal equipment and storage medium |
| CN111956203A (en) * | 2020-08-18 | 2020-11-20 | 广州视源电子科技股份有限公司 | Electrocardiosignal parameterization method, model training method, device, equipment and medium |
| CN114271829B (en) * | 2020-09-18 | 2023-07-25 | 深圳邦健生物医疗设备股份有限公司 | Method, device, equipment and medium for analyzing heart beat double marks |
| CN112528783B (en) * | 2020-11-30 | 2024-04-16 | 深圳邦健生物医疗设备股份有限公司 | Electrocardiogram and heart beat data clustering method, device, electronic equipment and medium |
| KR102653258B1 (en) * | 2022-02-18 | 2024-04-01 | 주식회사 뷰노 | Ecg delineation and inetegration method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101856225A (en) * | 2010-06-30 | 2010-10-13 | 重庆大学 | Method for detecting R wave crest of electrocardiosignal |
| CN102908135A (en) * | 2012-10-08 | 2013-02-06 | 中国科学院深圳先进技术研究院 | ECG diagnosis system and operating method of ECG diagnosis system |
| CN106778685A (en) * | 2017-01-12 | 2017-05-31 | 司马大大（北京）智能***有限公司 | Electrocardiogram image-recognizing method, device and service terminal |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8232866B2 (en) * | 2006-04-05 | 2012-07-31 | California Institute Of Technology | Systems and methods for remote long standoff biometric identification using microwave cardiac signals |
| US20080002775A1 (en) * | 2006-06-30 | 2008-01-03 | Ricci Carlos A | Signal analysis employing matched wavelet |
-
2017
- 2017-07-12 CN CN201710564424.4A patent/CN107358196B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101856225A (en) * | 2010-06-30 | 2010-10-13 | 重庆大学 | Method for detecting R wave crest of electrocardiosignal |
| CN102908135A (en) * | 2012-10-08 | 2013-02-06 | 中国科学院深圳先进技术研究院 | ECG diagnosis system and operating method of ECG diagnosis system |
| CN106778685A (en) * | 2017-01-12 | 2017-05-31 | 司马大大（北京）智能***有限公司 | Electrocardiogram image-recognizing method, device and service terminal |
Non-Patent Citations (2)
| Title |
|---|
| Classification of supraventricular and ventricular beats by QRS template matching and decision tree;Vessela Krasteva et al.;《Computing in Cardiology 2014》;20140910;1-4页 * |
| 基于特征匹配的心电数据分类算法的研究与应用;季春林;《中国优秀硕士学位论文全文数据库医药卫生科技辑》;20160315(第03期);E062-45 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107358196A (en) | 2017-11-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107358196B (en) | Heart beat type classification method and device and electrocardiograph | |
| CN107714023B (en) | Static electrocardiogram analysis method and device based on artificial intelligence self-learning | |
| CN106214145B (en) | Electrocardiogram classification method based on deep learning algorithm | |
| CN109411041B (en) | Electrocardio information processing method and electrocardio workstation system | |
| US7769434B2 (en) | Method of physiological data analysis and measurement quality check using principal component analysis | |
| CN109411042B (en) | Electrocardio information processing method and electrocardio workstation | |
| AL-Ziarjawey et al. | Heart rate monitoring and PQRST detection based on graphical user interface with Matlab | |
| WO2019161609A1 (en) | Method for analyzing multi-parameter monitoring data and multi-parameter monitor | |
| CN107951485A (en) | Ambulatory ECG analysis method and apparatus based on artificial intelligence self study | |
| CN106815570B (en) | Electrocardiosignal ST-T segment identification method based on dynamic pattern identification | |
| Chanwimalueang et al. | Enabling R-peak detection in wearable ECG: Combining matched filtering and Hilbert transform | |
| CN109288515B (en) | Periodicity monitoring method and device based on premature beat signal in wearable electrocardiosignal | |
| Thurner et al. | Complex-Pan-Tompkins-Wavelets: Cross-channel ECG beat detection and delineation | |
| Prashar et al. | Novel cardiac arrhythmia processing using machine learning techniques | |
| Khaustov et al. | An open-source standard T-wave alternans detector for benchmarking | |
| Pang et al. | Real time heart ischemia detection in the smart home care system | |
| CN111839505A (en) | Atrial fibrillation classification method based on cardiovascular system electrical-mechanical activity information | |
| Jekova et al. | Arrhythmia classification via time and frequency domain analyses of ventricular and atrial contractions | |
| Mincholé et al. | Discrimination between ischemic and artifactual ST segment events in Holter recordings | |
| Setiawidayat et al. | Method for Obtain Peak Amplitude Value on Discrete Electrocardiogram | |
| Mayapur | Detection and Processing of the R Peak | |
| Gharaviri et al. | Ischemia detection via ECG using ANFIS | |
| Chan et al. | ECG parameter extractor of intelligent home healthcare embedded system | |
| Sanamdikar et al. | Using the GAN method, analysis several characteristics of the ECG signal in order to detect cardiac arrhythmia | |
| Schulte et al. | Identification of a signal for an optimal heart beat detection in multimodal physiological datasets |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20201110 Termination date: 20210712 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |