CN102835954B - Method for generating heart beating waveform template and treating ventricular premature beat detection and device - Google Patents
Method for generating heart beating waveform template and treating ventricular premature beat detection and device Download PDFInfo
- Publication number
- CN102835954B CN102835954B CN201210328503.2A CN201210328503A CN102835954B CN 102835954 B CN102835954 B CN 102835954B CN 201210328503 A CN201210328503 A CN 201210328503A CN 102835954 B CN102835954 B CN 102835954B
- Authority
- CN
- China
- Prior art keywords
- template
- waveform
- clapped
- thought
- read
- 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
Landscapes
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a method for generating a template and detecting a premature beat, and a device. The method comprises the following steps: sequentially and respectively putting M waveforms for measuring a heart beating in N templates relative to the waveforms by sequentially performing a relativity comparison between the waveforms for measuring the heart beating and the N templates, wherein N is less than M; and taking the template which obtains most heart beating times from the first template to the N template as a normal template. The method for detecting the premature beat comprises the following steps: pretreating an inputted electrocardiogram signal to obtain a pretreated signal; extracting the feather parameters of the heart beating for measuring from the pretreated signal, wherein the feather parameters include the width of a QRS wave, an existing RR interphase and a related average RR interphase; obtaining the waveform of the heart beating for measuring from the pretreated signal, and calculating the association coefficient between the waveform of the heart beating for measuring and the waveform of the template; and according to judging conditions that whether the association coefficient between the waveform for measuring the heart beating and the waveform of the template is less than a set threshold and whether the feather parameters are deviated from normal values, judging whether a ventricular premature beat exists.
Description
Technical field
The present invention relates to medical electronics technical field, be specifically related to a kind of heart and clap waveform template generation, ventricular premature contraction check processing method and device.
Background technology
Arrhythmia is the instable important indicator of reflection cardiac electrical activity.Arrhythmia can affect cardiac synchronous and shrink sequential, reduces the efficiency of cardiac pumping.Ventricular premature contraction (Premature Ventricular Contraction, PVC) is one the most common in abnormal heart rhythm, have important clinical meaning, and most arrhythmia is all relevant with its existence.Therefore, the correct and detection fast of PVC, plays a crucial role to the performance improving arrhythmia monitoring and dynamic cardiograph analysis system.
PVC detection method of the prior art, generally has specialist system, feature extraction, template matching, neutral net and fuzzy technology etc.Wherein, the physical significance of neural network is not easily understood, and the time of training study is longer; Fuzzy technology method is in order to reach the object detected fast, and Rulemaking is relatively simple, is difficult to adapt to complicated electrocardiographic recording; And the method for Pattern matching is comparatively conventional, be generally used for the quick identification of PVC, but the method rule is too simple, and accuracy rate has much room for improvement.
Summary of the invention
The invention provides a kind of heart and clap waveform template generation, ventricular premature contraction check processing method and device, make the heart clap waveform template generation more reasonable, more accurate to the detection of ventricular premature contraction, and operand is less.
Technical scheme provided by the invention is as follows:
An object of the present invention is to provide a kind of heart and clap waveform template generation method, comprise the following steps:
First the waveform that first is treated that thought-read is clapped is put in the first template, second is treated the waveform that the heart in the waveform that thought-read is clapped and described first template is clapped compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, described second being treated thought-read bat gathers is the first template class, and if not, then gathering is the second template class; Again the 3rd is treated that the waveform that the heart in the waveform that thought-read is clapped and described first template is clapped compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, described 3rd being treated thought-read bat gathers is the first template class, and if not, then the waveform clapped with the heart in described second template again compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if so, then gathering is the second template class, if not, then gathering is the 3rd template class; By that analogy, treat that the waveform that thought-read is clapped puts into the N number of template class relevant to it, wherein N<M respectively by M successively;
The maximum class of heart umber of beats is obtained as normal template using in described first template class to described N template class.
Preferably, when the heart umber of beats order inputted is greater than M, successively the waveform and first that the input heart is clapped is clapped waveform to the heart in N number of template to compare, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, treat that thought-read bat gathers for this template class by described, if not, then compare with next template, gather for corresponding template class.
Preferably, when the quantity of template completely sets number N, if the next one treats that thought-read claps the thresholding being all less than or equal to setting with the waveform correlation coefficient of N number of template, then template is merged or delete processing, comprising:
Calculate the correlation coefficient between other N-1 template outside the maximum template of acquisition heart umber of beats, find out maximum correlation coefficient; If maximum correlation coefficient is more than or equal to the second thresholding of setting, then corresponding for this maximum correlation coefficient two templates are merged, otherwise the template finding out the minimum number that N-1 template center claps is deleted;
New is treated that thought-read bat is classified as template.
Preferably, when new is clapped after waveform clustering to corresponding template until thought-read, also comprise: add 7/8 of the waveform of original template and treat 1/8 of the waveform that thought-read is clapped, form the waveform of new template.
Another object of the present invention is to provide a kind of heart and claps waveform template generation module, comprising:
Modular unit, total N number of template, claps waveform for storing N number of dissimilar heart;
Template cluster cell, for first the waveform that first is treated that thought-read is clapped being put in the first template, second is treated the waveform that the heart in the waveform that thought-read is clapped and described first template is clapped compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if so, described second being treated that thought-read bat gathers is the first template class, if not, then gathering is the second template class; Again the 3rd is treated that the waveform that the heart in the waveform that thought-read is clapped and described first template is clapped compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, described 3rd being treated thought-read bat gathers is the first template class, and if not, then the waveform clapped with the heart in described second template again compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if so, then gathering is the second template class, if not, then gathering is the 3rd template class; By that analogy, treat that the waveform that thought-read is clapped puts into the N number of template class relevant to it, wherein N<M respectively by M successively;
Judging unit, obtains the maximum class of heart umber of beats as normal template using in described first template class to described N template class.
Another object of the present invention is to provide a kind of ventricular premature contraction check processing method, comprising:
Pretreatment is carried out to the ECG signal of input, obtains pretreated signal;
Extract from described pretreated signal and treat the characteristic parameter that thought-read is clapped, described characteristic parameter comprises the average RR interval of the width of QRS ripple, current RR interval and association;
Obtain from described pretreated signal and treat the waveform that thought-read is clapped, calculate the correlation coefficient treating the waveform of the waveform that thought-read is clapped and template;
Whether the correlation coefficient according to the waveform treating the waveform that thought-read is clapped and described template is less than setting thresholding, and whether the average RR interval of the width of QRS ripple, current RR interval and association departs from the first Rule of judgment of normal value, determines whether as ventricular premature contraction.
Preferably, the first Rule of judgment comprises:
Judge to treat the correlation coefficient of the waveform that thought-read is clapped and the normal waveform of template whether be greater than 0.99 and current RR interval be in average RR interval 90% ~ 120% between;
If result is yes, to treat that thought-read is clapped gathers for normal template class, and it is normal to be defined as room property, if the result is negative, judge whether the correlation coefficient of the waveform treated beyond the waveform that thought-read is clapped and normal template is greater than 0.99 successively, if, then treat that thought-read bat gathers for this template class by described, and calculate described by the correlation coefficient of the template class of cluster and described normal template class, further according to by the correlation coefficient of the template class of cluster and described normal template class, current RR interval, accounts for the ratio of average interval and the second Rule of judgment of QRS wave width, determine whether as ventricular premature contraction.
Preferably, second Rule of judgment comprises: judge whether the described correlation coefficient by the template class of cluster and described normal template class is greater than 0.96 further, or the described correlation coefficient by the template class of cluster and described normal template class whether be greater than 0.78 and current RR interval be average interval 90% ~ 120% between; Or the described correlation coefficient by the template class of cluster and described normal template class whether be greater than 0.5 and current RR interval be average interval 90% ~ 120% between and QRS wave width is less than 120 milliseconds; If so, be defined as room property normal, if the result is negative, further according to the ratio and the QRS wave width Rule of judgment that are accounted for average interval by the correlation coefficient of the template class of cluster and described normal template class, current RR interval, determine whether as ventricular premature contraction.
Preferably, the 3rd Rule of judgment comprises: the described correlation coefficient by the template class of cluster and described normal template class whether is greater than 0.78 and QRS width is greater than 120 milliseconds; Or be greater than 0.5 by the correlation coefficient of the template class of cluster and described normal template class and current RR interval is greater than average interval 120% or is less than average interval 90%, if so, be then defined as ventricular premature contraction, otherwise be uncertain.
Another object of the present invention is to provide a kind of ventricular premature contraction check processing device, comprising:
Pretreatment unit, for carrying out pretreatment to the ECG signal of input, obtains pretreated signal;
Feature extraction unit, treat for extracting from described pretreated signal the characteristic parameter that thought-read is clapped, described characteristic parameter comprises the average RR interval of the width of QRS ripple, current RR interval and association;
Calculation of correlation factor unit, treating for obtaining from described pretreated signal the waveform that thought-read is clapped, calculating the correlation coefficient treating the waveform of the waveform that thought-read is clapped and template;
Whether judging unit, be less than certain thresholding according to described correlation coefficient, and whether the average RR interval of the width of QRS ripple, current RR interval and association departs from normal value Rule of judgment, determines whether as ventricular premature contraction
Preferably, described judging unit specifically comprises following determination methods:
Judge to treat the correlation coefficient of the waveform that thought-read is clapped and the normal waveform of template whether be greater than 0.99 and current RR interval be in average RR interval 90% ~ 120% between; If result is yes, to treat that thought-read is clapped gathers for normal template class, and it is normal to be defined as room property, if the result is negative, judge successively to treat whether the correlation coefficient of the waveform of the waveform that thought-read is clapped and other template is greater than 0.99, if so, then treat that thought-read is clapped gather described for this template class, and calculate described by the correlation coefficient of the template class of cluster and described normal template class; If with normal template is not mated, judge whether the described correlation coefficient by the template class of cluster and described normal template class is greater than 0.96 further, or the described correlation coefficient by the template class of cluster and described normal template class whether be greater than 0.78 and current RR interval be average interval 90% ~ 120% between; Or the described correlation coefficient by the template class of cluster and described normal template class whether be greater than 0.5 and current RR interval be average interval 90% ~ 120% between and QRS wave width is less than 120 milliseconds; If it is normal to be defined as room property, if the result is negative, be defined as treating further judgement; If be defined as treating further judgement, then comprise further: the described correlation coefficient by the template class of cluster and described normal template class whether is greater than 0.78 and QRS width is greater than 120 milliseconds; Or be greater than 0.5 by the correlation coefficient of the template class of cluster and described normal template class and current RR interval is greater than average interval 120% or is less than average interval 90%, if so, be then defined as ventricular premature contraction, otherwise be uncertain.
Technique scheme can be found out, the embodiment of the present invention takes multiple template in template process of establishing, makes template accurately reflect the situation of normal waveform; After pretreatment is carried out to electrocardiosignal, be extracted the parameter such as average RR interval comprising the width treating the QRS ripple that thought-read is clapped, current RR interval and association, and calculate the correlation coefficient treating the waveform of the waveform that thought-read is clapped and normal template, then after these parameters comprehensive judge that meeting setting requires, export the result being defined as ventricular premature contraction, thus can more accurately detect ventricular premature contraction PVC relative to art methods, improve accuracy rate, and operand is less.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, be briefly described to the accompanying drawing used required in embodiment or description of the prior art below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, other accompanying drawing can also be obtained according to these accompanying drawings.
Fig. 1 is the embodiment of the present invention one ventricular premature contraction check processing method flow diagram.
Fig. 2 is the block flow diagram of the embodiment of the present invention two ventricular premature contraction check processing method.
Fig. 3 is the particular flow sheet of the embodiment of the present invention two ventricular premature contraction check processing method.
Fig. 4 is the template matching flow chart in the embodiment of the present invention two method.
Fig. 5 is the template generation flow chart in the embodiment of the present invention two method.
Fig. 6 is R wave height, QRS wave width marked graph in the embodiment of the present invention two method.
Fig. 7 is that the heart in the embodiment of the present invention two method is clapped and interval relation figure.
Fig. 8 is template matching procedure chart in the embodiment of the present invention two method.
Fig. 9 is embodiment of the present invention ventricular premature contraction check processing apparatus structure schematic diagram.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, be clearly and completely described the technical scheme in the embodiment of the present invention, obviously, described embodiment is only the present invention's part embodiment, instead of whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art, not making other embodiments all obtained under creative work prerequisite, belong to the scope of protection of the invention.
The invention provides a kind of heart and clap waveform template generation, ventricular premature contraction check processing method and device, generate more fast, template easily, it is more accurate to make the detection of ventricular premature contraction, and operand is less.
Be described in detail respectively below in conjunction with accompanying drawing.
A kind of heart claps waveform template generation method, comprises the following steps:
First the waveform that first is treated that thought-read is clapped is put in the first template, second is treated the waveform that the heart in the waveform that thought-read is clapped and described first template is clapped compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, described second being treated thought-read bat gathers is the first template class, and if not, then gathering is the second template class; Again the 3rd is treated that the waveform that the heart in the waveform that thought-read is clapped and described first template is clapped compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, described 3rd being treated thought-read bat gathers is the first template class, and if not, then the waveform clapped with the heart in described second template again compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if so, then gathering is the second template class, if not, then gathering is the 3rd template class; By that analogy, treat that the waveform that thought-read is clapped puts into the N number of template class relevant to it, wherein N<M respectively by M successively;
The maximum class of heart umber of beats is obtained as normal template using in described first template class to described N template class.
As embodiment, the first thresholding >0.8, preferably 0.95.
As embodiment M=10, a N=8.
Template along with the new arrival that thought-read is clapped for the treatment of, continuous real-time update.When the heart umber of beats order inputted is greater than M, successively the waveform and first that the input heart is clapped is clapped waveform to the heart in N number of template to compare, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if so, treat that thought-read bat gathers for this template class by described, if not, then compare with next template, gather for corresponding template class.
When the quantity of template completely sets number N, treat that if new thought-read claps the thresholding being all less than or equal to setting with the waveform correlation coefficient of N number of template, then template merged or delete processing, comprising:
Calculate the correlation coefficient between other N-1 template outside the maximum template of acquisition heart umber of beats, find out maximum correlation coefficient; If maximum correlation coefficient is more than or equal to the second thresholding of setting, then corresponding for this maximum correlation coefficient two templates are merged, otherwise the template finding out the minimum number that N-1 template center claps is deleted;
New is treated that thought-read bat is classified as template.
As embodiment, the second thresholding >0.7, is preferably 0.95.
When new is clapped after waveform clustering to corresponding template until thought-read, also comprise: add 7/8 of the waveform of original template and treat 1/8 of the waveform that thought-read is clapped, form the waveform of new template.
A kind of heart claps waveform template generation module, comprising:
Modular unit, total N number of template, claps waveform for storing N number of dissimilar heart;
Template cluster cell, for first the waveform that first is treated that thought-read is clapped being put in the first template, second is treated the waveform that the heart in the waveform that thought-read is clapped and described first template is clapped compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if so, described second being treated that thought-read bat gathers is the first template class, if not, then gathering is the second template class; Again the 3rd is treated that the waveform that the heart in the waveform that thought-read is clapped and described first template is clapped compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, described 3rd being treated thought-read bat gathers is the first template class, and if not, then the waveform clapped with the heart in described second template again compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if so, then gathering is the second template class, if not, then gathering is the 3rd template class; By that analogy, treat that the waveform that thought-read is clapped puts into the N number of template class relevant to it, wherein N<M respectively by M successively;
Judging unit, obtains the maximum class of heart umber of beats as normal template using in described first template class to described N template class.
Fig. 1 is the embodiment of the present invention one ventricular premature contraction check processing method flow diagram.As shown in Figure 1, comprising:
Step 101, to input ECG signal carry out pretreatment, obtain pretreated signal;
Step 102, extract and treat the characteristic parameter that thought-read is clapped from described pretreated signal, described characteristic parameter comprises the average RR interval of the width of QRS ripple, current RR interval and association;
Here the average RR interval said, the average of setting number (such as 8) the RR interval that current RR interval is nearest is referred to.
Step 103, obtain and treat the waveform that thought-read is clapped from described pretreated signal, the correlation coefficient of the waveform of the waveform that thought-read is clapped and template is treated in calculating;
Step 104, according to treating whether the correlation coefficient of waveform of the waveform that thought-read is clapped and described template is less than certain thresholding, and the average RR interval of the width of QRS ripple, current RR interval and association whether depart from normal value Rule of judgment, determine whether as ventricular premature contraction.
It should be noted that, step 102 and 103 does not have inevitable ordering relation.Treat that thought-read claps premature beat check processing and template renewal does not have sequencing to limit yet.
Can find, the embodiment of the present invention is after carrying out pretreatment to electrocardiosignal, be extracted comprise treat the QRS ripple that thought-read is clapped width, treat the parameters such as the average RR interval that thought-read is clapped, and calculate the correlation coefficient treating the waveform of the waveform that thought-read is clapped and template, then, after these parameters comprehensive judge that meeting setting requires, export the result being defined as ventricular premature contraction, thus can more accurately detect ventricular premature contraction PVC relative to art methods, improve accuracy rate, and operand is less.
Fig. 2 is the block flow diagram of the embodiment of the present invention two ventricular premature contraction check processing method.
As shown in Figure 2, whole handling process comprises: after input signal, carries out Signal Pretreatment, then carries out feature extraction to signal, then input template matching system, carries out template matching and logical judgment, last output category result.Concrete:
Digitized electrocardiogram ECG signal carries out pretreatment, namely carries out pretreatment through high and low pass filter, baseline drift can be made to be corrected, and substantially inhibit high-frequency noise by pretreatment, thus reduces interference.Carry out feature extraction on this basis, namely extract QRS wave characteristic parameter, the accuracy of parameter can be ensured.Because the main morphological features of PVC is the roomy lopsided QRS wave group occurred ahead of time, the feature so RR interval (RR represents the interval between two QRS ripples), QRS wave width are absolutely necessary, therefore extracts these as characteristic parameter.
In addition, also obtain it when carrying out template matching further and treat that thought-read claps the correlation coefficient of waveform and template waveforms, judged by parametric synthesis such as RR interval, QRS wave width and correlation coefficient CC, draw classification results.
The inventive method after often inputting data, by flow process shown in Fig. 2 real-time carry out PVC detection.Detect QRS ripple when a heartbeat arrives, and measure its a stack features parameter, then send into template-matching systems, draw beat classification result.When detecting, the inventive method postpones a heart and claps, so when the classification results returning upper heart bat when front center is clapped being detected.
Fig. 3 is the particular flow sheet of the embodiment of the present invention two ventricular premature contraction check processing method.As shown in Figure 3, step is comprised:
Step 301, pretreatment is carried out to input signal.
The noises such as power frequency, myoelectricity, baseline drift can be comprised in the electrocardiogram ECG signal of general input.Baseline drift is by breathing or kinetic low-frequency disturbance between electrode and human body, generally at below 0.5Hz, can change the form of ST section in ECG, the accuracy that affecting parameters extracts, and therefore filtering baseline drift is very important.The inventive method, in order to reduce noise jamming, carries out Filtering Processing in advance to input signal, such as, carry out high and low pass filter.
Step 302, extraction characteristic parameter.
What the present invention extracted treats that characteristic parameter that thought-read is clapped mainly comprises the average RR interval of the QRS wave width treated thought-read and clap, current RR interval, association.
Wherein, as shown in Figure 6, the width between Q ripple and S ripple is QRS wave width.Fig. 6 is R wave height in the embodiment of the present invention two method, QRS wave width marked graph.RR interval, needs last interval RR, a current interval RR and rear interval RR, and as shown in Figure 7, Fig. 7 is that the heart in the embodiment of the present invention two method is clapped and interval relation figure to the relation that itself and the heart are clapped.
Step 303, carry out template matching.Specifically see the description in Fig. 4.
Step 304, according to matching result output category result.
Fig. 4 is the template matching flow chart in the embodiment of the present invention two method, and Fig. 8 then shows the schematic process of template matching.
Before introducing this flow chart, first relational language is defined.
Average RR interval, refer to the average of setting number (such as 8) the RR interval that current RR interval is nearest.
It should be noted that, here the threshold value compared set by correlation coefficient CC is as 0.95,0.99 etc., the threshold value 90% ~ 120% etc. compared with average RR interval, the threshold value compared with QRS width 120 milliseconds etc., it is all rule of thumb gained, not immobilize, can arrange according to result of the test.
As shown in Figure 4, M before signal is clapped and first carries out training study, after having learnt, new-comer is treated that thought-read is clapped, first and normal template is mated, if meet correlation coefficient CC to be greater than 0.99, current RR interval be average RR interval 90% ~ 120% between, the match is successful, then gather for normal template class, return normal; If mate unsuccessful, then successively with other template matching, if the match is successful, then gather for this class template, if now meet correlation coefficient CC to be more than or equal to 0.96; Or correlation coefficient be greater than 0.78 and current RR interval be average RR interval 90% ~ 120% between; Or correlation coefficient CC be greater than 0.5 and current RR interval be average RR interval 90% ~ 120% between and QRS wave width is less than 120 milliseconds, then export normal result; If do not met, but meet correlation coefficient CC and be greater than 0.78 and QRS width is greater than 120 milliseconds; Or correlation coefficient CC is greater than 0.5 and current RR interval is greater than average interval 120%, or correlation coefficient whether is greater than 0.5 and current RR interval is less than average interval 90%, then return the result confirming PVC, if do not met again, then return uncertain result.
Concrete, Fig. 4 comprises step:
Step 401, input treat that thought-read is clapped.
Treating that thought-read is clapped can be one section of waveform of R wavefront 600ms, rear 400ms.
Step 402, judge whether existing normal template, if so, enter step 403, if not, enter step 414 and generate normal template.The method generating normal template is see described by flow chart 5.Said normal template in the present invention, representative be normal waveform.
Step 403, calculating treat the correlation coefficient CC parameter of the waveform of the waveform that thought-read is clapped and normal template.
The inventive method, by learning front 10 waveforms of every segment signal data, obtains the waveform of normal template, treats that thought-read is clapped and carries out cluster, obtain the correlation coefficient CC of the waveform of itself and normal template to each.It should be noted that, be illustrate with 10 waveforms but be not limited to this here.
The Calculation of correlation factor formula of extraction of the present invention is as follows:
Wherein, N for the length of the number of winning the confidence, also namely refer to the signal number of collection in a period of time (such as 320ms) before and after centered by QRS ripple, such as sample rate is 600, then N=320*600/1000=192; X (n), y (n) are respectively the waveshape signal of template and treat the waveshape signal that thought-read is clapped,
be respectively x (n), the meansigma methods of y (n).
Step 404, according to parameter comparative result, judge whether that meeting correlation coefficient CC is greater than 0.99, current RR interval be average RR interval 90% ~ 120% between, if meet, enter step 405, if do not meet, enter step 408.
Average RR interval in step 302 extract acquisition.
Step 405, because meet correlation coefficient CC and be greater than 0.99, current RR interval be average RR interval 90% ~ 120% between, be defined as coupling, therefore will treat that thought-read bat gathers for normal template class, and enter step 406.
Step 406, upgrade normal template, enter step 407.
Normal template is upgraded according to the result of step 405 in this step.
After cluster, if satisfied condition template upgraded, then adopt 7/8 of original normal template waveforms to add and treat now 1/8 of the waveform that thought-read is clapped, form the waveform of new normal template.In addition, the maximum quantity of template can be set to M (being such as 8), when template number is expired, newly to clap with the unmatched heart of waveform that is template if having, then template is merged or delete processing, handling process is as follows: 1) calculate the correlation coefficient between other 7 templates except normal template, find out maximum correlation coefficient; 2) if maximum correlation coefficient is more than or equal to 0.80, then corresponding for this maximum correlation coefficient two templates merged, if maximum correlation coefficient is less than 0.80, then the template finding out the minimum number that 7 template center clap is deleted.
Step 407, this step export normal result.
Step 408, because do not meet correlation coefficient CC and be greater than 0.99, current RR interval be average RR interval 90% ~ 120% between, be defined as not mating, therefore will treat thought-read bat successively with other template matching, if the match is successful, then gather for this class template.Enter step 409.
Step 409, calculate the correlation coefficient CC of this class template and normal template class.
Step 410, according to parameter comparative result, judge whether that meeting correlation coefficient CC is more than or equal to 0.96; Or correlation coefficient be greater than 0.78 and current RR interval be average RR interval 90% ~ 120% between; Or correlation coefficient CC be greater than 0.5 and current RR interval be average RR interval 90% ~ 120% between and QRS wave width is less than 120 milliseconds, if so, enter step 407 and export normal outcome, if not, enter step 411.
Step 411, judge whether to meet correlation coefficient CC further and be greater than 0.78 and QRS width is greater than 120 milliseconds; Or correlation coefficient CC is greater than 0.5 and current RR interval is greater than average interval 120%, or correlation coefficient whether is greater than 0.5 and current RR interval is less than average interval 90%, if so, enters step 412, if not, enters step 413.
Step 412, output are defined as the result of PVC.
By above-mentioned judgement flow process, finally determine PVC.
Step 413, export uncertain result.
Step 414, generate normal template.Specifically see described by Fig. 5.
Above template renewal and premature beat detect judgement and there is no sequencing relation, should not be limited to above-described embodiment.
Fig. 5 is the template generation flow chart in the embodiment of the present invention two method.
Template generation process of the present invention, adopt setting number (such as 8) template altogether, place a QRS ripple inside each template, each template has a template waveforms.Learn front ten hearts to clap, QRS ripple cluster forms template set, using templates maximum for cluster as advantage template, as the optimum object that later ECG signal compares, also namely as the normal template for comparing.
Concrete, as shown in Figure 5, template generation process comprises step:
Step 501, input treat that thought-read is clapped.
Step 502, the waveform that first is treated that thought-read is clapped is put in the first template, second is treated the waveform that the heart in the waveform that thought-read is clapped and the first template is clapped compares, calculate correlation coefficient.
Step 503, according to parameter comparative result, judge whether that meeting correlation coefficient CC is greater than 0.95, if so, enters step 504, if not, enters step 505.
Step 504, because meet correlation coefficient CC and be greater than 0.95, therefore second is treated that thought-read claps that to gather be the first template class.Enter step 506.
Step 505, because do not meet correlation coefficient CC and be greater than 0.95, therefore treat second that the waveform that thought-read is clapped carries out correlation ratio comparatively with the waveform of other templates successively, if relevant, then gathering for this template class, is also other classes.Enter step 507.
Step 506, more new template.Enter step 507.
Description more in the mode of new template and earlier figures 4 is similar, repeats no more herein.
Step 507, judgement are treated that thought-read is clapped and whether have been inputted N bat, if so, enter step 508, if not, return step 501 and input new bat.
N in this step claps and sets as required.
Step 508, will the maximum template of heart umber of beats be obtained as normal template, namely as and treat the template that the waveform that thought-read is clapped compares.
This step obtains the maximum class of heart umber of beats as normal template using in the first template class to described N template class.
Can find from said method flow process, the inventive method has following beneficial effect: the embodiment of the present invention is after carrying out pretreatment to electrocardiosignal, be extracted the parameter such as average RR interval comprising the width treating the QRS ripple that thought-read is clapped, current RR interval and association, and calculate the correlation coefficient treating the waveform of the waveform that thought-read is clapped and normal template, then after these parameters comprehensive judge that meeting setting requires, export the result being defined as ventricular premature contraction, thus can more accurately detect ventricular premature contraction PVC relative to art methods, improve accuracy rate, and operand is less.
The present invention is corresponding also provides a kind of ventricular premature contraction check processing device.
Fig. 9 is embodiment of the present invention apparatus structure schematic diagram, and as shown in Figure 9, this device comprises:
Pretreatment unit 91, for carrying out pretreatment to the ECG signal of input, obtains pretreated signal;
Feature extraction unit 92, treat for extracting from described pretreated signal the characteristic parameter that thought-read is clapped, described characteristic parameter comprises the average RR interval of the width of QRS ripple, current RR interval and association;
Calculation of correlation factor unit 93, treating for obtaining from described pretreated signal the waveform that thought-read is clapped, calculating the correlation coefficient treating the waveform of the waveform that thought-read is clapped and normal template;
Whether judging unit 93, be less than certain thresholding according to described correlation coefficient, and whether the average RR interval of the width of QRS ripple, current RR interval and association departs from normal value Rule of judgment, determines whether as ventricular premature contraction.
Further, judging unit comprises:
Judge to treat the correlation coefficient of the waveform that thought-read is clapped and the normal waveform of template whether be greater than 0.99 and current RR interval be in average RR interval 90% ~ 120% between; If result is yes, to treat that thought-read is clapped gathers for normal template class, and it is normal to be defined as room property, if the result is negative, judge successively to treat whether the correlation coefficient of the waveform of the waveform that thought-read is clapped and other template is greater than 0.99, if so, then treat that thought-read is clapped gather described for this template class, and calculate described by the correlation coefficient of the template class of cluster and described normal template class; If with normal template is not mated, judge whether the described correlation coefficient by the template class of cluster and described normal template class is greater than 0.96 further, or the described correlation coefficient by the template class of cluster and described normal template class whether be greater than 0.78 and current RR interval be average interval 90% ~ 120% between; Or the described correlation coefficient by the template class of cluster and described normal template class whether be greater than 0.5 and current RR interval be average interval 90% ~ 120% between and QRS wave width is less than 120 milliseconds; If it is normal to be defined as room property, if the result is negative, be defined as treating further judgement; If be defined as treating further judgement, then comprise further: the described correlation coefficient by the template class of cluster and described normal template class whether is greater than 0.78 and QRS width is greater than 120 milliseconds; Or be greater than 0.5 by the correlation coefficient of the template class of cluster and described normal template class and current RR interval is greater than average interval 120% or is less than average interval 90%, if so, be then defined as ventricular premature contraction, otherwise be uncertain.
Above to the scheme that the embodiment of the present invention provides, be described in detail, apply specific case herein and set forth principle of the present invention and embodiment, the explanation of above embodiment just understands method of the present invention and core concept thereof for helping; Meanwhile, for one of ordinary skill in the art, according to thought of the present invention, all will change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.
Claims (4)
1. the heart claps a waveform template generation method, it is characterized in that, comprises the following steps:
First the waveform that first is treated that thought-read is clapped is put in the first template, second is treated the waveform that the heart in the waveform that thought-read is clapped and described first template is clapped compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, described second being treated thought-read bat gathers is the first template class, and add that by 7/8 of the waveform of the first original template second is treated 1/8 of the waveform that thought-read is clapped, form the waveform of the first new template, if not, then gathering is the second template class, and described second waveform treating that thought-read is clapped is put in the second template, again the 3rd is treated that the waveform that the heart in the waveform that thought-read is clapped and described first template is clapped compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, described 3rd being treated thought-read bat gathers is the first template class, and add that by 7/8 of the waveform of the first original template the 3rd is treated 1/8 of the waveform that thought-read is clapped, forms the waveform of the first new template, if not, if the waveform then not having the heart to clap in the second template, the waveform that 3rd is treated that thought-read is clapped is put in the second template, if the waveform intentionally clapped in the second template, then again the waveform that the heart in the waveform of described 3rd heart to be measured bat and described second template is clapped is compared, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, then gathering is the second template class, and add that by 7/8 of the waveform of the second original template the 3rd is treated 1/8 of the waveform that thought-read is clapped, form the waveform of the second new template, if not, then gathering is the 3rd template class, and described 3rd waveform treating that thought-read is clapped is put in the 3rd template, by that analogy, treat that the waveform that thought-read is clapped puts into the N number of template class relevant to it, wherein N<M respectively by M successively,
Using described first template class to the template obtained in described N template class corresponding to the maximum class of heart umber of beats as normal template.
2. the heart according to claim 1 claps waveform template generation method, it is characterized in that, further comprising the steps of:
When the heart umber of beats order inputted is greater than M, successively the waveform and first that the input heart is clapped is clapped waveform to the heart in N number of template to compare, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if so, treat that thought-read bat gathers for this template class by described, if not, then compare with next template, gather for corresponding template class.
3. the heart according to claim 1 and 2 claps waveform template generation method, it is characterized in that: described method also comprises
When the quantity of template completely sets number N, if the next one treats that thought-read claps the thresholding being all less than or equal to setting with the waveform correlation coefficient of N number of template, then template is merged or delete processing, comprising:
Calculate the correlation coefficient between other N-1 template outside the maximum template of acquisition heart umber of beats, find out maximum correlation coefficient; If maximum correlation coefficient is more than or equal to the second thresholding of setting, then corresponding for this maximum correlation coefficient two templates are merged, otherwise the template finding out the minimum number that N-1 template center claps is deleted;
New is treated that thought-read bat is classified as template.
4. the heart claps a waveform template generation module, it is characterized in that, comprising:
Modular unit, total N number of template, claps waveform for storing N number of dissimilar heart;
Template cluster cell, for first the waveform that first is treated that thought-read is clapped being put in the first template, second is treated the waveform that the heart in the waveform that thought-read is clapped and described first template is clapped compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, described second being treated thought-read bat gathers is the first template class, and add that by 7/8 of the waveform of the first original template second is treated 1/8 of the waveform that thought-read is clapped, form the waveform of the first new template, if not, then gathering is the second template class, and described second waveform treating that thought-read is clapped is put in the second template, again the 3rd is treated that the waveform that the heart in the waveform that thought-read is clapped and described first template is clapped compares, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, described 3rd being treated thought-read bat gathers is the first template class, and add that by 7/8 of the waveform of the first original template the 3rd is treated 1/8 of the waveform that thought-read is clapped, forms the waveform of the first new template, if not, if the waveform then not having the heart to clap in the second template, the waveform that 3rd is treated that thought-read is clapped is put in the second template, if the waveform intentionally clapped in the second template, then the waveform that the heart in the waveform of described 3rd heart to be measured bat and described second template is clapped is compared, calculate correlation coefficient, judge whether described correlation coefficient is greater than the first thresholding of setting, if, then gathering is the second template class, and add that by 7/8 of the waveform of the second original template the 3rd is treated 1/8 of the waveform that thought-read is clapped, form the waveform of the second new template, if not, then gathering is the 3rd template class, and described 3rd waveform treating that thought-read is clapped is put in the 3rd template, by that analogy, treat that the waveform that thought-read is clapped puts into the N number of template class relevant to it, wherein N<M respectively by M successively,
Judging unit, using described first template class to the template obtained in described N template class corresponding to the maximum class of heart umber of beats as normal template.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210328503.2A CN102835954B (en) | 2012-09-07 | 2012-09-07 | Method for generating heart beating waveform template and treating ventricular premature beat detection and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210328503.2A CN102835954B (en) | 2012-09-07 | 2012-09-07 | Method for generating heart beating waveform template and treating ventricular premature beat detection and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102835954A CN102835954A (en) | 2012-12-26 |
| CN102835954B true CN102835954B (en) | 2015-03-11 |
Family
ID=47363880
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210328503.2A Active CN102835954B (en) | 2012-09-07 | 2012-09-07 | Method for generating heart beating waveform template and treating ventricular premature beat detection and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102835954B (en) |
Families Citing this family (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104398256B (en) * | 2014-11-13 | 2017-01-11 | 北京海思敏医疗技术有限公司 | Method and device for detecting electrocardio waveforms through computer |
| US20160310029A1 (en) * | 2015-04-23 | 2016-10-27 | Medtronic, Inc. | Method and apparatus for determining a premature ventricular contraction in a medical monitoring device |
| CN106108880B (en) * | 2016-06-28 | 2020-10-16 | 吉林大学 | Automatic heart beat identification method and system |
| US10433749B2 (en) * | 2016-08-10 | 2019-10-08 | Biosense Webster (Israel) Ltd. | Identifying ECG signals having the same morphology |
| CN107495960A (en) * | 2017-01-19 | 2017-12-22 | 山东医联万家健康科技有限公司 | A kind of heart real time signal monitoring processing method |
| CN107397547A (en) * | 2017-08-01 | 2017-11-28 | 深圳大学 | A kind of ECG typical case's heart beat template generation method and system |
| CN107456227B (en) * | 2017-08-16 | 2021-03-30 | 北京蓬阳丰业医疗设备有限公司 | Full-lead electrocardiogram clustering template system and 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 |
| CN111655142B (en) * | 2018-01-18 | 2024-07-16 | 深圳迈瑞生物医疗电子股份有限公司 | Arrhythmia real-time analysis method, electrocardiograph monitoring device and storage medium |
| CN109394206B (en) * | 2018-11-14 | 2021-08-27 | 东南大学 | Real-time monitoring method and device based on premature beat signal in wearable electrocardiosignal |
| CN109288515B (en) * | 2018-11-14 | 2021-06-22 | 东南大学 | Periodicity monitoring method and device based on premature beat signal in wearable electrocardiosignal |
| CN110327031A (en) * | 2018-11-29 | 2019-10-15 | 武汉中旗生物医疗电子有限公司 | A method of removal electrocardiosignal motion artifacts |
| 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 |
| CN113712525B (en) * | 2020-05-21 | 2024-07-12 | 深圳市理邦精密仪器股份有限公司 | Physiological parameter processing method and device and medical equipment |
| CN113876335B (en) * | 2020-06-17 | 2023-05-02 | 深圳邦健生物医疗设备股份有限公司 | Electrocardiogram and heart beat classification method, device, electronic equipment and medium |
| 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 |
| CN112971800B (en) * | 2021-01-27 | 2023-01-03 | 东南大学 | Deep learning-based electrocardio premature beat signal processing method, module and device |
| CN112842349B (en) * | 2021-02-20 | 2022-03-11 | 无锡市中健科仪有限公司 | Intelligent analysis method for supraventricular premature beat based on attention transfer mechanism |
| CN113180687B (en) * | 2021-04-29 | 2024-02-09 | 深圳邦健生物医疗设备股份有限公司 | Multi-guide dynamic heartbeat real-time classification method, device, equipment and storage medium |
| CN114027853B (en) * | 2021-12-16 | 2022-09-27 | 安徽心之声医疗科技有限公司 | QRS complex detection method, device, medium and equipment based on feature template matching |
| CN114224353B (en) * | 2022-02-21 | 2023-03-21 | 深圳泰和智能医疗科技有限公司 | Electrocardio detection classification method based on body temperature monitor |
| CN116350199B (en) * | 2023-05-31 | 2023-08-18 | 合肥心之声健康科技有限公司 | Dynamic electrocardiogram and heart beat template generation method and system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2103253A1 (en) * | 2008-03-18 | 2009-09-23 | BIOTRONIK CRM Patent AG | Device, method and computer-readable storage medium for SVT and VT Classification |
| CN101810476A (en) * | 2009-12-22 | 2010-08-25 | 李顶立 | Classification method of heart beat template of dynamic electrocardiogram |
| CN101953682A (en) * | 2010-10-15 | 2011-01-26 | 张辉 | Heartbeat detection method based on cuff device |
-
2012
- 2012-09-07 CN CN201210328503.2A patent/CN102835954B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2103253A1 (en) * | 2008-03-18 | 2009-09-23 | BIOTRONIK CRM Patent AG | Device, method and computer-readable storage medium for SVT and VT Classification |
| CN101810476A (en) * | 2009-12-22 | 2010-08-25 | 李顶立 | Classification method of heart beat template of dynamic electrocardiogram |
| CN101953682A (en) * | 2010-10-15 | 2011-01-26 | 张辉 | Heartbeat detection method based on cuff device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102835954A (en) | 2012-12-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102835954B (en) | Method for generating heart beating waveform template and treating ventricular premature beat detection and device | |
| CN101828918B (en) | Electrocardiosignal R peak detection method based on waveform characteristic matching | |
| CN102779234B (en) | A kind of cardiogram classification processing method and device | |
| CN103584854B (en) | Extraction method of electrocardiosignal R waves | |
| CN101766484B (en) | Method and equipment for identification and classification of electrocardiogram | |
| CN104887220B (en) | A kind of method and system that Fetal ECG signal is extracted by stomach wall electrocardiosignal | |
| CN109276242A (en) | The method and apparatus of electrocardiosignal type identification | |
| CN103705234B (en) | Demodulation method in ambulatory ecg signal data and device | |
| CN108056773A (en) | Based on the algorithms of QRS complexes detection in electrocardiogram signal for improving variation mode decomposition | |
| CN106725420A (en) | VPB recognition methods and VPB identifying system | |
| CN108647565A (en) | A kind of data preprocessing method classified to electrocardiosignal based on deep learning model | |
| CN109117730A (en) | Electrocardiogram auricular fibrillation real-time judge method, apparatus, system and storage medium | |
| CN104161510A (en) | Multistage lead electrocardiograph signal QRS waveform identification method | |
| CN113057648A (en) | ECG signal classification method based on composite LSTM structure | |
| CN108992053B (en) | Method for real-time non-binding detection of heart rate and heartbeat interval | |
| CN106874722A (en) | A kind of personal identification method and its device based on electrocardiosignal | |
| CN104840186A (en) | Evaluation method of autonomic nervous function of patient suffering from CHF (Congestive Heart-Failure) | |
| CN103829944B (en) | Based on the thoracic impedance signal processing method of pattern recognition | |
| CN115486855B (en) | Electrocardiogram and heart beat classification method based on QRS complex indefinite cycle superposition | |
| CN104771178A (en) | Method and device of identity recognition | |
| CN103970975A (en) | Electrocardio data processing method and electrocardio data processing system | |
| CN108509823A (en) | The detection method and device of QRS complex | |
| CN110327032A (en) | It is a kind of singly to lead the accurate recognizer of electrocardiosignal PQRST wave joint | |
| CN113116328A (en) | Heart rate detection method based on ballistocardiogram | |
| CN106137184A (en) | Electrocardiosignal QRS complex detection method based on wavelet transformation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C53 | Correction of patent for invention or patent application | ||
| CB02 | Change of applicant information |
Address after: 518102, Shenzhen, Guangdong Province, Baoan District, Xixiang Bao Road 168, Shenzhen famous industrial products display procurement center, block A, building seven Applicant after: Shenzhen Biocare Bio-Medical Equipment Co., Ltd. Address before: 518102, Shenzhen, Guangdong Province, Baoan District, Xixiang Bao Road 168, Shenzhen famous industrial products display procurement center, block A, building seven Applicant before: Shenzhen Biocare Electronics Co., Ltd. |
|
| COR | Change of bibliographic data |
Free format text: CORRECT: APPLICANT; FROM: SHENZHEN BIOCARE ELECTRONICS CO., LTD. TO: SHENZHEN BIOCARE BIO-MEDICAL EQUIPMENT CO., LTD. |
|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |