WO2021253709A1 - 心电图心搏分类方法、装置、电子设备和介质 - Google Patents

心电图心搏分类方法、装置、电子设备和介质 Download PDF

Info

Publication number
WO2021253709A1
WO2021253709A1 PCT/CN2020/126352 CN2020126352W WO2021253709A1 WO 2021253709 A1 WO2021253709 A1 WO 2021253709A1 CN 2020126352 W CN2020126352 W CN 2020126352W WO 2021253709 A1 WO2021253709 A1 WO 2021253709A1
Authority
WO
WIPO (PCT)
Prior art keywords
template
classified
data
heartbeat
type
Prior art date
Application number
PCT/CN2020/126352
Other languages
English (en)
French (fr)
Inventor
刘盛捷
Original Assignee
深圳邦健生物医疗设备股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳邦健生物医疗设备股份有限公司 filed Critical 深圳邦健生物医疗设备股份有限公司
Publication of WO2021253709A1 publication Critical patent/WO2021253709A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
    • A61B5/346—Analysis of electrocardiograms
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
    • A61B5/346—Analysis of electrocardiograms
    • A61B5/349—Detecting specific parameters of the electrocardiograph cycle
    • A61B5/35—Detecting specific parameters of the electrocardiograph cycle by template matching
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
    • A61B5/346—Analysis of electrocardiograms
    • A61B5/349—Detecting specific parameters of the electrocardiograph cycle
    • A61B5/352—Detecting R peaks, e.g. for synchronising diagnostic apparatus; Estimating R-R interval

Definitions

  • the present invention relates to the technical field of electrocardiogram data processing, in particular to an electrocardiogram heartbeat classification method, device, electronic equipment and medium.
  • the ECG heartbeat classification is defined as the type recognition of the heartbeat.
  • the electrocardiogram software can automatically identify the type of heartbeat by using different algorithms, and provide the doctor with information about whether the heartbeat of the members contained in the template is normal, which can improve the efficiency and accuracy of the doctor's diagnosis.
  • the current ECG heartbeat classification can use artificial feature extraction method.
  • This method requires the professional clinical knowledge of professional doctors, and the amount of calculation is large when acquiring various characteristic parameters; and the method of training neural network models for classification processing requires the collection of a large number of clinical data as samples, which is difficult and time-consuming to develop Longer.
  • This application provides an ECG heartbeat classification method, device, electronic equipment and medium.
  • an ECG heartbeat classification method which includes:
  • the type identifier of each heartbeat segment is set as the type identifier of the corresponding data template.
  • the performing type judgment on the template to be classified in the data template other than the dominant template, and determining the type identifier of the template to be classified includes:
  • the similarity difference is less than a preset difference threshold, it is determined that the type of the template to be classified is identified as the first type.
  • the judging the types of templates to be classified in the data templates other than the dominant template, and determining the type identification of the templates to be classified further includes:
  • the obtaining a scatter diagram of the template to be classified includes:
  • the first interval and the second interval of the heartbeat segment in the template to be classified where the first interval is the distance between the heartbeat segment and the adjacent previous heartbeat segment; the second interval The period is the distance between the heartbeat segment and the next adjacent heartbeat segment;
  • the determining whether the coordinate points in the scatter diagram of the template to be classified are distributed near a preset reference line includes:
  • the method further includes:
  • the ratio is not greater than the preset standard deviation ratio, or the representative waveform width of the template to be classified is not greater than the preset width threshold, it is determined that the type of the template to be classified is identified as the third type.
  • the method before the obtaining the electrocardiogram data, the method further includes:
  • the dividing the electrocardiogram data to obtain multiple heart beat segments includes:
  • the electrocardiogram data is divided according to the R wave position information obtained by the detection, and a plurality of equal-length heart beat segments centered on the R wave position are obtained.
  • an ECG heartbeat classification device which includes:
  • the dividing module is used to obtain electrocardiogram data, divide the electrocardiogram data, and obtain multiple heart beat segments;
  • a matching module which matches the multiple heart beat segments with preset data templates, and determines the data template corresponding to each heart beat segment in the multiple heart beat segments, and the template identifier of the data template;
  • a processing module configured to determine a dominant template among the data templates corresponding to each heartbeat segment, and set the type identifier of the dominant template as the first type;
  • the processing module is further configured to perform type judgment on templates to be classified in the data template other than the dominant template, and determine the type identification of the template to be classified;
  • the setting module is used to set the type identification of each heartbeat segment as the type identification of the corresponding data template.
  • an electronic device including a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the same as in the first aspect and any of them.
  • the memory stores a computer program
  • the processor executes the same as in the first aspect and any of them.
  • a computer storage medium stores one or more instructions, and the one or more instructions are suitable for being loaded by a processor and executed as in the first aspect and any one of them. Possible implementation steps.
  • This application obtains electrocardiogram data, divides the above-mentioned electrocardiogram data, obtains multiple heart beat fragments, matches the multiple heart beat fragments with a preset data template, and determines that each of the multiple heart beat fragments corresponds to each heart beat fragment
  • the data template of the above-mentioned data template, and the template identifier of the above-mentioned data template determine a dominant template in the data template corresponding to each of the above-mentioned heartbeat fragments, set the type identifier of the above-mentioned dominant template as the first type, and then remove the above-mentioned dominant template from the above-mentioned data templates
  • Type judgments for templates other than those to be classified determine the type identification of the above-mentioned template to be classified, and then set the type identification of each heartbeat segment to the type identification of the corresponding data template.
  • FIG. 1 is a schematic flowchart of an ECG heartbeat classification method according to an embodiment of the application
  • FIG. 2 is a schematic flowchart of another ECG heartbeat classification method provided by an embodiment of the application.
  • FIG. 3 is a schematic diagram of the whole process of an ECG heartbeat classification method provided by an embodiment of the application.
  • FIG. 4 is a schematic diagram of a template matching result provided by an embodiment of the application.
  • FIG. 5 is a schematic diagram of a template waveform provided by an embodiment of the application.
  • FIG. 6 is a schematic diagram of a scatter diagram of a template provided by an embodiment of the application.
  • FIG. 7 is a schematic diagram of another template waveform provided by an embodiment of the application.
  • FIG. 8 is a schematic diagram of a scatter diagram of another template provided by an embodiment of the application.
  • FIG. 9 is a schematic structural diagram of an electrocardiogram heartbeat classification device provided by an embodiment of the application.
  • FIG. 10 is a schematic structural diagram of an electronic device provided by an embodiment of this application.
  • FIG. 1 is a schematic flowchart of an ECG heartbeat classification method according to an embodiment of the present application.
  • the method may include:
  • the execution subject of the embodiments of the present application may be an electrocardiogram and heartbeat classification device, which may be an electronic device.
  • the above-mentioned electronic device is a terminal, which may also be called a terminal device, including but not limited to, for example, having a touch-sensitive surface (E.g., touch screen display and/or touch pad) other portable devices such as mobile phones, laptop computers, or tablet computers.
  • a touch-sensitive surface E.g., touch screen display and/or touch pad
  • the above-mentioned devices are not portable communication devices, but desktop computers with touch-sensitive surfaces (for example, touch screen displays and/or touch pads).
  • the above-mentioned ECG heartbeat classification method may be supported and run by a server in response to a user's operation instruction.
  • the electrocardiogram data in the embodiments of the present application may be the electrocardiogram signal collected by the electrocardiogram acquisition device over a period of time.
  • the position of the heartbeat can be determined according to the characteristics of the electrocardiogram, and then the signal can be divided into multiple equal-length segments reflecting the characteristics of the heartbeat .
  • the foregoing division of the foregoing electrocardiogram data to obtain multiple heart beat segments includes:
  • the electrocardiogram data is divided according to the R wave position information obtained by the detection, and a plurality of equal-length heart beat segments centered on the R wave position are obtained.
  • a complete electrocardiogram signal is usually composed of P wave, QRS complex and T wave.
  • QRS complex is the main part of the electrocardiogram, and R wave is dominant in the QRS complex, so the mark position of the heartbeat can be R Wave prevails. It is possible to obtain ECG fragments of equal length centered on the R wave position for template matching.
  • the method may further include:
  • Band-pass filtering is performed on the above-mentioned dual-lead ECG data to obtain the above-mentioned ECG data.
  • Holter is an ECG signal recording system. It can use portable ECG signal acquisition and continuous measurement of the patient's heart activity for up to 24 hours, providing a powerful information aid for doctors to comprehensively analyze the patient's heart condition.
  • the Holter data can be read first, which includes the 24-hour multi-lead Holter data collected by the Holter device, and can also include related parameters, such as:
  • the R wave position (RPos, with sampling point or time as the unit) obtained by QRS wave detection;
  • RR interval (rr_interval, represents the distance between the current heartbeat and the previous adjacent heartbeat, in units of sampling point or time);
  • QRS wave width width, in units of sampling points or time
  • Primary analysis lead number and secondary analysis lead number are Primary analysis lead number and secondary analysis lead number.
  • the primary analysis lead number and the secondary analysis lead number are the two dominant lead data numbers determined by analysis in the above-mentioned multi-lead Holter data.
  • signal preprocessing can be performed: the dual-lead Holter data is constructed according to the above-mentioned primary analysis lead number and the above-mentioned secondary analysis lead number, and then the above-mentioned band-pass filtering is performed to filter out low-frequency baseline drift interference and high-frequency noise.
  • the Holter data is divided into the center of the fragment corresponding to the R wave position, and the ECG data fragment X ⁇ R N*2*W with a length of W, where N means the division is good
  • N means the division is good
  • the number of ECG data, 2 means dual-lead
  • W means single-lead data length.
  • the K-Means method can be used to output template numbers according to ECG data segments, and ECG data segments corresponding to the same template number have similar waveforms. That is, according to the comparison of the similarity between each heartbeat segment and the template, the heartbeat segment is classified under the matching template and becomes the template member of the template.
  • Template member number beat_index ⁇ R temp_num each element represents a list containing member indexes
  • the template represents the heartbeat width temp_width ⁇ R temp_num (the average value of the QRS wave width of the template members);
  • the template represents the pre-beat RR interval temp_prerr ⁇ R temp_num (the average value of the distance between the template members and the corresponding adjacent previous heartbeat);
  • the template represents the post-beat RR interval temp_prerr ⁇ R temp_num (the average value of the distance between the template member and the corresponding next heartbeat);
  • the template represents the waveform temp_wave ⁇ R temp_num*2*W (average value of the waveform of the template members).
  • the above-mentioned dominant template is the most representative template among the data templates corresponding to each of the above-mentioned heartbeat segments.
  • the above step 103 includes: determining, according to the number of heart beat segments corresponding to different data templates in the above data template, the one data template with the largest number of heart beat segments in the data template as the dominant template.
  • the template corresponding to the maximum value can be searched and determined as the dominant template.
  • the dominant template type can be set as the first type.
  • this type can represent the sinus type and is marked as N. You can set the types and logos of different templates according to your needs.
  • step 104 includes:
  • the similarity difference is less than the preset difference threshold, it is determined that the type of the template to be classified is identified as the first type.
  • the dominant template waveform may be acquired from domi_wave ⁇ R temp_wave waveform templates represent information to be classified and the 2 * W template waveform temp_wave (temp_id) ⁇ R 2 * W .
  • abs( ⁇ ) represents the absolute value
  • ppv( ⁇ ) represents the peak-to-peak value, that is, the maximum value-the minimum value.
  • step 104 further includes:
  • the ECG heartbeat classification result is obtained.
  • This application obtains electrocardiogram data, divides the above-mentioned electrocardiogram data, obtains multiple heart beat fragments, matches the multiple heart beat fragments with a preset data template, and determines that each of the multiple heart beat fragments corresponds to each heart beat fragment
  • the data template of the above-mentioned data template, and the template identifier of the above-mentioned data template determine a dominant template in the data template corresponding to each of the above-mentioned heartbeat fragments, set the type identifier of the above-mentioned dominant template as the first type, and then remove the above-mentioned dominant template from the above-mentioned data templates
  • Type judgments for templates other than those to be classified determine the type identification of the above-mentioned template to be classified, and then set the type identification of each heartbeat segment to the type identification of the corresponding data template.
  • FIG. 2 is a schematic flowchart of another ECG heartbeat classification method provided by an embodiment of the present application.
  • the method may be a refinement of step 4 of the embodiment shown in FIG. 1, that is, the method for judging the type of the template to be classified, which may specifically include:
  • the execution subject of the embodiments of the present application may be an electrocardiogram and heartbeat classification device, which may be an electronic device.
  • the above-mentioned electronic device is a terminal, which may also be called a terminal device, including but not limited to, for example, having a touch-sensitive surface (E.g., touch screen display and/or touch pad) other portable devices such as mobile phones, laptop computers, or tablet computers.
  • a touch-sensitive surface E.g., touch screen display and/or touch pad
  • the above-mentioned devices are not portable communication devices, but desktop computers with touch-sensitive surfaces (for example, touch screen displays and/or touch pads).
  • the above-mentioned ECG heartbeat classification method may be supported and run by a server in response to a user's operation instruction.
  • the aforementioned templates to be classified are matched templates other than the determined dominant template in the embodiment shown in FIG. 1.
  • the scatter diagram of the template to be classified can be used to determine whether the coordinate points in the scatter diagram of the template to be classified are distributed near the preset reference line.
  • the former RR interval beat_prerr ⁇ [beat_num(temp_id)] and the later RR interval beat_postrr ⁇ [beat_num(temp_id) can be obtained according to the template number temp_id to be classified, the template member index beat_index, and the heartbeat RR interval rr_interval. )], that is, the above-mentioned first interval and the above-mentioned second interval.
  • the former RR interval can be the abscissa and the latter RR interval can be the ordinate to construct a template member scatter plot (beat_prerr, beat_postrr).
  • Other reference and distance reference rules can be set as needed.
  • the signed distance X ⁇ [beat_num(temp_id)] between each member (heartbeat segment) and the 45-degree line of the zero crossing point can be calculated, and the calculation formula can be as follows:
  • a normal distribution curve can be constructed. Specifically, it is assumed that the distance distribution obeys a normal distribution with a mean of ⁇ and a standard deviation of ⁇ , as shown in the following formula. Set the step size to step, calculate the distance histogram hist(distance), and obtain the statistical number Y(i) corresponding to each distance X(i). The nonlinear least squares method is used to fit the normal distribution curve expressed by the following formula with X and Y, and the estimated mean ⁇ and estimated standard deviation ⁇ are obtained.
  • the offset term ⁇ can be eliminated from the original distance X, and some abnormal points can be deleted using the 3 ⁇ principle to obtain a new distance value Xnew (target distance value).
  • the above-mentioned 3 ⁇ principle is to first assume that a set of test data contains only random errors, calculate and process them to obtain the standard deviation, and determine an interval according to a certain probability. It is considered that any error exceeding this interval is not a random error but a gross error. The data with this error should be eliminated.
  • the standard deviation X_std of the target distance value can be calculated, and it is determined whether the standard deviation is less than the preset standard deviation threshold TX_std. If X_std ⁇ TX_std, it means that the scatter plot of the template to be classified is distributed near the 45-degree line; if X_std is not less than TX_std, it is determined that the scatter plot of the template to be classified is not distributed near the 45-degree line.
  • the type of the template to be classified whose scatter diagram is distributed near the preset reference line can be set to the first type, such as N.
  • the former RR interval beat_prerr ⁇ [beat_num(temp_id)] and the later RR interval beat_postrr ⁇ [beat_num(temp_id) can be obtained according to the template number temp_id to be classified, the template member index beat_index, and the heartbeat RR interval rr_interval.
  • step 211 or step 212 It is determined to perform step 211 or step 212 through the above two comparisons.
  • the type of the template to be classified with the scatter diagram determined in step 210 having a vertical strip-like distribution and a larger waveform width is set to a second type, such as a ventricular type (represented by a V type).
  • the ratio is not greater than the preset standard deviation ratio, or the representative waveform width of the template to be classified is not greater than the preset width threshold, determine the type identification of the template to be classified as the third type.
  • the type of the template to be classified that does not meet the judgment condition of step 210 may be set to a third type, such as other types (represented by type O).
  • the first type, second type, and third type and their corresponding flags in the embodiments of the present application can be set as required.
  • the embodiment of the application obtains the first interval and the second interval of the heart beat segment in the template to be classified, the first interval is the distance between the heart beat segment and the adjacent previous heart beat segment; the second interval The period is the distance between the heartbeat segment and the next adjacent heartbeat segment, and the first interval is the abscissa and the second interval is the ordinate to generate the scatter plot of the template to be classified to obtain the heartbeat segment.
  • Data processing is performed on the scatter diagram of the template to obtain the target distance value between the coordinate points corresponding to the heartbeat segment and the preset reference line, obtain the standard deviation of the target distance value, and determine whether the standard deviation is less than the preset standard deviation threshold; If it is less than, it is determined that the coordinate points in the scatter diagram of the template to be classified are distributed near the preset reference line, and the type identification of the template to be classified is the first type; if it is not less than, the scatter of the template to be classified is determined
  • the coordinate points in the point map are not distributed near the aforementioned preset reference line, and the standard deviation of the first interval and the standard deviation of the second interval of the heartbeat segment in the template to be classified can be obtained; the aforementioned first interval can be obtained.
  • the ratio of the standard deviation of the period to the standard deviation of the second interval determine whether the ratio is greater than the preset standard deviation ratio, and obtain the representative waveform width of the template to be classified, and determine whether the representative waveform width of the template to be classified Is greater than the preset width threshold, if both are greater, the type identification of the template to be classified is determined to be the second type, if the ratio is not greater than the preset standard deviation ratio, or the representative waveform width of the template to be classified is not greater than the preset Set the width threshold to determine the type identification of the template to be classified as the third type.
  • This application uses the concept of template data group to aggregate multiple heartbeats with similar shapes into one template through template matching, and only needs to classify the representative waveforms of the template to complete the classification of all members in the template, and use the nonlinear least squares method
  • the fitting curve can resist the influence of interference to a certain extent, avoid misclassification, and improve the classification accuracy.
  • the ECG heartbeat classification method in the embodiments of the present application can be applied to the ECG heartbeat classification to assist doctors in analyzing the patient's heart condition. It can be introduced in conjunction with the flowchart shown in Figure 3 and Figures 4 to 8.
  • the first step in ECG heartbeat classification is to read Holter data.
  • each Holter data file contains a DAT file and a DATRst file.
  • the former stores 12-lead Holter raw signal data containing 24 hours, and the latter stores Holter data analysis results, including R wave position information and RR. Interval, QRS wave width, sampling rate information fs, and primary analysis lead number and secondary analysis lead number.
  • the 24h Holter data of patient A After the 24h Holter data of patient A is collected with the Holter collection device, it can be processed by the supporting software to output a DAT file and a DATRst file.
  • the segment width W According to the position of the R wave, set the segment width W to 0.1s, where the front and back of the R wave occupies 0.05s.
  • the number of sample points corresponding to the fragment width is 0.1*fs.
  • the Holter data is divided to obtain approximately 110,000 ECG data fragments.
  • the template matching results are obtained, as shown in Figure 4.
  • the templates have been sorted in descending order of the number of members, and the template numbers are 0,1,2,....
  • FIG. 4 For example, refer to a schematic diagram of a template matching result shown in FIG. 4.
  • the figure includes eight templates, numbered from 0-7.
  • the number in the upper right corner of each template image indicates the number of members of the template (corresponding heartbeat segment).
  • template 0 has 104,236 members, the most, so it is determined to be the dominant template.
  • only the numbers 1, 2, and 6 are used as examples to illustrate the execution process of the algorithm.
  • the template type is set to N;
  • the ECG heartbeat classification method based on template matching in the embodiment of this application can identify sinus type (indicated by N type), ventricular type (indicated by V type) and other types (indicated by O type), and can also be based on Different templates and parameters need to be set to identify different types, and there is no restriction on this.
  • the manual feature extraction method uses the doctor's professional clinical experience to design many parameters related to the heartbeat classification, and then input these parameters into the classifier to identify the heartbeat type.
  • This method usually needs to calculate a large number of characteristic parameters, and the algorithm generally takes a long time and is greatly affected by noise interference.
  • the neural network method is a research hotspot in the field of target recognition and classification in recent years. This method can directly input the original signal and the corresponding artificial label, and obtain a satisfactory classification effect by training the neural network model. However, this method needs to spend a lot of time in advance to establish an accurate and diverse training database. If the database does not conform to the real world, the effect of model training will be greatly reduced.
  • the ECG heartbeat classification method combined with the template matching technology in the embodiment of the application is compared with the previous method: without any signal transformation, only three characteristic parameters of the original signal, RR interval and QRS wave width are needed to complete the heartbeat classification. , High computational efficiency;
  • the template matching technology is used to aggregate multiple heartbeats with similar shapes into a template. Only by classifying the representative waveforms of the template, the classification of all members in the template can be completed, and multi-core CPU parallel computing technology can also be used to process different templates at the same time.
  • the algorithm takes a short time;
  • the representative waveform of the template and the corresponding representative parameters are the result of the average value of the members of the template, which can prevent some members from being misclassified due to parameter calculation errors after being affected by noise, thereby improving the anti-interference ability of the classification;
  • the optimization method technology of curve fitting is introduced into the template classification logic to further eliminate the overall deviation of the template and improve the robustness.
  • the electrocardiogram and heartbeat classification device 900 includes:
  • the dividing module 910 is used to obtain electrocardiogram data, divide the above-mentioned electrocardiogram data, and obtain multiple heart beat segments;
  • the matching module 920 matches the multiple heart beat segments with preset data templates, and determines the data template corresponding to each heart beat segment in the multiple heart beat segments, and the template identifier of the data template;
  • the processing module 930 is configured to determine a dominant template among the data templates corresponding to each of the above-mentioned heartbeat segments, and set the type identifier of the above-mentioned dominant template as the first type;
  • the processing module 930 is further configured to perform type judgment on templates to be classified in the data templates other than the dominant template, and determine the type identification of the template to be classified;
  • the setting module 940 is configured to set the type identification of each heartbeat segment described above as the type identification of the corresponding data template.
  • the steps involved in the methods shown in FIG. 1 and FIG. 2 may all be executed by each module in the electrocardiogram heartbeat classification device 900 shown in FIG. 9, and will not be repeated here.
  • the electrocardiogram and heartbeat classification device 900 in the embodiment of the present application obtains the electrocardiogram data and divides the above-mentioned electrocardiogram data to obtain a plurality of heartbeat fragments, and the plurality of heartbeat fragments are matched with a preset data template to determine the above
  • the data template corresponding to each heartbeat segment of the multiple heartbeat segments, and the template identifier of the aforementioned data template determine a dominant template in the data template corresponding to each of the aforementioned heartbeat segments, and set the type identifier of the aforementioned dominant template as the first type , And then perform type judgment on the template to be classified in the above data template except for the dominant template, determine the type identification of the template to be classified, and then set the type identification of each heartbeat segment to the type identification of the corresponding data template, and pass Template matching aggregates multiple heartbeats with similar shapes into one template. It only needs to classify the representative waveforms of the template to complete the classification of all members in the template. Different template data can be processed in parallel,
  • the electronic device 1000 includes at least a processor 1001, an input device 1002, an output device 1003, and a computer storage medium 1004.
  • the processor 1001, the input device 1002, the output device 1003, and the computer storage medium 1004 in the terminal may be connected by a bus or other methods.
  • the computer storage medium 1004 may be stored in the memory of the terminal.
  • the computer storage medium 1004 is used to store computer programs, the computer programs include program instructions, and the processor 1001 is used to execute the program instructions stored in the computer storage medium 1004.
  • the processor 1001 (or CPU (Central Processing Unit, central processing unit)) is the computing core and control core of the terminal. It is suitable for implementing one or more instructions, and specifically for loading and executing one or more instructions to achieve corresponding Method flow or corresponding functions; in one embodiment, the above-mentioned processor 1001 in the embodiment of the present application may be used to perform a series of processing, including the method in the embodiment shown in FIG. 1 and FIG. 2 and so on.
  • the embodiment of the present application also provides a computer storage medium (Memory).
  • the above-mentioned computer storage medium is a memory device in a terminal for storing programs and data. It can be understood that the computer storage medium herein may include a built-in storage medium in the terminal, and of course, may also include an extended storage medium supported by the terminal.
  • the computer storage medium provides storage space, and the storage space stores the operating system of the terminal.
  • one or more instructions suitable for being loaded and executed by the processor 1001 are stored in the storage space, and these instructions may be one or more computer programs (including program codes).
  • the computer storage medium here can be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as at least one disk memory; optionally, it can also be at least one located far away from the aforementioned processor.
  • Computer storage media can be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as at least one disk memory; optionally, it can also be at least one located far away from the aforementioned processor.
  • the processor 1001 can load and execute one or more instructions stored in the computer storage medium to implement the corresponding steps in the foregoing embodiment; in specific implementation, one or more instructions in the computer storage medium may be The processor 1001 loads and executes any steps of the method in FIG. 1 and/or FIG. 2, which will not be repeated here.
  • the disclosed system, device, and method can be implemented in other ways.
  • the division of the modules is only a logical function division, and there can be other divisions in actual implementation.
  • multiple modules or components can be combined or integrated into another system, or some features can be ignored or not. implement.
  • the displayed or discussed mutual coupling, or direct coupling, or communication connection may be indirect coupling or communication connection through some interfaces, devices or modules, and may be in electrical, mechanical, or other forms.
  • modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, that is, they may be located in one place, or they may be distributed to multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the objectives of the solutions of the embodiments.
  • the computer may be implemented in whole or in part by software, hardware, firmware, or any combination thereof.
  • software it can be implemented in the form of a computer program product in whole or in part.
  • the computer program product includes one or more computer instructions.
  • the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.
  • the computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium.
  • the computer instructions can be sent from one website, computer, server, or data center to another via wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.) A website, computer, server or data center for transmission.
  • the computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or data center integrated with one or more available media.
  • the usable medium can be a read-only memory (ROM), or a random access memory (RAM), or a magnetic medium, such as a floppy disk, a hard disk, a magnetic tape, a magnetic disk, or an optical medium, for example, Digital versatile disc (DVD), or semiconductor media, for example, solid state disk (SSD), etc.
  • ROM read-only memory
  • RAM random access memory
  • magnetic medium such as a floppy disk, a hard disk, a magnetic tape, a magnetic disk, or an optical medium, for example, Digital versatile disc (DVD), or semiconductor media, for example, solid state disk (SSD), etc.

Landscapes

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

Abstract

一种心电图心搏分类方法、装置、电子设备和介质，其中心电图心搏分类方法包括：获取心电图数据，对心电图数据进行划分，获得多个心搏片段（101）；将多个心搏片段分别与预设的数据模板匹配，确定多个心搏片段中各个心搏片段对应的数据模板以及模板标识（102）；在各个心搏片段对应的数据模板中确定一个主导模板，设置主导模板的类型标识为第一类型（103）；对数据模板中除主导模板以外的待分类模板进行类型判断，确定待分类模板的类型标识（104）；将各个心搏片段的类型标识设置为对应的数据模板的类型标识（105），多个具有相似形态的心搏聚合为一个模板处理，快速完成心搏分类，处理效率高。

Description

心电图心搏分类方法、装置、电子设备和介质 技术领域
本发明涉及心电图数据处理技术领域，尤其是涉及一种心电图心搏分类方法、装置、电子设备和介质。
背景技术
为了更快捷清楚地对采集的心电图数据进行分析，开发了越来越多的计算机辅助诊断心电图软件。其中，心电图心搏分类定义为对心搏进行类型识别。心电图软件可通过采用不同算法自动识别心搏的类型，为医生提供有关模板所包含的成员心搏是否正常的信息，可提高医生的诊断效率和准确性。
目前心电图心搏分类可以使用人工提取特征法。利用专家知识，设计许多与心搏类型相关的参数，包括形态学特征、统计学特征、频率域特征、相空间特征、时域特征、信息熵特征和心律变异性特征等，然后借助机器学习算法对以上特征参数进行综合计算，输出心搏类型标记。该方法需要借助专业的医生专业临床知识，并且在获取各种特征参数时计算量大；而通过训练神经网络模型进行分类处理的方式需要采集大量的临床数据作样本，开发难度较大且耗时较长。
申请内容
本申请提供了一种心电图心搏分类方法、装置、电子设备和介质。
第一方面，提供了一种心电图心搏分类方法，包括：
获取心电图数据，对所述心电图数据进行划分，获得多个心搏片段；
将所述多个心搏片段分别与预设的数据模板匹配，确定所述多个心搏片段中各个心搏片段对应的数据模板，以及所述数据模板的模板标识；
在所述各个心搏片段对应的数据模板中确定一个主导模板，设置所述主导模板的类型标识为第一类型；
对所述数据模板中除所述主导模板以外的待分类模板进行类型判断，确定所述待分类模板的类型标识；
将所述各个心搏片段的类型标识设置为对应的数据模板的类型标识。
在一种可选的实施方式中，所述对所述数据模板中除所述主导模板以外的待分类模板进行类型判断，确定所述待分类模板的类型标识包括：
获取各个所述待分类模板的波形与所述主导模板的波形的相似度差值；
若所述相似度差值小于预设差值阈值，确定所述待分类模板的类型标识为所述第一类型。
在一种可选的实施方式中，所述对所述数据模板中除所述主导模板以外的待分类模板进行类型判断，确定所述待分类模板的类型标识，还包括：
获取所述待分类模板的散点图，判断所述待分类模板的散点图中的坐标点是否分布在预设参考线附近；
若是，确定所述待分类模板的类型标识为所述第一类型。
在一种可选的实施方式中，所述获取所述待分类模板的散点图包括：
获取所述待分类模板中的心搏片段的第一间期和第二间期，所述第一间期为所述心搏片段与相邻前一个心搏片段的间距；所述第二间期为所述心搏片段与相邻后一个心搏片段的间距；
以所述第一间期为横坐标，以所述第二间期为纵坐标，生成所述待分类模板的散点图。
在一种可选的实施方式中，所述判断所述待分类模板的散点图中的坐标点是否分布在预设参考线附近，包括：
获取所述心搏片段对应的坐标点与预设参考线的有符号距离；
获取所述有符号距离分布服从正态分布情况下的估计均值和估计标准差；
根据所述估计均值和所述估计标准差对所述待分类模板的散点图进行数据处理，获得所述心搏片段对应的坐标点与预设参考线的目标距离值；
获取所述目标距离值的标准差，判断所述标准差是否小于预设的标准差阈值；
若小于，确定所述待分类模板的散点图中的坐标点分布在所述预设参考线附近；若不 小于，确定所述待分类模板的散点图中的坐标点未分布在所述预设参考线附近。
在一种可选的实施方式中，所述确定所述待分类模板的散点图中的坐标点未分布在所述预设参考线附近之后，所述方法还包括：
获取所述待分类模板中的心搏片段的第一间期的标准差和第二间期的标准差；获得所述第一间期的标准差和所述第二间期的标准差的比值；
判断所述比值是否大于预设的标准差比值；
获取所述待分类模板的代表波形宽度，判断所述待分类模板的代表波形宽度是否大于预设的宽度阈值；
若均大于，确定所述待分类模板的类型标识为第二类型；
若所述比值不大于所述预设的标准差比值，或者，所述待分类模板的代表波形宽度不大于预设的宽度阈值，确定所述待分类模板的类型标识为第三类型。
在一种可选的实施方式中，所述获取心电图数据之前，所述方法还包括：
获取采集的多导联心电图数据；根据所述多导联心电图数据中的主分析导联编号和次分析导联编号构建双导联心电图数据；
对所述双导联心电图数据进行带通滤波，获得所述心电图数据；
所述对所述心电图数据进行划分，获得多个心搏片段包括：
根据检测获得的R波位置信息对所述心电图数据进行划分，获得多个以所述R波位置为中心的等长度的心搏片段。
第二方面，提供了一种心电图心搏分类装置，包括：
划分模块，用于获取心电图数据，对所述心电图数据进行划分，获得多个心搏片段；
匹配模块，将所述多个心搏片段分别与预设的数据模板匹配，确定所述多个心搏片段中各个心搏片段对应的数据模板，以及所述数据模板的模板标识；
处理模块，用于在所述各个心搏片段对应的数据模板中确定一个主导模板，设置所述主导模板的类型标识为第一类型；
所述处理模块还用于，对所述数据模板中除所述主导模板以外的待分类模板进行类型判断，确定所述待分类模板的类型标识；
设置模块，用于将所述各个心搏片段的类型标识设置为对应的数据模板的类型标识。
第三方面，提供了一种电子设备，包括存储器和处理器，所述存储器存储有计算机程序，所述计算机程序被所述处理器执行时，使得所述处理器执行如第一方面及其任一种可能的实现方式的步骤。
第四方面，提供了一种计算机存储介质，所述计算机存储介质存储有一条或多条指令，所述一条或多条指令适于由处理器加载并执行如上述第一方面及其任一种可能的实现方式的步骤。
实施本申请实施例，将具有如下有益效果：
本申请通过获取心电图数据，对上述心电图数据进行划分，获得多个心搏片段，将上述多个心搏片段分别与预设的数据模板匹配，确定上述多个心搏片段中各个心搏片段对应的数据模板，以及上述数据模板的模板标识，在上述各个心搏片段对应的数据模板中确定一个主导模板，设置上述主导模板的类型标识为第一类型，再对上述数据模板中除上述主导模板以外的待分类模板进行类型判断，确定上述待分类模板的类型标识，然后将上述各个心搏片段的类型标识设置为对应的数据模板的类型标识，通过模板匹配将多个具有相似形态的心搏聚合为一个模板，仅需对模板代表波形分类即可完成模板内所有成员的分类，可以并行处理不同的模板数据，耗时更短，并且无需复杂的信号变换，处理效率高。
附图说明
为了更清楚地说明本申请实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
其中：
图1为本申请实施例提供的一种心电图心搏分类方法的流程示意图；
图2为本申请实施例提供的另一种心电图心搏分类方法的流程示意图；
图3为本申请实施例提供的一种心电图心搏分类方法全流程示意图；
图4为本申请实施例提供的一种模板匹配结果示意图；
图5为本申请实施例提供的一种模板波形示意图；
图6为本申请实施例提供的一种模板的散点图示意图；
图7为本申请实施例提供的另一种模板波形示意图；
图8为本申请实施例提供的另一种模板的散点图示意图；
图9为本申请实施例提供的一种心电图心搏分类装置的结构示意图；
图10为本申请实施例提供的一种电子设备的结构示意图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
本申请的说明书和权利要求书及上述附图中的术语“第一”、“第二”等是用于区别不同对象，而不是用于描述特定顺序。此外，术语“包括”和“具有”以及它们任何变形，意图在于覆盖不排他的包含。例如包含了一系列步骤或单元的过程、方法、系统、产品或设备没有限定于已列出的步骤或单元，而是可选地还包括没有列出的步骤或单元，或可选地还包括对于这些过程、方法、产品或设备固有的其他步骤或单元。
在本文中提及“实施例”意味着，结合实施例描述的特定特征、结构或特性可以包含在本申请的至少一个实施例中。在说明书中的各个位置出现该短语并不一定均是指相同的实施例，也不是与其它实施例互斥的独立的或备选的实施例。本领域技术人员显式地和隐式地理解的是，本文所描述的实施例可以与其它实施例相结合。
下面结合本申请实施例中的附图对本申请实施例进行描述。
请参阅图1，图1是本申请实施例提供的一种心电图心搏分类方法的流程示意图。该方法可包括：
101、获取心电图数据，对上述心电图数据进行划分，获得多个心搏片段。
本申请实施例的执行主体可以为一种心电图心搏分类装置，可以为电子设备，具体实现中，上述电子设备为一种终端，也可称为终端设备，包括但不限于诸如具有触摸敏感表面(例如，触摸屏显示器和/或触摸板)的移动电话、膝上型计算机或平板计算机之类的其它便携式设备。还应当理解的是，在某些实施例中，上述设备并非便携式通信设备，而是具有触摸敏感表面(例如，触摸屏显示器和/或触摸板)的台式计算机。在一种可选的实施方式中，上述心电图心搏分类方法可以响应于用户的操作指令，由服务器支撑运行。
本申请实施例中的心电图数据可以是心电图采集设备所采集的一段时间内的心电图信号，一般可以根据心电图的特性确定心搏位置，进而将该信号划分为多个反应心搏特征的等长片段。
可选的，上述对上述心电图数据进行划分，获得多个心搏片段，包括：
根据检测获得的R波位置信息对上述心电图数据进行划分，获得多个以上述R波位置为中心的等长度的心搏片段。
一个完整的心电图信号通常由P波、QRS波群和T波构成，其中QRS波群是心电图的主体部分，而R波又在QRS波群中占主导地位，因此心搏的标记位置可以以R波为准。可以获取以R波位置为中心的等长度的ECG片段，以进行模板匹配。
可选的，上述步骤101之前，该方法还可以包括：
获取采集的多导联心电图数据；根据上述多导联心电图数据中的主分析导联编号和次分析导联编号构建双导联心电图数据；
对上述双导联心电图数据进行带通滤波，获得上述心电图数据。
本申请实施例中可以使用Holter采集的多导联心电图数据进行处理。Holter是一种心电信号记录系统，它可以通过采用便携心电信号采集和对患者的心脏活动进行长达24小时的持续测量，为医生全面分析患者心脏病情提供了强有力的信息辅助。
具体的，可以首先读取Holter数据，该数据包括由Holter设备采集的24小时多导联Holter数据，还可以包括相关参数，如：
经过QRS波检测得到的R波位置(RPos，以采样点或时间为单位)；
RR间期(rr_interval，表示当前心搏与相邻前一个心搏的间距，以采样点或时间为单 位)；
QRS波宽度(width，以采样点或时间为单位)；
Holter设备采样率fs；
主分析导联编号和次分析导联编号。
其中，上述主分析导联编号和次分析导联编号是在上述多导联Holter数据中分析确定的占主导地位的两个导联数据编号。进一步的，可以进行信号预处理：根据上述主分析导联编号和上述次分析导联编号构建双导联Holter数据，然后经过上述带通滤波，滤除低频的基线漂移干扰和高频噪声。
对于前述ECG划分，具体可以为：根据R波位置RPos，将Holter数据划分为片段中心对应该R波位置，长度为W的ECG数据片段X∈R N*2*W，其中N表示划分好的ECG数据个数，2表示双导联，W表示单导联数据长度。
102、将上述多个心搏片段分别与预设的数据模板匹配，确定上述多个心搏片段中各个心搏片段对应的数据模板，以及上述数据模板的模板标识。
可以采用K-Means方法，根据ECG数据片段输出模板编号，相同模板编号对应的ECG数据片段具有相似的波形。即根据比较各心搏片段分别与模板的相似度，将心搏片段归类到匹配的模板下，成为该模板的模板成员。
还可以根据模板匹配结果，获取用于心搏分类的信息。如：
(1)模板个数temp_num；
(2)模板成员编号beat_index∈R temp_num(每个元素代表一个包含成员索引的列表)；
(3)模板成员个数beat_num∈R temp_num；
(4)模板代表心搏宽度temp_width∈R temp_num(模板成员QRS波宽度均值)；
(5)模板代表心搏前RR间期temp_prerr∈R temp_num(模板成员与对应相邻前一个心搏的间距均值)；
(6)模板代表心搏后RR间期temp_prerr∈R temp_num(模板成员与对应相邻后一个心搏的间距均值)；
(7)模板代表波形temp_wave∈R temp_num*2*W(模板成员波形均值)。
103、在上述各个心搏片段对应的数据模板中确定一个主导模板，设置上述主导模板的类型标识为第一类型。
上述主导模板是上述各个心搏片段对应的数据模板中最具代表性的一个模板。在一种实施方式中，上述步骤103包括：根据上述数据模板中不同数据模板对应的心搏片段数量，确定上述数据模板中上述心搏片段数量最多的一个数据模板为上述主导模板。
可以根据不同模板的模板成员个数，搜索最大值对应的模板确定为主导模板。在确定主导模板后，可以设置主导模板类型为第一类型，比如该类型可以表示窦性类型，标识为N。可以根据需要设置不同模板的类型及标识。
104、对上述数据模板中除上述主导模板以外的待分类模板进行类型判断，确定上述待分类模板的类型标识。
具体的，对于除主导模板外的待分类模板，可以继续进行类型判断。可以初始化除该主导模板外的待分类模板编号temp_id＝1:temp_num-1：对除主导模板以外的其他不同模板进行遍历，依次判断模板类型。
可选的，上述步骤104包括：
获取各个上述待分类模板的波形与上述主导模板的波形的相似度差值；
若上述相似度差值小于预设差值阈值，确定上述待分类模板的类型标识为上述第一类型。
可以根据波形相似度判断待分类模板波形与主导模板波形是否相似。在一种可选的实施方式中，可以从模板代表波形temp_wave信息中获取主导模板波形domi_wave∈R 2*W和待分类模板波形temp_wave(temp_id)∈R 2*W。依次对每个导联分别计算两个模板波形的累计差值Sdiff1和Sdiff2，计算公式如下所示：
Figure PCTCN2020126352-appb-000001
其中abs(·)表示取绝对值，ppv(·)表示取峰峰值，即最大值-最小值。则待分类模板波形与主导模板波形的累计差值S diff＝(S diff_1+S diff_2)/2。设置累计差值阈值(上述预设差值阈值)为TSdiff，若Sdiff<TSdiff，则判断待分类模板波形与主导模板波形相似，也可确定该待分类模板的类型标识为上述第一类型；否则即为不相似，该待分类模板可以进一步 进行类型判断。
在一种可选的实施方式中，上述步骤104还包括：
获取上述待分类模板的散点图，判断上述待分类模板的散点图中的坐标点是否分布在预设参考线附近；
若是，确定上述待分类模板的类型标识为上述第一类型。
对于上述方法和不同模板的类型分析方法还可以见图2所示实施例中的具体描述。
105、将上述各个心搏片段的类型标识设置为对应的数据模板的类型标识。
通过将模板中的成员类型设置为对应的模板类型，更新模板成员类型，获得心电图心搏分类结果。
本申请通过获取心电图数据，对上述心电图数据进行划分，获得多个心搏片段，将上述多个心搏片段分别与预设的数据模板匹配，确定上述多个心搏片段中各个心搏片段对应的数据模板，以及上述数据模板的模板标识，在上述各个心搏片段对应的数据模板中确定一个主导模板，设置上述主导模板的类型标识为第一类型，再对上述数据模板中除上述主导模板以外的待分类模板进行类型判断，确定上述待分类模板的类型标识，然后将上述各个心搏片段的类型标识设置为对应的数据模板的类型标识，通过模板匹配将多个具有相似形态的心搏聚合为一个模板，仅需对模板代表波形分类即可完成模板内所有成员的分类，可以并行处理不同的模板数据，耗时更短，并且无需复杂的信号变换，处理效率高。
请参阅图2，图2是本申请实施例提供的另一种心电图心搏分类方法的流程示意图。如图2所示，该方法可以为图1所示实施例的步骤4的细化，即对待分类模板进行类型判断的方法，具体可以包括：
201、获取待分类模板中的心搏片段的第一间期和第二间期，上述第一间期为上述心搏片段与相邻前一个心搏片段的间距；上述第二间期为上述心搏片段与相邻后一个心搏片段的间距。
本申请实施例的执行主体可以为一种心电图心搏分类装置，可以为电子设备，具体实现中，上述电子设备为一种终端，也可称为终端设备，包括但不限于诸如具有触摸敏感表面(例如，触摸屏显示器和/或触摸板)的移动电话、膝上型计算机或平板计算机之类的其 它便携式设备。还应当理解的是，在某些实施例中，上述设备并非便携式通信设备，而是具有触摸敏感表面(例如，触摸屏显示器和/或触摸板)的台式计算机。在一种可选的实施方式中，上述心电图心搏分类方法可以响应于用户的操作指令，由服务器支撑运行。
其中，上述待分类模板为如图1所示实施例中除确定的主导模板以外的其他匹配的模板。本申请实施例可以通过待分类模板的散点图，判断所述待分类模板的散点图中的坐标点是否分布在预设参考线附近。具体的，可以根据待分类模板编号temp_id、模板成员索引beat_index和心搏RR间期rr_interval，获取模板成员的前RR间期beat_prerr∈[beat_num(temp_id)]和后RR间期beat_postrr∈[beat_num(temp_id)]，即上述第一间期和上述第二间期。
202、以上述第一间期为横坐标，以上述第二间期为纵坐标，生成上述待分类模板的散点图。
进一步的，可以以前RR间期为横坐标，后RR间期为纵坐标构建模板成员散点图(beat_prerr,beat_postrr)。
203、获取上述心搏片段对应的坐标点与预设参考线的有符号距离。
本申请实施例中提到的预设参考线可以为45度线，比如表示为y＝x+b，其中b为常数。可以根据需要设置其他的参考和距离参考规则。具体的，可以计算每个成员(心搏片段)与过零点的45度线的有符号距离X∈[beat_num(temp_id)]，计算公式可如下所示：
Figure PCTCN2020126352-appb-000002
204、获取上述有符号距离分布服从正态分布情况下的估计均值和估计标准差。
在一种实施方式中，可以构建正态分布曲线。具体的，假设距离分布服从均值为μ，标准差为σ的正态分布，如下公式所示。设置步长为step，统计距离直方图hist(distance)，获取每个距离X(i)对应的统计个数Y(i)。采用非线性最小二乘法利用X和Y拟合由以下公式表示的正态分布曲线，得到估计均值μ和估计标准差σ。
Figure PCTCN2020126352-appb-000003
205、根据上述估计均值和上述估计标准差对上述待分类模板的散点图进行数据处理， 获得上述心搏片段对应的坐标点与预设参考线的目标距离值。
具体的，可以对原始距离X消除偏移项μ，并采用3σ原理删除部分异常点后，得到新的距离值Xnew(目标距离值)。上述3σ原理是先假设一组检测数据只含随机误差，对其进行计算处理得到标准偏差，按一定概率确定一个区间，认为凡超过这个区间的误差，就不属于随机误差而是粗大误差，含有该误差的数据应予以剔除。
206、获取上述目标距离值的标准差，判断上述标准差是否小于预设的标准差阈值；若小于，确定上述待分类模板的散点图中的坐标点分布在上述预设参考线附近；若不小于，确定上述待分类模板的散点图中的坐标点未分布在上述预设参考线附近。
具体的，可以计算目标距离值的标准差X_std，判断该标准差是否小于预设的标准差阈值TX_std。若X_std<TX_std，说明待分类模板散点图分布在45度线附近；若X_std不小于TX_std，确定待分类模板的散点图未分布在45度线附近。
207、确定上述待分类模板的类型标识为第一类型。
与判断与主导模板波形相似的待分类模板类似的，可以将散点图分布在预设参考线附近的待分类模板的类型设置为第一类型，如N。
208、获取上述待分类模板中的心搏片段的第一间期的标准差和第二间期的标准差；获得上述第一间期的标准差和上述第二间期的标准差的比值。
209、判断上述比值是否大于预设的标准差比值。
210、获取上述待分类模板的代表波形宽度，判断上述待分类模板的代表波形宽度是否大于预设的宽度阈值。
进一步的，可以判断剩下的待分类模板的散点图是否呈竖带状分布且波形宽度较大，来进行分类。具体的，可以根据待分类模板编号temp_id、模板成员索引beat_index和心搏RR间期rr_interval，获取模板成员的前RR间期beat_prerr∈[beat_num(temp_id)]和后RR间期beat_postrr∈[beat_num(temp_id)]，分别计算前RR间期的标准差beat_prerr_std和后RR间期的标准差beat_postrr_std，并计算两者的比值Ratio＝beat_postrr_std/beat_prerr_std。同时获取待分类模板代表波形宽度Width。设置预设的标准差比值TRatio和预设的宽度阈值TWidth。若Ratio>TRatio，判断待分类模板散点图呈 竖带状分布。若Width>TWidth，判断待分类模板波形宽度较大。
通过上述两项比较确定执行步骤211或者步骤212。
211、若均大于，确定上述待分类模板的类型标识为第二类型。
将步骤210判断出的散点图呈竖带状分布且波形宽度较大的待分类模板的类型设置为第二类型，比如室性类型(以V类表示)。
212、若上述比值不大于上述预设的标准差比值，或者，上述待分类模板的代表波形宽度不大于预设的宽度阈值，确定上述待分类模板的类型标识为第三类型。
可以将不满足步骤210判断条件的待分类模板的类型设置为第三类型，如其他类型(以O类表示)。本申请实施例中的第一类型、第二类型和第三类型及其对应标志可以根据需要进行设置。
本申请实施例通过获取待分类模板中的心搏片段的第一间期和第二间期，上述第一间期为上述心搏片段与相邻前一个心搏片段的间距；上述第二间期为上述心搏片段与相邻后一个心搏片段的间距，以上述第一间期为横坐标，以上述第二间期为纵坐标，生成上述待分类模板的散点图，获取上述心搏片段对应的坐标点与预设参考线的有符号距离，再获取上述有符号距离分布服从正态分布情况下的估计均值和估计标准差，根据上述估计均值和上述估计标准差对上述待分类模板的散点图进行数据处理，获得上述心搏片段对应的坐标点与预设参考线的目标距离值，获取上述目标距离值的标准差，判断上述标准差是否小于预设的标准差阈值；若小于，确定上述待分类模板的散点图中的坐标点分布在上述预设参考线附近，可以确定上述待分类模板的类型标识为第一类型；若不小于，确定上述待分类模板的散点图中的坐标点未分布在上述预设参考线附近，进一步可以获取上述待分类模板中的心搏片段的第一间期的标准差和第二间期的标准差；获得上述第一间期的标准差和上述第二间期的标准差的比值，判断上述比值是否大于预设的标准差比值，并且，获取上述待分类模板的代表波形宽度，判断上述待分类模板的代表波形宽度是否大于预设的宽度阈值，若均大于，确定上述待分类模板的类型标识为第二类型，若上述比值不大于上述预设的标准差比值，或者，上述待分类模板的代表波形宽度不大于预设的宽度阈值，确定上述待分类模板的类型标识为第三类型。本申请利用模板数据群的概念，通过模板匹配将多个 具有相似形态的心搏聚合为一个模板，仅需对模板代表波形分类即可完成模板内所有成员的分类，并借助非线性最小二乘法拟合曲线，能在一定程度抵抗干扰的影响，避免误分类，提高了分类准确度。
可以参见图3所示的一种心电图心搏分类方法全流程示意图，如图3所示，在步骤4模板匹配之后可以通过步骤8、9、11判断待分类模板类型，从而确定其中心搏片段的类型。具体步骤内容可以参考前述实施例中的描述，此处不再赘述。
本申请实施例中的心电图心搏分类方法可以应用于辅助医生分析患者心脏病情的心电图心搏分类。可以结合图3所示的流程示意图和图4-图8进行介绍。
心电图心搏分类的第一步是读取Holter数据。具体的，每份Holter数据文件包含一个DAT文件和一个DATRst文件，前者存储了包含24小时的12导联Holter原始信号数据，后者存储的是Holter数据分析结果信息，包括R波位置信息、RR间期、QRS波宽度、采样率信息fs以及主分析导联编号和次分析导联编号。
借助Holter采集设备采集到了患者A的24h的Holter数据后，可借由配套软件处理后，输出一个DAT文件和一个DATRst文件。
首先读取DAT文件，获取12导联的Holter数据后，提取由主分析导联(本例中为V5)和次分析导联(本例中为V4)构成的双导联Holter数据。
然后将双导联Holter数据经过0.5Hz-50Hz的带通滤波器，输出得到滤除基线漂移和高频噪声的带通滤波信号。
依据R波位置，设置片段宽度W为0.1s，其中R波前后各占0.05s。片段宽度对应的样本点个数为0.1*fs。对Holter数据进行划分，得到大约11万个ECG数据片段。
采用K-Means聚类后，得到模板匹配结果，如图4所示，此时模板已按成员个数从大到小顺序排序，模板编号依次为0,1,2,...。
将模板成员个数最多的模板(即编号为0)设置为主导模板，模板类型设置为N。
依次遍历剩下的待分类模板，识别每个模板的类型。
举例来讲，可以参考图4所示的一种模板匹配结果示意图。图中包括八个模板，编号分别为0-7，每个模板图像右上角的数字表示该模板成员(对应的心搏片段)数量，其中模 板0的成员有104236，最多，因此被确定为主导模板。此处仅以编号为1、2和6的编号举例说明算法的执行过程。
(1)编号为1的模板波形如图5，在图3中步骤8计算待分类模板波形与主导模板波形的累计差值Sdiff＝0.0867，小于给定阈值TSdiff＝0.13，因此在步骤10中将该模板类型设置为N；
(2)编号为2的模板散点图分布如图6所示。当该模板因不满足图3中步骤8和9的判断后，开始进行步骤11的条件判断。可获知模板波形宽度Width＝160ms，后RR间期标准差与前RR间期标准差比值Ratio＝3.147。由于Width大于宽度阈值TWitdh＝120ms，Ratio大于宽度阈值TRatio＝3，因此在第12步中将该模板类型设置为V；
(3)编号为6的模板波形图和散点图分别如7和图8所示。从图7可见，由于受到干扰影响，R波位置并没有正确地定在如图5所示的R波上，而是在S波上，使得在步骤8中判断该模板波形与主导模板波形不相似。然后在步骤9中，算法首先获取模板成员的前RR间期和后RR间期，并计算得到与45度线的距离X＝(-181.019,-181.019,-1991.21,181.019,271.529,…)，采用曲线拟合方法求解得到正态分布系数μ＝-16.679，σ＝128.0。消除原始距离X的偏移项，并根据3σ原理删除异常点后，得到新的无偏差距离Xnew＝(-164.34,-164.34,197.699,288.208,288.208,…)。计算距离标准差X_std＝161.972，小于给定阈值TX_std＝384.0，因此在步骤10中将该模板类型设置为N。
遍历完所有待分类模板后，更新每个模板包含的成员的类型，则完成心电图心搏分类。本申请实施例中的基于模板匹配的心电图心搏分类方法，可识别窦性类型(以N类表示)、室性类型(以V类表示)和其他类型(以O类表示)，也可以根据需要设置不同的模板和参数来识别不同的类型，对此不做限制。
一般地，人工提取特征法是利用医生的专业临床经验，设计许多与心搏分类相关的参数，然后将这些参数输入到分类器进行心搏类型的识别。该方法通常需要计算大量的特征参数，算法耗时一般较长，而且受噪声干扰影响较大。而神经网络法是近年来目标识别与分类领域的研究热点，该方法可直接输入原始信号和对应人工标签，通过训练神经网络模型获得满意的分类效果。但是这种方法需要事先花费大量时间建立准确且具备多样性的训 练数据库，若数据库与真实世界不符合，则模型训练效果将大打折扣。
本申请实施例中的心电图心搏分类方法结合模板匹配技术，相比前述方法而言：无需任何信号变换，仅需原始信号、RR间期和QRS波宽度三个特征参数即可完成心搏分类，计算效率高；
采用模板匹配技术将多个具有相似形态的心搏聚合为一个模板，仅需对模板代表波形分类即可完成模板内所有成员的分类，并且还可以采用多核CPU并行运算技术同时处理不同的模板，算法耗时短；
模板代表波形以及相应代表参数是模板成员平均值的结果，避免部分成员受噪声影响后参数计算错误而出现误分类的情形，从而可以提升分类的抗干扰能力；
在模板分类逻辑中引入曲线拟合的最优化方法技术，进一步消除模板的整体偏差，提高了鲁棒性。
基于上述心电图心搏分类方法实施例的描述，本申请实施例还公开了一种心电图心搏分类装置。请参见图9，心电图心搏分类装置900包括：
划分模块910，用于获取心电图数据，对上述心电图数据进行划分，获得多个心搏片段；
匹配模块920，将上述多个心搏片段分别与预设的数据模板匹配，确定上述多个心搏片段中各个心搏片段对应的数据模板，以及上述数据模板的模板标识；
处理模块930，用于在上述各个心搏片段对应的数据模板中确定一个主导模板，设置上述主导模板的类型标识为第一类型；
上述处理模块930还用于，对上述数据模板中除上述主导模板以外的待分类模板进行类型判断，确定上述待分类模板的类型标识；
设置模块940，用于将上述各个心搏片段的类型标识设置为对应的数据模板的类型标识。
根据本申请的一个实施例，图1和图2所示的方法所涉及的各个步骤均可以是由图9所示的心电图心搏分类装置900中的各个模块执行的，此处不再赘述。
本申请实施例中的心电图心搏分类装置900，通过获取心电图数据，对上述心电图数 据进行划分，获得多个心搏片段，将上述多个心搏片段分别与预设的数据模板匹配，确定上述多个心搏片段中各个心搏片段对应的数据模板，以及上述数据模板的模板标识，在上述各个心搏片段对应的数据模板中确定一个主导模板，设置上述主导模板的类型标识为第一类型，再对上述数据模板中除上述主导模板以外的待分类模板进行类型判断，确定上述待分类模板的类型标识，然后将上述各个心搏片段的类型标识设置为对应的数据模板的类型标识，通过模板匹配将多个具有相似形态的心搏聚合为一个模板，仅需对模板代表波形分类即可完成模板内所有成员的分类，可以并行处理不同的模板数据，耗时更短，并且无需复杂的信号变换，处理效率高。
基于上述方法实施例以及装置实施例的描述，本申请实施例还提供一种电子设备。请参见图10，该电子设备1000至少包括处理器1001、输入设备1002、输出设备1003以及计算机存储介质1004。其中，终端内的处理器1001、输入设备1002、输出设备1003以及计算机存储介质1004可通过总线或其他方式连接。
计算机存储介质1004可以存储在终端的存储器中，上述计算机存储介质1004用于存储计算机程序，上述计算机程序包括程序指令，上述处理器1001用于执行上述计算机存储介质1004存储的程序指令。处理器1001(或称CPU(Central Processing Unit，中央处理器))是终端的计算核心以及控制核心，其适于实现一条或多条指令，具体适于加载并执行一条或多条指令从而实现相应方法流程或相应功能；在一个实施例中，本申请实施例上述的处理器1001可以用于进行一系列的处理，包括如图1和图2所示实施例中方法等等。
本申请实施例还提供了一种计算机存储介质(Memory)，上述计算机存储介质是终端中的记忆设备，用于存放程序和数据。可以理解的是，此处的计算机存储介质既可以包括终端中的内置存储介质，当然也可以包括终端所支持的扩展存储介质。计算机存储介质提供存储空间，该存储空间存储了终端的操作系统。并且，在该存储空间中还存放了适于被处理器1001加载并执行的一条或多条的指令，这些指令可以是一个或一个以上的计算机程序(包括程序代码)。需要说明的是，此处的计算机存储介质可以是高速RAM存储器，也可以是非不稳定的存储器(non-volatile memory)，例如至少一个磁盘存储器；可选的还可以是至少一个位于远离前述处理器的计算机存储介质。
在一个实施例中，可由处理器1001加载并执行计算机存储介质中存放的一条或多条指令，以实现上述实施例中的相应步骤；具体实现中，计算机存储介质中的一条或多条指令可以由处理器1001加载并执行图1和/或图2中方法的任意步骤，此处不再赘述。
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的装置和模块的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
在本申请所提供的几个实施例中，应该理解到，所揭露的系统、装置和方法，可以通过其它的方式实现。例如，该模块的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如，多个模块或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。所显示或讨论的相互之间的耦合、或直接耦合、或通信连接可以是通过一些接口，装置或模块的间接耦合或通信连接，可以是电性，机械或其它的形式。
作为分离部件说明的模块可以是或者也可以不是物理上分开的，作为模块显示的部件可以是或者也可以不是物理模块，即可以位于一个地方，或者也可以分布到多个网络模块上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。
在上述实施例中，可以全部或部分地通过软件、硬件、固件或者其任意组合来实现。当使用软件实现时，可以全部或部分地以计算机程序产品的形式实现。该计算机程序产品包括一个或多个计算机指令。在计算机上加载和执行该计算机程序指令时，全部或部分地产生按照本申请实施例的流程或功能。该计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。该计算机指令可以存储在计算机可读存储介质中，或者通过该计算机可读存储介质进行传输。该计算机指令可以从一个网站站点、计算机、服务器或数据中心通过有线(例如同轴电缆、光纤、数字用户线(digital subscriber line，DSL))或无线(例如红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。该计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包含一个或多个可用介质集成的服务器、数据中心等数据存储设备。该可用介质可以是只读存储器(read-only memory，ROM)，或随机存储存储器(random access memory，RAM)，或磁性介质，例如，软盘、硬盘、磁带、磁碟、或光介质，例如，数字通用光盘(digital versatile disc，DVD)、或者半导体介质，例如，固态硬盘(solid state disk，SSD)等。

Claims (10)

  1. 一种心电图心搏分类方法，其特征在于，包括：
    获取心电图数据，对所述心电图数据进行划分，获得多个心搏片段；
    将所述多个心搏片段分别与预设的数据模板匹配，确定所述多个心搏片段中各个心搏片段对应的数据模板，以及所述数据模板的模板标识；
    在所述各个心搏片段对应的数据模板中确定一个主导模板，设置所述主导模板的类型标识为第一类型；
    对所述数据模板中除所述主导模板以外的待分类模板进行类型判断，确定所述待分类模板的类型标识；
    将所述各个心搏片段的类型标识设置为对应的数据模板的类型标识。
  2. 根据权利要求1所述的心电图心搏分类方法，其特征在于，所述对所述数据模板中除所述主导模板以外的待分类模板进行类型判断，确定所述待分类模板的类型标识包括：
    获取各个所述待分类模板的波形与所述主导模板的波形的相似度差值；
    若所述相似度差值小于预设差值阈值，确定所述待分类模板的类型标识为所述第一类型。
  3. 根据权利要求2所述的心电图心搏分类方法，其特征在于，所述对所述数据模板中除所述主导模板以外的待分类模板进行类型判断，确定所述待分类模板的类型标识，还包括：
    获取所述待分类模板的散点图，判断所述待分类模板的散点图中的坐标点是否分布在预设参考线附近；
    若是，确定所述待分类模板的类型标识为所述第一类型。
  4. 根据权利要求3所述的心电图心搏分类方法，其特征在于，所述获取所述待分类模板的散点图包括：
    获取所述待分类模板中的心搏片段的第一间期和第二间期，所述第一间期为所述心搏片段与相邻前一个心搏片段的间距；所述第二间期为所述心搏片段与相邻后一个心搏片段的间距；
    以所述第一间期为横坐标，以所述第二间期为纵坐标，生成所述待分类模板的散点图。
  5. 根据权利要求3或4所述的心电图心搏分类方法，其特征在于，所述判断所述待分类模板的散点图中的坐标点是否分布在预设参考线附近，包括：
    获取所述心搏片段对应的坐标点与预设参考线的有符号距离；
    获取所述有符号距离分布服从正态分布情况下的估计均值和估计标准差；
    根据所述估计均值和所述估计标准差对所述待分类模板的散点图进行数据处理，获得所述心搏片段对应的坐标点与预设参考线的目标距离值；
    获取所述目标距离值的标准差，判断所述标准差是否小于预设的标准差阈值；
    若小于，确定所述待分类模板的散点图中的坐标点分布在所述预设参考线附近；若不小于，确定所述待分类模板的散点图中的坐标点未分布在所述预设参考线附近。
  6. 根据权利要求5所述的心电图心搏分类方法，其特征在于，所述确定所述待分类模板的散点图中的坐标点未分布在所述预设参考线附近之后，所述方法还包括：
    获取所述待分类模板中的心搏片段的第一间期的标准差和第二间期的标准差；获得所述第一间期的标准差和所述第二间期的标准差的比值；
    判断所述比值是否大于预设的标准差比值；
    获取所述待分类模板的代表波形宽度，判断所述待分类模板的代表波形宽度是否大于预设的宽度阈值；
    若均大于，确定所述待分类模板的类型标识为第二类型；
    若所述比值不大于所述预设的标准差比值，或者，所述待分类模板的代表波形宽度不大于预设的宽度阈值，确定所述待分类模板的类型标识为第三类型。
  7. 根据权利要求1-4任一项所述的心电图心搏分类方法，其特征在于，所述获取心电图数据之前，所述方法还包括：
    获取采集的多导联心电图数据；根据所述多导联心电图数据中的主分析导联编号和次分析导联编号构建双导联心电图数据；
    对所述双导联心电图数据进行带通滤波，获得所述心电图数据；
    所述对所述心电图数据进行划分，获得多个心搏片段包括：
    根据检测获得的R波位置信息对所述心电图数据进行划分，获得多个以所述R波位置为中心的等长度的心搏片段。
  8. 一种心电图心搏分类装置，其特征在于，包括：
    划分模块，用于获取心电图数据，对所述心电图数据进行划分，获得多个心搏片段；
    匹配模块，将所述多个心搏片段分别与预设的数据模板匹配，确定所述多个心搏片段中各个心搏片段对应的数据模板，以及所述数据模板的模板标识；
    处理模块，用于在所述各个心搏片段对应的数据模板中确定一个主导模板，设置所述主导模板的类型标识为第一类型；
    所述处理模块还用于，对所述数据模板中除所述主导模板以外的待分类模板进行类型判断，确定所述待分类模板的类型标识；
    设置模块，用于将所述各个心搏片段的类型标识设置为对应的数据模板的类型标识。
  9. 一种电子设备，其特征在于，包括存储器和处理器，所述存储器存储有计算机程序，所述计算机程序被所述处理器执行时，使得所述处理器执行如权利要求1至7中任一项所述的心电图心搏分类方法的步骤。
  10. 一种计算机可读存储介质，其特征在于，存储有计算机程序，所述计算机程序被处理器执行时，使得所述处理器执行如权利要求1至7中任一项所述的心电图心搏分类方法的步骤。
PCT/CN2020/126352 2020-06-17 2020-11-04 心电图心搏分类方法、装置、电子设备和介质 WO2021253709A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010554239.9 2020-06-17
CN202010554239.9A CN113876335B (zh) 2020-06-17 2020-06-17 心电图心搏分类方法、装置、电子设备和介质

Publications (1)

Publication Number Publication Date
WO2021253709A1 true WO2021253709A1 (zh) 2021-12-23

Family

ID=79011777

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/126352 WO2021253709A1 (zh) 2020-06-17 2020-11-04 心电图心搏分类方法、装置、电子设备和介质

Country Status (2)

Country Link
CN (1) CN113876335B (zh)
WO (1) WO2021253709A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20220071546A1 (en) * 2020-09-04 2022-03-10 Atsens Co., Ltd. Electrocardiogram signal processing apparatus, method, and program for identifying supraventricular arrhythmia and ventricular arrhythmia
CN115486855A (zh) * 2022-09-15 2022-12-20 浙江好络维医疗技术有限公司 一种基于qrs波群不定次循环叠加的心电图心搏分类方法
CN116350199A (zh) * 2023-05-31 2023-06-30 合肥心之声健康科技有限公司 一种动态心电图心搏模板生成方法及系统
CN117958835A (zh) * 2024-03-28 2024-05-03 北华大学 用于患者术后康复的心电图分析方法

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114469041A (zh) * 2022-01-30 2022-05-13 北京理工大学 一种运动过程中心率变化数据特征分析方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102835954A (zh) * 2012-09-07 2012-12-26 深圳市邦健电子有限公司 一种心拍波形模板生成、室性早搏检测处理方法及装置
CN105411567A (zh) * 2015-12-15 2016-03-23 北京世纪今科医疗器械有限公司 一种心电图数据分析方法及系统
US20170181655A1 (en) * 2014-05-09 2017-06-29 Universiteit Gent Detection of pulmonary vein isolation
CN109171704A (zh) * 2018-09-19 2019-01-11 深圳市理邦精密仪器股份有限公司 心电波形分析的方法、装置、终端和计算机可读存储介质
CN110464333A (zh) * 2019-07-23 2019-11-19 深圳邦健生物医疗设备股份有限公司 一种心电图数据的存储方法及装置
CN110693483A (zh) * 2019-09-02 2020-01-17 乐普智芯(天津)医疗器械有限公司 一种动态心电图自动分析的方法

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH09173310A (ja) * 1995-11-29 1997-07-08 Hewlett Packard Co <Hp> Ｅｃｇ波形の心搏を分類する方法および装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102835954A (zh) * 2012-09-07 2012-12-26 深圳市邦健电子有限公司 一种心拍波形模板生成、室性早搏检测处理方法及装置
US20170181655A1 (en) * 2014-05-09 2017-06-29 Universiteit Gent Detection of pulmonary vein isolation
CN105411567A (zh) * 2015-12-15 2016-03-23 北京世纪今科医疗器械有限公司 一种心电图数据分析方法及系统
CN109171704A (zh) * 2018-09-19 2019-01-11 深圳市理邦精密仪器股份有限公司 心电波形分析的方法、装置、终端和计算机可读存储介质
CN110464333A (zh) * 2019-07-23 2019-11-19 深圳邦健生物医疗设备股份有限公司 一种心电图数据的存储方法及装置
CN110693483A (zh) * 2019-09-02 2020-01-17 乐普智芯(天津)医疗器械有限公司 一种动态心电图自动分析的方法

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20220071546A1 (en) * 2020-09-04 2022-03-10 Atsens Co., Ltd. Electrocardiogram signal processing apparatus, method, and program for identifying supraventricular arrhythmia and ventricular arrhythmia
CN115486855A (zh) * 2022-09-15 2022-12-20 浙江好络维医疗技术有限公司 一种基于qrs波群不定次循环叠加的心电图心搏分类方法
CN115486855B (zh) * 2022-09-15 2024-05-03 浙江好络维医疗技术有限公司 一种基于qrs波群不定次循环叠加的心电图心搏分类方法
CN116350199A (zh) * 2023-05-31 2023-06-30 合肥心之声健康科技有限公司 一种动态心电图心搏模板生成方法及系统
CN116350199B (zh) * 2023-05-31 2023-08-18 合肥心之声健康科技有限公司 一种动态心电图心搏模板生成方法及系统
CN117958835A (zh) * 2024-03-28 2024-05-03 北华大学 用于患者术后康复的心电图分析方法

Also Published As

Publication number Publication date
CN113876335B (zh) 2023-05-02
CN113876335A (zh) 2022-01-04

Similar Documents

Publication Publication Date Title
WO2021253709A1 (zh) 心电图心搏分类方法、装置、电子设备和介质
Wang et al. An improved convolutional neural network based approach for automated heartbeat classification
CN109117730B (zh) 心电图心房颤动实时判断方法、装置、系统及存储介质
Jin et al. A novel domain adaptive residual network for automatic atrial fibrillation detection
Alaskar et al. The implementation of pretrained AlexNet on PCG classification
CA2686877C (en) Ecg signal analysis tool
WO2019161610A1 (zh) 心电信息处理方法和心电工作站系统
CN111046731B (zh) 基于表面肌电信号进行手势识别的迁移学习方法和识别方法
WO2022110524A1 (zh) 心电图心搏数据聚类方法、装置、电子设备和介质
WO2019161611A1 (zh) 心电信息处理方法和心电工作站
CN110464333B (zh) 一种心电图数据的存储方法及装置
CN109864736A (zh) 心电信号的处理方法、装置、终端设备及介质
Ahmed et al. An investigative study on motifs extracted features on real time big-data signals
CN106923820B (zh) 一种心电信号伪差识别方法及心电信号伪差识别装置
US10271741B2 (en) Biometric scale
WO2022134472A1 (zh) 一种检测心电信号的方法、装置、终端及存储介质
CN109077720B (zh) 信号处理方法、装置、设备和存储介质
WO2021143215A1 (zh) 心电图模板匹配方法、装置、计算机设备及存储介质
CN110648318A (zh) 用于皮肤病的辅助分析方法及装置、电子设备、存储介质
Prakash et al. A system for automatic cardiac arrhythmia recognition using electrocardiogram signal
WO2022202943A1 (ja) 心電図解析支援装置、プログラム、心電図解析支援方法、及び心電図解析支援システム
CN114494798A (zh) 一种心电图伪差确认方法、终端设备及存储介质
WO2021031155A1 (zh) 一种基于ecg的多尺度特征提取方法及装置
CN111449646B (zh) 一种心电图识别方法、装置、存储介质及电子设备
CN114617562A (zh) 心拍信号的分类方法、装置、设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20940709

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20940709

Country of ref document: EP

Kind code of ref document: A1