Electrocardiosignal quality evaluation method and device, electronic equipment and storage medium

Abstract

The invention relates to an electrocardiosignal quality evaluation method and device, electronic equipment and a storage medium. The method comprises the steps that a first electrocardiosignal is obtained; determining a baseline drift index of the first electrocardiosignal and the number of target abnormal waves in the first electrocardiosignal; determining feature information of the first electrocardiosignal based on a variation parameter of the first electrocardiosignal, the baseline drift index and the number of the abnormal waves; performing at least one classification processing on the feature information to obtain at least one classification result; and based on each classification result, obtaining a target evaluation result corresponding to the first electrocardiosignal, so that the calculation complexity and resource loss are reduced while the accuracy of evaluating the quality of the first electrocardiosignal is improved.

Description

Technical Field

[0001] The present disclosure relates to the technical field of signal processing, and in particular, to a method and device for evaluating the quality of electrocardiogram signals, an electronic device, and a storage medium. Background Art

[0002] With the development of technologies such as the Internet of Things, big data, and artificial intelligence, as well as the popularization of mobile terminals, wearable electrocardiogram monitoring has received extensive attention. In practical applications, the quality of electrocardiogram signals is often greatly affected by different noises and artifacts. Therefore, evaluating the quality of electrocardiogram signals plays an important role in improving the accuracy and reliability of electrocardiogram analysis systems.

[0003] In related technologies, by adopting feature extraction such as baseline, QRS complex, etc., and combining machine learning or heuristic decision-making methods, although the quality of electrocardiogram signals can be evaluated, the features extracted by this method are relatively single, which is likely to cause inaccurate evaluation problems. At the same time, deep learning algorithms have a large amount of computation, which is not conducive to wide application on resource-constrained wearable devices. Summary of the Invention

[0004] To overcome the problems existing in related technologies, the present disclosure provides a method and device for evaluating the quality of electrocardiogram signals, an electronic device, and a storage medium, so as to overcome the problems of inaccurate evaluation of the quality of electrocardiogram signals in the prior art and being not conducive to wide application in wearable devices.

[0005] According to a first aspect of an embodiment of the present disclosure, there is provided a method for evaluating the quality of electrocardiogram signals, including:

[0006] Obtaining a first electrocardiogram signal;

[0007] Determining a baseline drift index of the first electrocardiogram signal and the number of target abnormal waves in the first electrocardiogram signal;

[0008] Based on a variation parameter of the first electrocardiogram signal, the baseline drift index, and the number of abnormal waves, determining feature information of the first electrocardiogram signal;

[0009] Performing at least one classification process on the feature information to obtain at least one classification result;

[0010] Based on each classification result, obtaining a target evaluation result corresponding to the first electrocardiogram signal.

[0011] In some embodiments, the performing at least one classification process on the feature information to obtain at least one classification result includes:

[0012] Inputting the feature information into at least one classification model to obtain classification results corresponding to the respective classification models.

[0013] In some embodiments, determining the number of target abnormal waves in the first electrocardiogram (ECG) signal includes:

[0014] Determining a first wave group from the first ECG signal and a second wave group formed based on the first wave group and a preset wave;

[0015] Determining a first quantity of the first wave groups in an abnormal state and a second quantity of the abnormal waves in the first ECG signal excluding the second wave group;

[0016] Determining the characteristic information of the first ECG signal based on the variation parameter of the first ECG signal, the baseline drift index, and the number of abnormal waves includes:

[0017] Determining the characteristic information of the first ECG signal based on the variation parameter of the first ECG signal, the baseline drift index, the first quantity, and the second quantity.

[0018] In some embodiments, the method further includes:

[0019] Performing slicing processing on the first ECG signal to obtain a plurality of signal segments with equal lengths;

[0020] Determining the quantity of the first wave groups in each of the signal segments as a third quantity;

[0021] Determining the variation parameter corresponding to each signal segment based on the third quantity corresponding to each signal segment and the first morphological information of the first wave group;

[0022] Determining the variation parameter corresponding to each signal segment as the variation parameter of the first ECG signal.

[0023] In some embodiments, the first morphological information includes: the amplitudes of the waves in the first wave group; the first wave group includes: a first type of wave, a second type of wave, and a third type of wave;

[0024] Determining the variation parameter corresponding to each signal segment based on the third quantity corresponding to each signal segment and the first morphological information of the first wave group includes:

[0025] Determining a first amplitude difference based on the amplitude of the first type of wave and the amplitude of the second type of wave;

[0026] Determining a second amplitude difference based on the amplitude of the second type of wave and the amplitude of the third type of wave;

[0027] Determining the first amplitude of the first wave group based on the first amplitude difference and the second amplitude difference;

[0028] Determine an amplitude average value based on each of the first amplitudes and the third quantity;

[0029] Calculate an amplitude anomaly parameter corresponding to the signal segment based on each of the first amplitudes and the amplitude average value.

[0030] In some embodiments, the first morphological information includes: the width between each wave in the first wave group;

[0031] The determining of the variation parameter corresponding to each signal segment based on the third quantity corresponding to each signal segment and the first morphological information of the first wave group includes:

[0032] Determine a first width between the first type of wave and the third type of wave;

[0033] Determine a first width average value based on each of the first widths and the third quantity;

[0034] Calculate a first width anomaly parameter corresponding to the signal segment based on each of the first widths and the first width average value; and

[0035] Determine a second width between the second type of wave in two adjacent first wave groups;

[0036] Determine a second width average value based on each of the second widths and the third quantity;

[0037] Calculate a second width anomaly parameter corresponding to the signal segment based on each of the second widths and the second width average value.

[0038] In some embodiments, the method further includes:

[0039] Determine a first wave group with the first amplitude not within a first preset amplitude range as a first wave group in an abnormal state;

[0040] Determine a first wave group with the first width not within a preset width range as the first wave group in the abnormal state;

[0041] Wherein, the first preset amplitude range is related to the amplitude average value, and the preset width range is related to the first width average value.

[0042] In some embodiments, the method further includes:

[0043] Determine second morphological information of each candidate wave except the second wave group in each signal segment; wherein, the second morphological information includes the second amplitude of each candidate wave;

[0044] Determine the candidate wave whose second amplitude is not within the second preset amplitude range as the abnormal wave;

[0045] Wherein, the second preset amplitude range is related to the amplitude average value.

[0046] In some embodiments, determining the baseline drift index of the first electrocardiogram signal includes:

[0047] Determine the peak points of each wave in the first electrocardiogram signal;

[0048] Calculate the rising edge amplitude and the falling edge amplitude corresponding to each peak point;

[0049] Determine the baseline drift index based on the rising edge amplitude, the falling edge amplitude, and the amplitude average value.

[0050] In some embodiments, the method further includes:

[0051] Determine the third amplitude difference corresponding to each peak point based on the rising edge amplitude and the falling edge amplitude corresponding to each peak point;

[0052] The determining the baseline drift index based on the rising edge amplitude, the falling edge amplitude, and the amplitude average value includes:

[0053] When it is detected that a plurality of consecutive or adjacent third amplitude differences are greater than a preset threshold, select the largest third amplitude difference as the baseline drift threshold;

[0054] Calculate the baseline drift index based on the baseline drift threshold and the amplitude average value.

[0055] In some embodiments, the method further includes:

[0056] Perform filtering processing on the second electrocardiogram signal to obtain a third electrocardiogram signal, a low-frequency noise signal, and a high-frequency noise signal;

[0057] Perform normalization processing on the data in the third electrocardiogram signal to obtain the first electrocardiogram signal.

[0058] In some embodiments, the method further includes:

[0059] Determine the third amplitude of the high-frequency noise signal based on the peak and valley values of each wave in the high-frequency noise signal;

[0060] Determine the fourth amplitude of the low-frequency noise signal based on the peak and valley values of each wave in the low-frequency noise signal;

[0061] Determining the characteristic information of the first electrocardiogram signal based on the variation parameter of the first electrocardiogram signal, the baseline drift index, and the number of abnormal waves, includes:

[0062] Determine the variation parameter, the baseline drift index, the number of abnormal waves, the third amplitude, and the fourth amplitude as the characteristic information of the electrocardiogram signal.

[0063] According to a second aspect of the embodiments of the present disclosure, there is provided an electrocardiogram signal quality evaluation device, including:

[0064] A first acquisition module, configured to acquire a first electrocardiogram signal;

[0065] A first determination module, configured to determine the baseline drift index of the first electrocardiogram signal and the number of target abnormal waves in the first electrocardiogram signal;

[0066] A second determination module, configured to determine the characteristic information of the first electrocardiogram signal based on the variation parameter of the first electrocardiogram signal, the baseline drift index, and the number of abnormal waves;

[0067] A second acquisition module, configured to perform at least one classification process on the characteristic information to obtain at least one classification result;

[0068] A third acquisition module, configured to obtain a target evaluation result corresponding to the first electrocardiogram signal based on each classification result.

[0069] In some embodiments, the second acquisition module is specifically configured to:

[0070] Input the characteristic information into at least one classification model to obtain classification results corresponding to each classification model.

[0071] In some embodiments, the first determination module is specifically configured to:

[0072] Determine a first wave group from the first electrocardiogram signal, and a second wave group formed based on the first wave group and a preset wave form;

[0073] Determine a first number of the first wave groups in an abnormal state, and a second number of abnormal waves in the first electrocardiogram signal excluding the second wave group;

[0074] The second determination module is specifically configured to:

[0075] Determine the characteristic information of the first electrocardiogram signal based on the variation parameter of the first electrocardiogram signal, the baseline drift index, the first number, and the second number.

[0076] In some embodiments, the apparatus further comprises:

[0077] A fourth acquisition module, configured to perform slicing processing on the first electrocardiogram signal to obtain a plurality of signal segments with equal lengths;

[0078] A third determination module, configured to determine the number of the first wave groups in each of the signal segments as a third quantity;

[0079] A fourth determination module, configured to determine a variation parameter corresponding to each of the signal segments based on the third quantity corresponding to each of the signal segments and the first morphological information of the first wave group;

[0080] A fifth determination module, configured to determine the variation parameter corresponding to each of the signal segments as the variation parameter of the first electrocardiogram signal.

[0081] In some embodiments, the first morphological information includes: the amplitudes of the waves in the first wave group; the first wave group includes: a first type of wave, a second type of wave, and a third type of wave;

[0082] The fourth determination module is specifically configured to:

[0083] Determine a first amplitude difference based on the amplitude of the first type of wave and the amplitude of the second type of wave;

[0084] Determine a second amplitude difference based on the amplitude of the second type of wave and the amplitude of the third type of wave;

[0085] Determine a first amplitude of the first wave group based on the first amplitude difference and the second amplitude difference;

[0086] Determine an amplitude average value based on each of the first amplitudes and the third quantity;

[0087] Calculate an amplitude anomaly parameter corresponding to the signal segment based on each of the first amplitudes and the amplitude average value.

[0088] In some embodiments, the first morphological information includes: the widths between the waves in the first wave group;

[0089] The fourth determination module is further configured to:

[0090] Determine a first width between the first type of wave and the third type of wave;

[0091] Determine a first width average value based on each of the first widths and the third quantity;

[0092] Calculate a first width anomaly parameter corresponding to the signal segment based on each of the first widths and the first width average value; and

[0093] Determine a second width between second type waves in two adjacent first wave groups;

[0094] Determine a second width average value based on each of the second widths and the third quantity;

[0095] Calculate a second width anomaly parameter corresponding to the signal segment based on each of the second widths and the second width average value.

[0096] In some embodiments, the apparatus further includes:

[0097] A sixth determination module configured to determine a first wave group with a first amplitude not within a first preset amplitude range as a first wave group in an abnormal state;

[0098] A seventh determination module configured to determine a first wave group with a first width not within a preset width range as the first wave group in the abnormal state;

[0099] Wherein, the first preset amplitude range is related to the amplitude average value, and the preset width range is related to the first width average value.

[0100] In some embodiments, the apparatus further includes:

[0101] An eighth determination module configured to determine second morphological information of each candidate wave in each signal segment except the second wave group; wherein, the second morphological information includes a second amplitude of each candidate wave;

[0102] A ninth determination module configured to determine a candidate wave with a second amplitude not within a second preset amplitude range as the abnormal wave;

[0103] Wherein, the second preset amplitude range is related to the amplitude average value.

[0104] In some embodiments, the first determination module includes:

[0105] A first sub-module configured to determine peak points of each wave in the first electrocardiogram signal;

[0106] A second sub-module configured to calculate a rising edge amplitude and a falling edge amplitude corresponding to each peak point;

[0107] A third sub-module configured to determine the baseline drift index based on the rising edge amplitude, the falling edge amplitude, and the amplitude average value.

[0108] In some embodiments, the apparatus further includes:

[0109] A tenth determination module, configured to determine a third amplitude difference corresponding to each of the peak points based on the rising edge amplitude and the falling edge amplitude corresponding to each of the peak points;

[0110] The third sub-module is specifically configured to:

[0111] When it is detected that a plurality of consecutive or adjacent third amplitude differences are all greater than a preset threshold, select the largest third amplitude difference as the baseline drift threshold;

[0112] Calculate the baseline drift index based on the baseline drift threshold and the amplitude average value.

[0113] In some embodiments, the apparatus further includes:

[0114] A first processing module, configured to perform filtering processing on the second electrocardiogram signal to obtain a third electrocardiogram signal, a low-frequency noise signal, and a high-frequency noise signal;

[0115] A second processing module, configured to perform normalization processing on the data in the third electrocardiogram signal to obtain the first electrocardiogram signal.

[0116] In some embodiments, the apparatus further includes:

[0117] A third processing module, configured to determine a third amplitude of the high-frequency noise signal based on the peak and valley values of each wave in the high-frequency noise signal;

[0118] A fourth processing module, configured to determine a fourth amplitude of the low-frequency noise signal based on the peak and valley values of each wave in the low-frequency noise signal;

[0119] The second determination module is specifically configured to

[0120] Determine the mutation parameter, the baseline drift index, the number of abnormal waves, the third amplitude, and the fourth amplitude as the characteristic information of the electrocardiogram signal.

[0121] According to a third aspect of the embodiments of the present disclosure, there is provided an electronic device, including:

[0122] A processor;

[0123] A memory for storing processor-executable instructions;

[0124] Wherein, the processor is configured to: when executing the executable command, implement the steps in any one of the electrocardiogram signal quality assessment methods in the first aspect above.

[0125] According to a fourth aspect of the embodiments of the present disclosure, there is provided a non-transitory computer-readable storage medium, including:

[0126] When the instructions in the storage medium are executed by a processor of an electronic device, the steps in any one of the electrocardiogram signal quality assessment methods in the above first aspect are implemented.

[0127] The technical solutions provided by the embodiments of the present disclosure may include the following beneficial effects:

[0128] In the embodiments of the present disclosure, first, a first electrocardiogram signal is obtained, and then the baseline drift index of the first electrocardiogram signal and the number of target abnormal waves in the first electrocardiogram signal are determined; then, based on the variation parameter, baseline drift index, and number of abnormal waves of the first electrocardiogram signal, the characteristic information of the first electrocardiogram signal is determined, at least one classification process is performed on the characteristic information to obtain at least one classification result, and based on each classification result, a target evaluation result corresponding to the first electrocardiogram signal is obtained.

[0129] On the one hand, determining the characteristic information of the first electrocardiogram signal based on the variation parameter, baseline drift index, and number of abnormal waves of the first electrocardiogram signal can improve the diversity of the characteristic information, and based on the diverse characteristic information, evaluate the quality of the first electrocardiogram signal, thereby facilitating the improvement of the accuracy of evaluating the first electrocardiogram signal; on the other hand, performing at least one classification process on the characteristic information to obtain at least one classification result, and based on each classification result, determining the target evaluation result corresponding to the first electrocardiogram signal can reduce the computational complexity, reduce resource consumption, and facilitate the wide application of this algorithm to wearable devices.

[0130] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0131] The accompanying drawings herein are incorporated into the specification and constitute a part of the specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure.

[0132] Figure 1 is a flowchart showing a method for evaluating the quality of an electrocardiogram signal according to an exemplary embodiment Figure 1 ;

[0133] Figure 2 is a flowchart showing a method for evaluating the quality of an electrocardiogram signal according to an exemplary embodiment Figure 2 ;

[0134] Figure 3 is a schematic diagram showing the quality evaluation result of an electrocardiogram signal according to an exemplary embodiment;

[0135] Figure 4 It is a block diagram of a device for evaluating the quality of electrocardiogram (ECG) signals shown according to an exemplary embodiment;

[0136] Figure 5 It is a hardware structure block diagram of an electronic device shown according to an exemplary embodiment Figure 1 ;

[0137] Figure 6 It is a hardware structure block diagram of an electronic device shown according to an exemplary embodiment Figure 2 。 Detailed implementation manners

[0138] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the present disclosure. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0139] In one implementation manner, by acquiring a first ECG signal, calculating the baseline L of the first ECG signal, calculating the slope LS of the baseline L, comparing the slope LS with a preset slope threshold, if the slope LS exceeds the preset slope threshold, then the first ECG signal data segment is marked as an invalid segment LD of baseline drift. Then, when evaluating the quality of the ECG signal through artificial intelligence or algorithm technology, the invalid signal segment can be ignored. However, this method only judges the quality of the first ECG signal based on baseline drift, and the characteristic information is relatively single, which is not conducive to improving the accuracy of judging the quality of the first ECG signal.

[0140] In another implementation manner, a first ECG signal of twelve leads is collected, the first ECG signal is quality - marked, the waveform attributes of the signal and the characteristics of baseline drift are extracted, the extracted characteristics are used as input1 of the input model of a convolutional neural network (CNN), then the time - frequency two - dimensional image of the first ECG signal is calculated, the time - frequency spectrum matrix is used as input2 of the multi - input model of the CNN, and convolutional pooling operations are performed on the time - frequency spectrum matrix to extract the characteristic information of the time - frequency spectrum matrix; then the characteristic information input1 extracted from the time - frequency spectrum matrix is fused, a CNN classifier is used, and the fused characteristics are trained to obtain a classification model. Finally, the ECG signal to be classified is processed into a time - frequency two - dimensional image, and features are manually extracted, and the time - frequency two - dimensional image and the manually extracted features are input into the classification model to output the signal quality grade result. However, this method uses a CNN model to evaluate the quality of the ECG signal, with high computational complexity and large resource consumption, which is not conducive to wide application on resource - limited wearable devices.

[0141] Based on this, embodiments of the present disclosure propose to obtain a first electrocardiogram (ECG) signal, then determine characteristic information such as a baseline drift index and variation parameters from the first ECG signal, input the above characteristic information into a classification model, and obtain a target evaluation result corresponding to the first ECG signal, thereby improving the accuracy of evaluating the quality of the first ECG signal while reducing the computational complexity.

[0142] Figure 1 is a schematic flowchart of a method for evaluating the quality of an electrocardiogram (ECG) signal shown according to an exemplary embodiment Figure 1 , such as Figure 1 shown, the method for evaluating the quality of an electrocardiogram (ECG) signal mainly includes the following steps:

[0143] In step 101, obtain a first electrocardiogram (ECG) signal;

[0144] In step 102, determine the baseline drift index of the first electrocardiogram (ECG) signal and the number of target abnormal waves in the first electrocardiogram (ECG) signal;

[0145] In step 103, based on the variation parameters of the first electrocardiogram (ECG) signal, the baseline drift index, and the number of abnormal waves, determine the characteristic information of the first electrocardiogram (ECG) signal;

[0146] In step 104, perform at least one classification process on the characteristic information to obtain at least one classification result;

[0147] In step 105, based on each classification result, obtain a target evaluation result corresponding to the first electrocardiogram (ECG) signal.

[0148] It should be noted that the method for evaluating the quality of an electrocardiogram (ECG) signal proposed in the present disclosure can be applied to an electronic device. Here, the electronic device may include: a terminal device, for example, a mobile terminal or a fixed terminal. Among them, the mobile terminal may include: devices such as a mobile phone, a tablet computer, a laptop computer, a wearable electronic device, etc. The fixed terminal may include: a desktop computer, a smart TV, a vehicle-mounted device, etc. In other embodiments, the method for evaluating the quality of an electrocardiogram (ECG) signal can also be applied to an application program installed on an electronic device.

[0149] In other embodiments, the method for evaluating the quality of an electrocardiogram (ECG) signal in the embodiments of the present disclosure can be configured in an electrocardiogram (ECG) signal quality evaluation device, and the electrocardiogram (ECG) signal quality evaluation device can be set in an electronic device. The embodiments of the present disclosure do not limit this. It should be noted that the execution subject of the embodiments of the present disclosure can be a central processing unit (CPU) in an electronic device in terms of hardware, and can be a relevant background service in an electronic device in terms of software. The embodiments of the present disclosure do not limit this.

[0150] It can be understood that the first electrocardiogram (ECG) signal is a weak and low-frequency physiological electrical signal, usually with a frequency range of 0.05 - 100 Hz and an amplitude not exceeding 4 mV. It is a component of the vital sign signal parameters and one of the important physiological electrical signals. The ECG signal contains a large amount of information for clinical diagnosis of cardiovascular diseases. Therefore, evaluating the quality of the ECG signal is an important means to understand the heart function and condition, assist in diagnosing cardiovascular diseases, and evaluate the effectiveness of various treatment methods.

[0151] Here, the first ECG signal can be collected at rest when the organism is lying, standing, or sitting, that is, it can be determined according to the actual situation, and the embodiments of the present disclosure do not limit this.

[0152] It should be noted that the baseline drift index of the first ECG signal is one of the indicators for evaluating the change of the first ECG signal; the number of target abnormal waves in the first ECG signal is one of the indicators affecting the quality of the first ECG signal. Therefore, in order to determine the quality grade of the first ECG signal, it is necessary to first determine the baseline drift index and the number of abnormal waves.

[0153] Here, the abnormal wave can be a wave with abnormal waveform or a wave with abnormal data, such as frequency, phase, and pulse width. The embodiments of the present disclosure do not limit this.

[0154] In some embodiments, the frequency and pulse width of each wave in the first ECG signal can be obtained, and then it is judged whether each frequency is within a first preset range and whether each pulse width is within a second preset range; the wave corresponding to the frequency not within the first preset range is determined as an abnormal wave, and the wave corresponding to the pulse width not within the second preset range is determined as an abnormal wave, so as to count the number of abnormal waves.

[0155] In some other embodiments, a baseline drift separator can be used to remove the drift of the input first ECG signal (X1) to obtain the baseline drift component (X2) and the ECG signal with baseline drift removed (X3); an FIR low-pass filter is used to perform low-pass filtering on X3 to filter out the high-frequency components and obtain the ECG signal with baseline drift removed and high-frequency components filtered out (X4); a baseline identifier is used to identify the baseline positioning information (X5) for X4; X5 and X3 are input into a baseline signal similar mean square error parameter calculation module to calculate and output the baseline signal similar mean square error parameter (Y1); and X5 and X2 are input into a baseline drift amplitude parameter calculation module to calculate and output the baseline drift fluctuation amplitude parameter (Y2); finally, Y1 and Y2 are input into a Bayesian network, and the baseline drift index of the first ECG signal is output by this Bayesian network.

[0156] In addition to obtaining the baseline drift index of the first electrocardiogram (ECG) signal and the number of target abnormal waves in the first ECG signal, the variation parameters of the first ECG signal can also be obtained, so as to increase the characteristic information for evaluating the first ECG signal, thereby improving the accuracy of evaluating the quality of the first ECG signal.

[0157] Here, the variation parameters of the first ECG signal can be a spatial variation degree characteristic curve, a time variation degree characteristic curve, an amplitude variation parameter, etc., and the embodiments of the present disclosure do not limit this.

[0158] In some embodiments, the first ECG signal can be filtered to extract the R wave and T wave in the first ECG cycle, and the peak points of the R wave and T wave in several cardiac cycles are obtained; the peak points of the R wave and T wave are traversed, and for any adjacent R wave peak points and adjacent T wave peak points, according to the amplitude difference and time span difference between the peak points, the spatial variation degree characteristic curve and the time variation degree characteristic curve of the first ECG signal are extracted. Then, spatial reconstruction is performed on the spatial variation degree characteristic curve and the time variation degree characteristic curve to obtain spatial running curves respectively, and then the distance metric between all data points in the spatial running curve and the data start point is calculated, where the distance metric includes one or more of Euclidean distance, cosine similarity, and Minkowski distance.

[0159] It can be understood that after determining the variation parameters, baseline drift index, and the number of abnormal waves of the first ECG, the above-obtained information can be determined as the characteristic information of the first ECG signal.

[0160] It should be noted that in order to improve the accuracy of evaluating the quality of the first ECG signal while reducing the calculation amount, the characteristic information of the first ECG signal can be classified at least once to obtain at least one classification result.

[0161] In some embodiments, the classification result can be that the first ECG signal is of excellent signal level or of poor signal level.

[0162] Here, the classification process can be PCA (Principal Component Analysis) processing, LDA (Linear Discriminant Analysis) processing, ICA (Independent Component Analysis) processing, or feature selection processing, etc., and the embodiments of the present disclosure do not limit this.

[0163] In some embodiments, through PCA (Principal Component Analysis) processing, high-dimensional characteristic information can be mapped into a low-dimensional space, thereby reducing the dimension of the characteristic information and retaining the principal components that contribute significantly to the classification task, so as to reduce the computational complexity in the quality assessment process of the first electrocardiogram (ECG) signal; through LDA (Linear Discriminant Analysis), the vector projection of the characteristic information can be found, making the data points within the same class closer and the data points between different classes more dispersed, so as to improve the accuracy of the quality classification result of the first ECG signal; through feature selection, the contribution degrees of each feature in the characteristic information (such as the variation parameter of the first ECG, the baseline drift index, and the number of abnormal waves) can be evaluated, and the features with larger contribution degrees can be selected, so as to reduce the complexity of the quality assessment process of the first ECG signal.

[0164] Exemplarily, after performing PCA on the characteristic information of the first ECG signal and inputting the analyzed characteristic information into the learning model, the first classification result corresponding to the first ECG signal can be obtained; after performing LDA on the characteristic information of the first ECG signal and inputting the analyzed characteristic information into the learning model, the second classification result corresponding to the first ECG signal can be obtained.

[0165] To further improve the accuracy of evaluating the quality of the first ECG signal, each classification result can be integrated to obtain the target evaluation result corresponding to the first ECG signal.

[0166] In some embodiments, to facilitate determining the target evaluation result corresponding to the first ECG signal, each classification result can be scored. For example, a signal level of excellent is recorded as 1 point, a signal level of poor is recorded as 0 point, and then each scoring result is integrated to obtain the target evaluation result of the first ECG signal. For example, an evaluation of 0 - 1 point is considered excellent, 2 points is considered good, and 3 points is considered excellent.

[0167] In the embodiments of the present disclosure, the first ECG signal is first obtained, and then the baseline drift index of the first ECG signal and the number of target abnormal waves in the first ECG signal are determined; then, based on the variation parameter, baseline drift index, and the number of abnormal waves of the first ECG signal, the characteristic information of the first ECG signal is determined, at least one classification process is performed on the characteristic information to obtain at least one classification result, and based on each classification result, the target evaluation result corresponding to the first ECG signal is obtained.

[0168] On the one hand, determining the characteristic information of the first electrocardiogram signal based on the variation parameters, baseline drift index, and the number of abnormal waves of the first electrocardiogram signal can improve the diversity of the characteristic information, and based on the diverse characteristic information, evaluate the quality of the first electrocardiogram signal, thereby facilitating the improvement of the accuracy of evaluating the first electrocardiogram signal; on the other hand, performing at least one classification process on the characteristic information to obtain at least one classification result, and determining the target evaluation result corresponding to the first electrocardiogram signal based on each classification result can reduce the computational complexity, reduce resource consumption, and facilitate the wide application of this algorithm to wearable devices.

[0169] In some embodiments, the performing at least one classification process on the characteristic information to obtain at least one classification result includes:

[0170] Inputting the characteristic information into at least one classification model to obtain the classification results corresponding to the respective classification models.

[0171] It should be noted that in order to improve the accuracy of evaluating the quality of the first electrocardiogram signal while reducing the amount of calculation, the characteristic information of the first electrocardiogram signal can be input into at least one classification model, and the characteristic information is analyzed based on each classification model to obtain the classification results corresponding to the respective classification models.

[0172] Here, the classification model is a pre-trained model, and the classification model can be a logistic regression model, a decision tree model, a support vector machine model, a random forest model, etc., and the embodiments of the present disclosure do not limit this.

[0173] In some embodiments, the logistic regression model is a generalized linear model used for binary classification and multi-classification problems. The logistic regression model multiplies the input characteristic information by weights, and then maps the result to between 0 and 1 through the sigmoid function to obtain the classification result; the decision tree model is a tree-structured classification model that divides the input characteristic information into different categories through a series of judgment conditions to obtain the classification result; the support vector machine model is a classification model based on margin maximization that maps the input characteristic information to a high-dimensional space to obtain the classification result.

[0174] It can be understood that the classification model is a binary classification model, and the classification results include two results: the signal level is excellent and the signal level is poor. Therefore, after inputting the characteristic information into at least one classification model, the results with the signal level being excellent or the signal level being poor output by each classification model can be obtained.

[0175] Here, the characteristic information can be input into at least one classification model of the same type, or the characteristic information can be input into at least one classification model of different types, and the present disclosure does not limit this.

[0176] Exemplarily, the training process of the classification model is as follows: Manually label the electrocardiogram signals with known quality grades, that is, label the electrocardiogram signals with excellent signal grades and the electrocardiogram signals with poor signal grades, and input the labeled electrocardiogram signals into at least one classification model respectively to train each classification model, so as to obtain a pre-trained classification model.

[0177] In another example, the classification models are a logistic regression model, a decision tree model, and a support vector machine model; after inputting the feature information of the first electrocardiogram signal into the logistic regression model, the decision tree model, and the support vector machine model respectively, the classification result of the logistic regression model is that the signal grade is excellent, that is, recorded as 1 point; the classification result of the decision tree model is that the signal grade is poor, that is, recorded as 0 point; the classification result of the support vector machine model is that the signal grade is excellent, that is, recorded as 1 point. Then integrate the classification results of each model, that is, get 2 points, so as to determine that the target evaluation result corresponding to the first electrocardiogram signal is good.

[0178] In the embodiments of the present disclosure, by inputting the feature information into at least one classification model, the classification results corresponding to each classification model can be obtained, so that while reducing the computational complexity in the process of evaluating the quality of the first electrocardiogram signal, the accuracy of evaluating the quality of the first electrocardiogram signal can be improved.

[0179] In some embodiments, determining the number of target abnormal waves in the first electrocardiogram signal includes:

[0180] Determine a first wave group from the first electrocardiogram signal, and a second wave group formed based on the first wave group and a preset wave;

[0181] Determine a first quantity of the first wave group in an abnormal state, and a second quantity of abnormal waves in the first electrocardiogram signal except for the second wave group;

[0182] Determining the feature information of the first electrocardiogram signal based on the variation parameter, the baseline drift index, and the number of abnormal waves of the first electrocardiogram signal includes:

[0183] Determine the feature information of the first electrocardiogram signal based on the variation parameter, the baseline drift index, the first quantity, and the second quantity of the first electrocardiogram signal.

[0184] It should be noted that, in order to improve the accuracy of determining abnormal waves in the first electrocardiogram signal, waves or wave groups that have a greater impact on the quality of the first electrocardiogram signal can be located first, and then the number of abnormal waves can be counted from the waves or wave groups. Therefore, in the embodiments of the present disclosure, a first wave group is first determined from the first electrocardiogram signal, and a second wave group formed based on the first wave group and a preset wave; then, a first number of the first wave group in an abnormal state is determined, and a second number of abnormal waves in the first electrocardiogram signal except for the second wave group is determined.

[0185] Here, the first wave group and the preset wave can be set arbitrarily according to requirements, and the embodiments of the present disclosure do not limit this.

[0186] In some embodiments, the preset wave can be a P wave and a T wave, the first wave group is a QRS wave group, and the second wave group is a PQRST wave group; after obtaining the first electrocardiogram signal, the range of the QRS wave group in the first electrocardiogram signal can be determined by using empirical mode decomposition (EMD); then, within the range of the QRS wave group, the position of the R peak is determined, and based on the range of the QRS wave group and the determined position of the R peak, the positions of the Q peak and the S peak are determined; finally, in the QRS wave group and the adjacent QRS wave groups, relevant information of the T wave and the P wave is extracted, where the relevant information includes the peak positions of the T wave and the P wave and the starting points and ending points of the T wave and the P wave, so as to determine the first wave group and the second wave group from the first electrocardiogram signal.

[0187] Here, the number of target abnormal waves includes a first number of the first wave group in an abnormal state and a second number of abnormal waves in the first electrocardiogram signal except for the second wave group.

[0188] In some embodiments, after determining the QRS wave group and the PQRST wave group from the first electrocardiogram signal, the state of the QRS wave group can be judged, such as the waveform state, so as to determine a first number of the QRS wave group in an abnormal state; then, for other waves in the first electrocardiogram signal except for the PQRST wave group, such as the U wave or the T wave, the state is also judged in the same way, so as to determine a second number of abnormal waves; finally, based on the first number and the second number, the number of target abnormal waves is determined.

[0189] It can be understood that after determining the first number and the second number, the characteristic information of the first electrocardiogram signal can be determined based on the variation parameter, baseline drift index, the first number, and the second number of the first electrocardiogram signal.

[0190] In the embodiments of the present disclosure, first, a first wave group is determined from a first electrocardiogram signal, and a second wave group is formed based on the first wave group and a preset wave; then, a first quantity of the first wave group in an abnormal state is determined, and a second quantity of abnormal waves other than the second wave group in the first electrocardiogram signal is determined; finally, based on the variation parameter of the first electrocardiogram signal, the baseline drift index, the first quantity, and the second quantity, the characteristic information of the first electrocardiogram signal is determined. In this way, by obtaining the first quantity and the second quantity, it is beneficial to improve the accuracy of determining the quantity of target abnormal waves in the first electrocardiogram signal, thereby improving the accuracy of evaluating the first electrocardiogram signal, and being able to take into account both the first wave group in an abnormal state and the abnormal waves that do not belong to the second wave group, which can improve the comprehensiveness and accuracy of the obtained characteristic information.

[0191] In some embodiments, the method further includes:

[0192] Performing slicing processing on the first electrocardiogram signal to obtain a plurality of signal segments with equal lengths;

[0193] Determining the quantity of the first wave group in each of the signal segments as a third quantity;

[0194] Based on the third quantity corresponding to each of the signal segments and the first morphological information of the first wave group, determining the variation parameter corresponding to each of the signal segments;

[0195] Determining the variation parameter corresponding to each of the signal segments as the variation parameter of the first electrocardiogram signal.

[0196] It can be understood that, in order to improve the evaluation efficiency of the first electrocardiogram signal, slicing processing can be first performed on the first electrocardiogram signal to obtain a plurality of signal segments with equal lengths, so as to extract the characteristic information in each signal segment and evaluate the quality of the first electrocardiogram signal.

[0197] Here, the length of the signal segment can be set arbitrarily, and the embodiments of the present disclosure do not limit this.

[0198] Exemplarily, if the first electrocardiogram signal is a 30 - second signal, the first electrocardiogram signal can be sliced according to a fixed length of 3 seconds, so as to obtain 10 signal segments of 3 seconds.

[0199] It should be noted that after obtaining each signal segment, it is necessary to locate the first wave group in each signal segment and count the quantity of the first wave group in each signal segment, denoted as the third quantity, so as to calculate the variation parameter of each signal segment based on each third quantity.

[0200] To ensure the accuracy of calculating the variation parameters for each signal segment, in addition to determining the third quantity in each signal segment, the first morphological information of the first wave group in each signal segment should also be determined, such as the waveform of the first wave group.

[0201] It can be understood that after obtaining the variation parameters corresponding to each signal segment based on the third quantity and the first morphological information of the first wave group in each signal segment, the variation parameters corresponding to each signal segment can be determined as the variation parameters of the first electrocardiogram signal, that is, the variation parameters corresponding to each signal segment are used as one of the characteristic information of the first electrocardiogram signal and input into at least one classification model, so as to obtain the target evaluation result corresponding to the first electrocardiogram signal.

[0202] In the embodiments of the present disclosure, the first electrocardiogram signal is sliced to obtain multiple signal segments of equal length; then the number of the first wave groups in each signal segment is determined; finally, based on the number of the first wave groups and the first morphological information of the first wave group in each signal segment, the variation parameters corresponding to each signal segment are determined. Thus, by segmenting the first electrocardiogram signal, while improving the processing efficiency, it is beneficial to improve the accuracy of determining the variation parameters of the first electrocardiogram signal.

[0203] In some embodiments, the first morphological information includes: the amplitudes of the waves in the first wave group; the first wave group includes: the first type of wave, the second type of wave, and the third type of wave;

[0204] The determining the variation parameters corresponding to each signal segment based on the third quantity and the first morphological information of the first wave group corresponding to each signal segment includes:

[0205] Based on the amplitude of the first type of wave and the amplitude of the second type of wave, determine the first amplitude difference;

[0206] Based on the amplitude of the second type of wave and the amplitude of the third type of wave, determine the second amplitude difference;

[0207] Based on the first amplitude difference and the second amplitude difference, determine the first amplitude of the first wave group;

[0208] Based on each of the first amplitudes and the third quantity, determine the amplitude average value;

[0209] Based on each of the first amplitudes and the amplitude average value, calculate the amplitude anomaly parameter corresponding to the signal segment.

[0210] It can be understood that, in order to improve the accuracy of determining the mutation parameters corresponding to each of the signal segments, the first morphological information includes: the amplitudes of the waves in the first wave group, so that the amplitude anomaly parameter corresponding to the signal segment can be calculated based on the amplitudes of the waves and the third quantity.

[0211] Here, the first wave group includes: the first type of wave, the second type of wave, and the third type of wave.

[0212] In some embodiments, the first wave group is a QRS wave group, the first type of wave is a Q wave, the second type of wave is an R wave, and the third type of wave is an S wave.

[0213] Here, the amplitude represents the amplitude size of the wave, that is, the difference between the peak value and the valley value of the wave.

[0214] In the embodiments of the present disclosure, based on the first morphological information, the amplitudes of the first type of wave, the amplitudes of the second type of wave, and the amplitudes of the third type of wave are obtained; then, based on the amplitudes of the first type of wave, the amplitudes of the second type of wave, and the amplitudes of the third type of wave, the first amplitude of the first wave group is calculated; then, based on the first amplitudes of each of the first wave groups and the third quantity, the amplitude average value is determined; finally, based on the amplitude average value and each of the first amplitudes, the amplitude anomaly parameter is calculated.

[0215] Exemplarily, the first type of wave is a Q wave, the second type of wave is an R wave, and the third type of wave is an S wave; based on the amplitude of the Q wave and the amplitude of the R wave, the first amplitude difference is determined; based on the amplitude of the R wave and the amplitude of the S wave, the second amplitude difference is determined; based on the first amplitude difference and the second amplitude difference, the first amplitude qrs_amp of the first wave group is determined i ; the first amplitude in the signal segment can be recorded as QRS_AMP = {qrs_amp1, qrs_amp2,... qrs_amp n}; then, based on the third quantity n and each of the first amplitudes qrs_amp i , the amplitude average value E1 is determined; finally, based on the amplitude average value E1 and the third quantity n, the amplitude anomaly parameter is calculated.

[0216] Therefore, the calculation formula for the amplitude anomaly parameter corresponding to the signal segment can be as follows:

[0217]

[0218] In formula (1), CV amp represents the amplitude anomaly parameter, n represents the third quantity corresponding to the first wave group in the signal segment, qrs_amp i represents each of the first amplitudes in the signal segment, E1 represents the amplitude average value, represents the accumulation of n first amplitudes.

[0219] In the embodiments of the present disclosure, the first morphological information includes: the amplitudes of the waves in the first wave group; the first wave group includes: the first type of wave, the second type of wave, and the third type of wave; by first obtaining the first amplitude difference and the second amplitude difference to determine the first amplitude of the first wave group, and then calculating the amplitude anomaly parameter corresponding to the signal segment based on each first amplitude and the amplitude average value, it is possible to determine the variation parameter corresponding to the signal segment, thereby facilitating the improvement of the accuracy of evaluating the first electrocardiogram signal.

[0220] In some embodiments, the first morphological information includes: the widths between the waves in the first wave group;

[0221] The determining of the variation parameter corresponding to each signal segment based on the third quantity corresponding to each signal segment and the first morphological information of the first wave group includes:

[0222] Determining the first width between the first type of wave and the third type of wave;

[0223] Based on each first width and the third quantity, determining the first width average value;

[0224] Based on each first width and the first width average value, calculating the first width anomaly parameter corresponding to the signal segment; and

[0225] Determining the second width between the second type of wave in two adjacent first wave groups;

[0226] Based on each second width and the third quantity, determining the second width average value;

[0227] Based on each second width and the second width average value, calculating the second width anomaly parameter corresponding to the signal segment.

[0228] It can be understood that, in order to improve the accuracy of determining the variation parameter corresponding to each signal segment, the first morphological information includes: the widths between the waves in the first wave group, so that the first width anomaly parameter corresponding to the signal segment can be calculated based on the first width between the first type of wave and the third type of wave and the third quantity; and the second width anomaly parameter corresponding to the signal segment can be calculated based on the second width between the second type of wave in two adjacent first wave groups and the third quantity.

[0229] In the embodiments of the present disclosure, based on the first form information, the first width between the first type of wave and the third type of wave is obtained; then based on each of the first widths and the third quantity, the average value of the first width is determined, and then the first width anomaly parameter is further calculated. And the second width between the second type of waves in two adjacent first wave groups is obtained; then based on each of the second widths and the number corresponding to the second width in the signal segment (i.e., the third quantity minus 1), the average value of the second width is determined, and then the second width anomaly parameter is further calculated.

[0230] Exemplarily, the first type of wave is a Q wave, the second type of wave is an R wave, and the third type of wave is an S wave; calculate the first width qrs_widt between the amplitude of the Q wave and the S wave i h, the first width in the signal segment can be denoted as QRS_WIDTH = {qrs_width1, qrs_width2,... qrs_width n}; based on each first width qrs_widthi and the third quantity n, the average value of the first width E2 is determined; based on each first width qrs_width i and the average value of the first width E2, the first width anomaly parameter corresponding to the signal segment is calculated.

[0231] Therefore, the calculation formula for the first width anomaly parameter corresponding to the signal segment can be as follows:

[0232]

[0233] In formula (2), CV width represents the first width anomaly parameter, n represents the third quantity corresponding to the first wave group in the signal segment, qrs_width i represents each first width in the signal segment, E2 represents the average value of the first width, represents the accumulation of n first widths.

[0234] Another exemplarily, the second type of wave is an R wave; calculate the second width rr between the R wave in the first wave group M1 and the R wave in the first wave group M2 adjacent to the first wave group M1 i , the second width in the signal segment can be denoted as RR = {rr1, rr2,... rr t}; based on each second width rr i and the number t of the second widths in the signal segment (t = n - 1), the average value of the second width E3 is determined; based on each second width rr i and the average value of the second width E3, the second width anomaly parameter corresponding to the signal segment is calculated.

[0235] Therefore, the calculation formula for the second width anomaly parameter corresponding to the signal segment can be as follows:

[0236]

[0237] In formula (3), CV rr represents the second width anomaly parameter, t represents the number of the second widths in the signal segment, rr i represents each of the second widths in the signal segment, E3 represents the average value of the second widths, which represents the accumulation of t second widths.

[0238] In the embodiments of the present disclosure, the first morphological information includes: the widths between the waves in the first wave group; calculating the first width anomaly parameter through each first width and the average value of the first widths; and then calculating the second width anomaly parameter through each second width and the average value of the second widths, so that the variation parameter corresponding to the signal segment can be determined, which is beneficial to improving the accuracy of evaluating the first electrocardiogram signal.

[0239] In some embodiments, the method further includes:

[0240] determining the first wave group with the first amplitude not within the first preset amplitude range as the first wave group in an abnormal state;

[0241] determining the first wave group with the first width not within the preset width range as the first wave group in the abnormal state;

[0242] wherein, the first preset amplitude range is related to the amplitude average value, and the preset width range is related to the average value of the first widths.

[0243] It should be noted that, in order to increase the number of target abnormal waves determined in each signal segment, the first wave group in the abnormal state and the abnormal waves except the second wave group can be determined respectively. Therefore, by setting the first preset amplitude range, the first wave group in the abnormal state is screened from each signal segment; and then by setting the preset width range, the first wave group in the abnormal state is screened from each signal segment.

[0244] Here, the first preset amplitude range and the preset width range can be set arbitrarily according to requirements, as long as the first preset amplitude range is related to the amplitude average value and the preset width range is related to the average value of the first widths. The embodiments of the present disclosure do not limit this.

[0245] Exemplarily, the first preset amplitude range is 0.5E1 to 1.5E1; it is determined whether each obtained first amplitude is within the first preset amplitude range, and the first wave groups with the first amplitudes within the first preset amplitude range are determined as the first wave groups in the normal state, and then these first wave groups are not included in the target abnormal waves; the first wave groups with the first amplitudes not within the first preset amplitude range are determined as the first wave groups in the abnormal state, and then these first wave groups are included in the target abnormal waves.

[0246] Another exemplarily, the preset width range is 0.5E2 to 1.5E2, it is determined whether each obtained first width is within the preset width range, and the first wave groups with the first widths within the preset width range are determined as the first wave groups in the normal state, and then these first wave groups are not included in the target abnormal waves; the first wave groups with the first widths not within the preset width range are determined as the first wave groups in the abnormal state, and then these first wave groups are included in the target abnormal waves.

[0247] In the embodiments of the present disclosure, by setting the first preset amplitude range related to the amplitude average value and the preset width range related to the first width average value, the first wave groups in the abnormal state are determined, so that the first quantity of the first wave groups in the abnormal state can be determined, which is beneficial to improving the accuracy of evaluating the first electrocardiogram signal.

[0248] In some embodiments, the method further includes:

[0249] Determine the second morphological information of each candidate wave except the second wave group in each of the signal segments; wherein, the second morphological information includes the second amplitudes of the respective candidate waves;

[0250] The candidate waves with the second amplitudes not within the second preset amplitude range are determined as the abnormal waves;

[0251] Wherein, the second preset amplitude range is related to the amplitude average value.

[0252] It should be noted that, in order to determine the second quantity of the abnormal waves except the second wave group in each signal segment, it is necessary to first determine the second morphological information of each candidate wave except the second wave group in each of the signal segments, so as to determine the second amplitudes of the respective candidate waves and further judge the second quantity of the abnormal waves.

[0253] It can be understood that, in order to improve the accuracy of determining the second quantity of the abnormal waves, a second preset amplitude range can be set, and it is determined whether each second amplitude is within the second preset amplitude range, so as to screen the abnormal waves from each signal segment.

[0254] Here, the first preset amplitude range and the second preset amplitude range may be the same or different, and the embodiments of the present disclosure do not limit this either.

[0255] Exemplarily, the second preset amplitude range is 0.3E1 to E1. Determine whether each obtained second amplitude is within the second preset amplitude range. The candidate group with the second amplitude within the second preset amplitude range is determined as the wave in the normal state, and this wave is not included in the target abnormal wave; the candidate group with the second amplitude not within the second preset amplitude range is determined as the abnormal wave, and this abnormal wave is included in the target abnormal wave.

[0256] In the embodiments of the present disclosure, by setting the second preset amplitude range related to the amplitude average value, the abnormal waves in the first electrocardiogram signal except the second wave group are determined, so that the second quantity of the abnormal waves can be determined, which is beneficial to improving the accuracy of evaluating the first electrocardiogram signal.

[0257] In some embodiments, determining the baseline drift index of the first electrocardiogram signal includes:

[0258] Determine the peak points of each wave in the first electrocardiogram signal;

[0259] Calculate the rising edge amplitude and the falling edge amplitude corresponding to each peak point;

[0260] Based on the rising edge amplitude, the falling edge amplitude, and the amplitude average value, determine the baseline drift index.

[0261] It can be understood that by determining the peak points of each wave in the first electrocardiogram signal, the rising edge amplitude and the falling edge amplitude corresponding to each peak point can be calculated, so as to further calculate the baseline drift index of the first electrocardiogram signal.

[0262] Here, the rising edge amplitude is the amplitude difference AMP between the peak H of the peak point P and the left valley value V L ; the falling edge amplitude is the amplitude difference AMP between the peak H of the peak point P and the right valley value V up ; R ; down .

[0263] It should be noted that the baseline drift index is generally in the form of a ratio. After obtaining AMP up and AMP down , by calculating the ratios of AMP up , AMP down to the amplitude average value, the baseline drift index of the first electrocardiogram signal is obtained.

[0264] In the embodiments of the present disclosure, the peak points of each wave in the first electrocardiogram signal are first determined, then the rising edge amplitude and the falling edge amplitude corresponding to each peak point are determined, and finally, based on the rising edge amplitude, the falling edge amplitude, and the amplitude average value, the baseline drift index is determined, thereby improving the accuracy of calculating the baseline drift index.

[0265] In some embodiments, the method further includes:

[0266] Based on the rising edge amplitude and the falling edge amplitude corresponding to each peak point, determine the third amplitude difference corresponding to each peak point;

[0267] The determining the baseline drift index based on the rising edge amplitude, the falling edge amplitude, and the amplitude average value includes:

[0268] In the case where it is detected that a plurality of consecutive or adjacent third amplitude differences are all greater than a preset threshold, select the largest third amplitude difference as the baseline drift threshold;

[0269] Based on the baseline drift threshold and the amplitude average value, calculate the baseline drift index.

[0270] It should be noted that, in order to improve the accuracy of calculating the baseline drift index, the third amplitude difference between the rising edge amplitude and the falling edge amplitude corresponding to each peak point can be calculated first; then a preset threshold is set, and each third amplitude difference is compared with the preset threshold, and the baseline drift threshold AMP for calculating the baseline drift index is selected. bd 。

[0271] Here, the preset threshold can be set arbitrarily, and the embodiments of the present disclosure do not limit this.

[0272] It can be understood that, in order to further improve the accuracy of calculating the baseline drift index, in the case where it is detected that a plurality of consecutive or adjacent third amplitude differences are all greater than the preset threshold, the largest third amplitude difference is selected as the baseline drift threshold. Then calculate the ratio between the baseline drift threshold and the amplitude average value, and calculate the baseline drift index of the first electrocardiogram signal.

[0273] Exemplarily, the preset threshold can be the first amplitude average value corresponding to the R wave in the first electrocardiogram signal. If it is detected that 4 consecutive or adjacent third amplitude differences are all greater than the first amplitude average value, select the larger third amplitude difference among the 4 third amplitude differences as the baseline drift threshold; then calculate the ratio between the baseline drift threshold and the amplitude average value, and calculate the baseline drift index.

[0274] In the embodiments of the present disclosure, first, based on the rising edge amplitude and the falling edge amplitude corresponding to each peak point, the third amplitude difference corresponding to each peak point is determined; then, when it is detected that a plurality of consecutive or adjacent third amplitude differences are greater than a preset threshold, the largest third amplitude difference is selected as the baseline drift threshold; finally, based on the baseline drift threshold and the amplitude average value, the baseline drift index is calculated. In this way, by setting the preset threshold, selecting the baseline drift threshold, and further calculating the baseline drift index, the accuracy of calculating the baseline drift index can be improved, which is beneficial to improving the accuracy of evaluating the first electrocardiogram signal.

[0275] In some embodiments, the method further includes:

[0276] Performing a filtering process on the second electrocardiogram signal to obtain a third electrocardiogram signal, a low-frequency noise signal, and a high-frequency noise signal;

[0277] Performing a normalization process on the data in the third electrocardiogram signal to obtain the first electrocardiogram signal.

[0278] It can be understood that by filtering the second electrocardiogram signal (i.e., the directly collected original electrocardiogram signal), a high-frequency noise signal, a low-frequency noise signal, and a third electrocardiogram signal can be obtained, and then by performing a normalization process on the data in the third electrocardiogram signal, the first electrocardiogram signal is obtained.

[0279] Here, the data may include signal-related parameters such as frequency and amplitude, and the embodiments of the present disclosure do not limit this.

[0280] In some embodiments, the third electrocardiogram signal is collected by an electrode and converted into an analog electrocardiogram signal. The filtering process on the analog electrocardiogram signal mainly includes low-pass filtering and high-pass filtering. Among them, low-pass filtering removes high-frequency noise and retains low-frequency electrocardiogram information; high-pass filtering removes low-frequency noise and retains high-frequency electrocardiogram information. The third electrocardiogram signal after the filtering process is converted into a digital signal through an analog-to-digital converter.

[0281] Exemplarily, after filtering the second electrocardiogram signal by using a 40HZ low-pass filter and a 0.5HZ high-pass filter, the third electrocardiogram signal S f ={S f1 ,S f2 ,…S fw}, the high-frequency noise signal S H ={S h1 ,S h2 ,…S hw}, and the low-frequency noise signal S L ={S l1 ,S l2 ,…S lw} can be obtained.

[0282] The standardization processing formula can be as follows:

[0283]

[0284] In formula (4), S K represents the first electrocardiogram signal, w represents the number of signals in the third electrocardiogram signal, and S i represents any one signal in the third electrocardiogram signal, represents the signal with the smallest data in the third electrocardiogram signal, represents the signal with the largest data in the third electrocardiogram signal.

[0285] In the embodiments of the present disclosure, by filtering the second electrocardiogram signal, a third electrocardiogram signal, a low-frequency noise signal, and a high-frequency noise signal are obtained; then, the data in the third electrocardiogram signal is standardized to obtain the first electrocardiogram signal, so that a first electrocardiogram signal after noise reduction and standardization processing can be obtained, which is beneficial to improving the accuracy of evaluating the first electrocardiogram signal.

[0286] In some embodiments, the method further includes:

[0287] Based on the peak values and valley values of each wave in the high-frequency noise signal, determining a third amplitude of the high-frequency noise signal;

[0288] Based on the peak values and valley values of each wave in the low-frequency noise signal, determining a fourth amplitude of the low-frequency noise signal;

[0289] The determining the characteristic information of the first electrocardiogram signal based on the variation parameter of the first electrocardiogram signal, the baseline drift index, and the number of abnormal waves includes:

[0290] Determining the variation parameter, the baseline drift index, the number of abnormal waves, the third amplitude, and the fourth amplitude as the characteristic information of the electrocardiogram signal.

[0291] It can be understood that, in order to further improve the accuracy of evaluating the quality of the first electrocardiogram signal, the third amplitude of the high-frequency noise signal and the fourth amplitude of the low-frequency noise signal can also be used as one of the characteristic information of the first electrocardiogram signal respectively, that is, based on the variation parameter, the baseline drift index, the number of abnormal waves, the third amplitude, and the fourth amplitude, determining the target evaluation result corresponding to the first electrocardiogram signal.

[0292] Here, the third amplitude of the high-frequency noise signal can be determined based on the peak values and valley values of each wave in the high-frequency noise signal; the fourth amplitude of the low-frequency noise signal can be determined based on the peak values and valley values of each wave in the low-frequency noise signal.

[0293] Specifically, after obtaining the high-frequency noise signal, the peak values and valley values of each wave of the high-frequency noise signal can be determined. Based on the peak values of each wave, a first envelope line of the high-frequency noise signal is formed; based on the valley values of each wave, a second envelope line of the high-frequency noise signal is formed. And calculate the first peak average value corresponding to each peak value in the first envelope line and the first valley average value corresponding to each valley value in the second envelope line, then calculate the first average difference between the first peak average value and the first valley average value, and finally determine the first average difference as the third amplitude of the high-frequency noise signal.

[0294] After obtaining the low-frequency noise signal, the peak values and valley values of each wave of the low-frequency noise signal can be determined. Based on the peak values of each wave, a third envelope line of the low-frequency noise signal is formed; based on the valley values of each wave, a fourth envelope line of the low-frequency noise signal is formed. And calculate the second peak average value corresponding to each peak value in the third envelope line and the second valley average value corresponding to each valley value in the fourth envelope line, then calculate the second average difference between the second peak average value and the second valley average value, and finally determine the second average difference as the fourth amplitude of the low-frequency noise signal.

[0295] In the embodiments of the present disclosure, by calculating the third amplitude of the high-frequency noise signal and the fourth amplitude of the low-frequency noise signal, the third amplitude and the fourth amplitude are respectively determined as one of the characteristic information of the electrocardiogram signal, so that more characteristic information is input into the classification model, which is beneficial to improving the accuracy of evaluating the first electrocardiogram signal.

[0296] Figure 2 It is a schematic flowchart of a method for evaluating the quality of an electrocardiogram signal shown according to an exemplary embodiment Figure 2 , such as Figure 2 shown, based on inputting characteristic information such as the baseline drift index, the number of target abnormal waves, and the variation parameter of the first electrocardiogram signal into at least one classification model to obtain the classification results corresponding to each of the classification models; and then based on each of the classification results, obtaining the target evaluation result corresponding to the first electrocardiogram signal. The method mainly includes the following steps:

[0297] In step 201, a second electrocardiogram signal is collected.

[0298] Here, the second electrocardiogram signal (i.e., the original electrocardiogram signal) of the organism can be collected by a data acquisition device.

[0299] In step 202, the second electrocardiogram signal is subjected to filtering processing and normalization processing to obtain the first electrocardiogram signal.

[0300] In some embodiments, the second electrocardiogram signal is filtered to obtain a third electrocardiogram signal, a low-frequency noise signal, and a high-frequency noise signal; the data in the third electrocardiogram signal is normalized to obtain the first electrocardiogram signal.

[0301] In step 203, the first electrocardiogram signal is sliced to obtain a plurality of signal segments of equal length.

[0302] In step 204, the baseline drift index, the number of target abnormal waves, and the variation parameter of the first electrocardiogram signal are determined.

[0303] In some embodiments, after determining the baseline drift index, the number of target abnormal waves, and the variation parameter of the first electrocardiogram signal, based on the variation parameter, the baseline drift index, and the number of abnormal waves of the first electrocardiogram signal, the characteristic information of the first electrocardiogram signal is determined.

[0304] In step 205, the characteristic information is respectively input into a linear classification model, a decision tree classification model, and a support vector machine classification model.

[0305] In some embodiments, by manually labeling electrocardiogram signals of known quality grades, that is, labeling electrocardiogram signals with excellent signal grades and electrocardiogram signals with poor signal grades, and inputting the labeled electrocardiogram signals into a linear classification model, a decision tree classification model, and a support vector machine classification model respectively to train each classification model, thereby obtaining a pre-trained classification model.

[0306] In step 206, the classification results corresponding to each classification model are obtained.

[0307] In step 207, based on each of the classification results, the target evaluation result corresponding to the first electrocardiogram signal is obtained.

[0308] In some embodiments, to facilitate determining the target evaluation result corresponding to the first electrocardiogram signal, each classification result can be scored. For example, an electrocardiogram signal with an excellent signal grade is scored 1 point, and an electrocardiogram signal with a poor signal grade is scored 0 point. Then, the scored results are integrated to obtain the target evaluation result of the first electrocardiogram signal. For example, an evaluation of 0 - 1 point is considered excellent, 2 points is considered good, and 3 points is considered excellent.

[0309] Exemplarily, the accuracy of the classification algorithm can be evaluated by the precision, sensitivity, and accuracy of the classification algorithm.

[0310] Here, the sensitivity Se represents the proportion of samples that are actually positive and are judged to be positive. For example, for the first electrocardiogram signal with an actual excellent quality grade, the proportion of the first electrocardiogram signal that is judged to have an excellent quality grade.

[0311] Taking the first electrocardiogram signal with excellent quality grade as an example, the calculation formula for sensitivity can be as follows:

[0312]

[0313] In formula (5), S e优 represents the sensitivity corresponding to the first electrocardiogram signal with excellent detected quality grade, and TN 优 represents the number of the first electrocardiogram signals with actual excellent quality grade that are judged to be of excellent quality grade; TN1 represents the number of the first electrocardiogram signals with actual excellent quality grade.

[0314] Here, the accuracy P r represents the proportion of samples judged as positive among the samples judged as positive. For example, among all the first electrocardiogram signals judged to be of excellent quality grade, it is the proportion of the first electrocardiogram signals with actual excellent quality grade.

[0315] Taking the first electrocardiogram signal with excellent quality grade as an example, the calculation formula for accuracy can be as follows:

[0316]

[0317] In formula (6), P r优 represents the accuracy corresponding to the first electrocardiogram signal with excellent detected quality grade, and TN 优 represents the number of the first electrocardiogram signals with actual excellent quality grade that are judged to be of excellent quality grade; TN2 represents the number of all the first electrocardiogram signals with excellent quality grade.

[0318] Here, the accuracy rate A c represents the proportion of the number of all correctly classified samples to the total sample data.

[0319] The calculation formula for the accuracy rate can be as follows:

[0320]

[0321] In formula (7), A c represents the accuracy corresponding to the first electrocardiogram signal, N represents the total number of signals in the first electrocardiogram signal; TN 优 represents the number of the first electrocardiogram signals with actual excellent quality grade that are judged to be of excellent quality grade; TN 良 represents the number of the first electrocardiogram signals with actual good quality grade that are judged to be of good quality grade; TN 差 represents the number of the first electrocardiogram signals with actual poor quality grade that are judged to be of poor quality grade.

[0322] Figure 3 It is a schematic diagram showing the quality assessment result of an electrocardiogram (ECG) signal according to an exemplary embodiment. As Figure 3 shown, using the above classification algorithm, the classification results of a total of 6,820 actually collected first ECG signals in the embodiments of the present disclosure are as follows: among them, for the first ECG signals with an actual quality level of excellent, the number of first ECG signals judged to have a quality level of excellent is 3,441; for the first ECG signals with an actual quality level of good, the number of first ECG signals judged to have a quality level of good is 1,293; for the first ECG signals with an actual quality level of poor, the number of first ECG signals judged to have a quality level of poor is 1,735.

[0323] Table 1 is a statistical table of the accuracy of the classification algorithm. As shown in Table 1, the sensitivity, accuracy, and precision of the classification model algorithm in the embodiments of the present disclosure are relatively high. Therefore, the classification model algorithm proposed in the present disclosure can improve the efficiency of evaluating the quality of the first ECG signal, reduce the computational amount and resource consumption, and at the same time improve the accuracy of evaluating the quality of the first ECG signal, so that it can be widely applied to wearable devices.

[0324] Table 1 is a statistical table of the accuracy of the classification algorithm

[0325]

[0326]

[0327] Figure 4 It is a block diagram of an electrocardiogram signal quality assessment device according to an exemplary embodiment. As Figure 4 shown, the device 400 includes:

[0328] A first acquisition module 401, configured to acquire a first electrocardiogram signal;

[0329] A first determination module 402, configured to determine the baseline drift index of the first electrocardiogram signal and the number of target abnormal waves in the first electrocardiogram signal;

[0330] A second determination module 403, configured to determine the characteristic information of the first electrocardiogram signal based on the variation parameter, the baseline drift index, and the number of abnormal waves of the first electrocardiogram signal;

[0331] A second acquisition module 404, configured to perform at least one classification process on the characteristic information to obtain at least one classification result;

[0332] A third acquisition module 405, configured to obtain a target evaluation result corresponding to the first electrocardiogram signal based on each classification result.

[0333] In some embodiments, the second acquisition module 404 is specifically configured to:

[0334] Input the feature information into at least one classification model to obtain classification results corresponding to each of the classification models.

[0335] In some embodiments, the first determination module 402 is specifically configured to:

[0336] Determine a first wave group from the first electrocardiogram signal, and a second wave group formed based on the first wave group and a preset wave;

[0337] Determine a first quantity of the first wave group in an abnormal state, and a second quantity of abnormal waves in the first electrocardiogram signal except for the second wave group;

[0338] The second determination module 403 is specifically configured to:

[0339] Determine the feature information of the first electrocardiogram signal based on the variation parameter of the first electrocardiogram signal, the baseline drift index, the first quantity, and the second quantity.

[0340] In some embodiments, the apparatus 400 further includes:

[0341] A fourth acquisition module, configured to perform slicing processing on the first electrocardiogram signal to obtain a plurality of signal segments of equal length;

[0342] A third determination module, configured to determine the quantity of the first wave group in each of the signal segments as a third quantity;

[0343] A fourth determination module, configured to determine the variation parameter corresponding to each of the signal segments based on the third quantity corresponding to each of the signal segments and the first morphological information of the first wave group;

[0344] A fifth determination module, configured to determine the variation parameter corresponding to each of the signal segments as the variation parameter of the first electrocardiogram signal.

[0345] In some embodiments, the first morphological information includes: the amplitudes of the waves in the first wave group; the first wave group includes: a first type of wave, a second type of wave, and a third type of wave;

[0346] The fourth determination module is specifically configured to:

[0347] Determine a first amplitude difference based on the amplitude of the first type of wave and the amplitude of the second type of wave;

[0348] Determine a second amplitude difference based on the amplitude of the second type of wave and the amplitude of the third type of wave;

[0349] Based on the first amplitude difference and the second amplitude difference, determine the first amplitude of the first wave group;

[0350] Based on each of the first amplitudes and the third quantity, determine the average amplitude;

[0351] Based on each of the first amplitudes and the average amplitude, calculate the amplitude anomaly parameter corresponding to the signal segment.

[0352] In some embodiments, the first morphological information includes: the width between each wave in the first wave group;

[0353] The fourth determination module is further configured to:

[0354] Determine the first width between the first type of wave and the third type of wave;

[0355] Based on each of the first widths and the third quantity, determine the average first width;

[0356] Based on each of the first widths and the average first width, calculate the first width anomaly parameter corresponding to the signal segment; and

[0357] Determine the second width between the second type of waves in two adjacent first wave groups;

[0358] Based on each of the second widths and the third quantity, determine the average second width;

[0359] Based on each of the second widths and the average second width, calculate the second width anomaly parameter corresponding to the signal segment.

[0360] In some embodiments, the apparatus 400 further includes:

[0361] A sixth determination module, configured to determine, as the first wave group in an abnormal state, the first wave group whose first amplitude is not within the first preset amplitude range;

[0362] A seventh determination module, configured to determine, as the first wave group in the abnormal state, the first wave group whose first width is not within the preset width range;

[0363] Wherein, the first preset amplitude range is related to the average amplitude, and the preset width range is related to the average first width.

[0364] In some embodiments, the apparatus 400 further includes:

[0365] An eighth determination module, configured to determine second morphological information of each candidate wave in each of the signal segments except for the second wave group; wherein, the second morphological information includes second amplitudes of each of the candidate waves.

[0366] A ninth determination module, configured to determine, as the abnormal wave, a candidate wave whose second amplitude is not within a second preset amplitude range.

[0367] Wherein, the second preset amplitude range is related to the amplitude average value.

[0368] In some embodiments, the first determination module 402 includes:

[0369] A first sub-module, configured to determine peak points of each wave in the first electrocardiogram signal.

[0370] A second sub-module, configured to calculate a rising edge amplitude and a falling edge amplitude corresponding to each of the peak points.

[0371] A third sub-module, configured to determine the baseline drift index based on the rising edge amplitude, the falling edge amplitude, and the amplitude average value.

[0372] In some embodiments, the apparatus 400 further includes:

[0373] A tenth determination module, configured to determine a third amplitude difference corresponding to each of the peak points based on the rising edge amplitude and the falling edge amplitude corresponding to each of the peak points.

[0374] The third sub-module is specifically configured to:

[0375] In a case where it is detected that a plurality of consecutive or adjacent third amplitude differences are greater than a preset threshold, select the largest third amplitude difference as the baseline drift threshold.

[0376] Calculate the baseline drift index based on the baseline drift threshold and the amplitude average value.

[0377] In some embodiments, the apparatus 400 further includes:

[0378] A first processing module, configured to perform filtering processing on a second electrocardiogram signal to obtain a third electrocardiogram signal, a low-frequency noise signal, and a high-frequency noise signal.

[0379] A second processing module, configured to perform normalization processing on data in the third electrocardiogram signal to obtain the first electrocardiogram signal.

[0380] In some embodiments, the apparatus 400 further includes:

[0381] A third processing module, configured to determine a third amplitude of the high-frequency noise signal based on peaks and valleys of each wave in the high-frequency noise signal;

[0382] A fourth processing module, configured to determine a fourth amplitude of the low-frequency noise signal based on peaks and valleys of each wave in the low-frequency noise signal;

[0383] The second determination module 403 is specifically configured to

[0384] determine the mutation parameter, the baseline drift index, the number of abnormal waves, the third amplitude, and the fourth amplitude as characteristic information of the electrocardiogram signal.

[0385] Regarding the device in the above embodiments, the specific manners in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated herein.

[0386] Figure 5 is a hardware structure block diagram of an electronic device shown according to an exemplary embodiment Figure 1 . For example, the device 800 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, etc.

[0387] Referring to Figure 5 , the device 800 may include one or more of the following components: a processing component 802, a memory 804, a power supply component 806, a multimedia component 808, an audio component 810, an input / output (I / O) interface 812, a sensor component 814, and a communication component 816.

[0388] The processing component 802 generally controls the overall operation of the device 800, such as operations associated with display, telephone calls, data communication, camera operations, and recording operations. The processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the above method. In addition, the processing component 802 may include one or more modules to facilitate the interaction between the processing component 802 and other components. For example, the processing component 802 may include a multimedia module to facilitate the interaction between the multimedia component 808 and the processing component 802.

[0389] The memory 804 is configured to store various types of data to support the operation of the device 800. Examples of such data include instructions for any application or method operating on the device 800, contact data, phone book data, messages, pictures, videos, and the like. The memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disks, or optical disks.

[0390] The power supply component 806 provides power to various components of the device 800. The power supply component 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the device 800.

[0391] The multimedia component 808 includes a screen that provides an output interface between the device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen can be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors can not only sense the boundaries of touch or swipe actions but also detect the duration and pressure associated with the touch or swipe operation. In some embodiments, the multimedia component 808 includes a front camera and / or a rear camera. When the device 800 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera can receive external multimedia data. Each of the front camera and the rear camera can be a fixed optical lens system or have focal length and optical zoom capabilities.

[0392] The audio component 810 is configured to output and / or input audio signals. For example, the audio component 810 includes a microphone (MIC) that is configured to receive external audio signals when the device 800 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals can be further stored in the memory 804 or transmitted via the communication component 816. In some embodiments, the audio component 810 further includes a speaker for outputting audio signals.

[0393] The I / O interface 812 provides an interface between the processing component 802 and a peripheral interface module, which can be a keyboard, a click wheel, buttons, etc. These buttons can include, but are not limited to: a home button, a volume button, a power-on button, and a lock button.

[0394] The sensor assembly 814 includes one or more sensors for providing a status assessment of various aspects of the device 800. For example, the sensor assembly 814 can detect the on / off state of the device 800, the relative positioning of components, such as the display and keypad of the device 800. The sensor assembly 814 can also detect a change in the position of the device 800 or a component of the device 800, the presence or absence of user contact with the device 800, the orientation or acceleration / deceleration of the device 800, and a change in the temperature of the device 800. The sensor assembly 814 can include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 814 can also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 814 can also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0395] The communication component 816 is configured to facilitate communication between the device 800 and other devices in a wired or wireless manner. The device 800 can access a wireless network based on communication standards, such as WiFi, 4G, or 5G, or a combination thereof. In an exemplary embodiment, the communication component 816 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 816 further includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0396] In an exemplary embodiment, the device 800 can be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components for performing the above method.

[0397] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 804 including instructions, and the above instructions can be executed by the processor 820 of the device 800 to complete the above method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0398] A non-transitory computer-readable storage medium, when the instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to execute a method for electrocardiogram signal quality assessment, the method including:

[0399] Obtain a first electrocardiogram signal;

[0400] Determine the baseline drift index of the first electrocardiogram signal and the number of target abnormal waves in the first electrocardiogram signal;

[0401] Based on the variation parameter of the first electrocardiogram signal, the baseline drift index, and the number of abnormal waves, determine the characteristic information of the first electrocardiogram signal;

[0402] Perform at least one classification process on the characteristic information to obtain at least one classification result;

[0403] Based on each of the classification results, obtain the target evaluation result corresponding to the first electrocardiogram signal.

[0404] Figure 6 It is a hardware structure block diagram of an electronic device shown according to an exemplary embodiment Figure 2 . For example, device 1900 may be provided as a server. Referring Figure 6 , device 1900 includes a processing component 1922, which further includes one or more processors, and memory resources represented by a memory 1932 for storing instructions executable by the processing component 1922, such as application programs. The application programs stored in the memory 1932 may include one or more modules each corresponding to a set of instructions. In addition, the processing component 1922 is configured to execute instructions to perform the above method:

[0405] Obtain a first electrocardiogram signal;

[0406] Determine the baseline drift index of the first electrocardiogram signal and the number of target abnormal waves in the first electrocardiogram signal;

[0407] Based on the variation parameter of the first electrocardiogram signal, the baseline drift index, and the number of abnormal waves, determine the characteristic information of the first electrocardiogram signal;

[0408] Perform at least one classification process on the characteristic information to obtain at least one classification result;

[0409] Based on each of the classification results, obtain the target evaluation result corresponding to the first electrocardiogram signal.

[0410] The apparatus 1900 may further include a power supply component 1926 configured to perform power management of the apparatus 1900, a wired or wireless network interface 1950 configured to connect the apparatus 1900 to a network, and an input / output (I / O) interface 1958. The apparatus 1900 may operate based on an operating system stored in the memory 1932, such as Windows ServerTM, MacOS XTM, UnixTM, LinuxTM, FreeBSDTM or the like.

[0411] Other embodiments of the present disclosure will be readily apparent to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. The present disclosure is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common general knowledge or conventional technical means in the technical field not disclosed herein. The specification and examples are only exemplary, and the true scope and spirit of the present disclosure are pointed out by the following claims.

[0412] It should be understood that the present disclosure is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present disclosure is only limited by the appended claims.

Claims

1. A method for evaluating the quality of electrocardiogram signals, characterized in that, The method includes: Obtaining a first electrocardiogram signal; Determining a baseline drift index of the first electrocardiogram signal and the number of target abnormal waves in the first electrocardiogram signal; Determining characteristic information of the first electrocardiogram signal based on a variation parameter of the first electrocardiogram signal, the baseline drift index, and the number of abnormal waves; Performing at least one classification process on the characteristic information to obtain at least one classification result; Obtaining a target evaluation result corresponding to the first electrocardiogram signal based on each of the classification results.

2. The method according to claim 1, wherein The performing at least one classification process on the characteristic information to obtain at least one classification result includes: Inputting the characteristic information into at least one classification model to obtain a classification result corresponding to each of the classification models.

3. The method according to claim 1, wherein The determining the number of target abnormal waves in the first electrocardiogram signal includes: Determining a first wave group from the first electrocardiogram signal and a second wave group formed based on the first wave group and a preset wave form; Determining a first number of the first wave groups in an abnormal state and a second number of abnormal waves in the first electrocardiogram signal excluding the second wave group; The determining the characteristic information of the first electrocardiogram signal based on the variation parameter of the first electrocardiogram signal, the baseline drift index, and the number of abnormal waves includes: Determining the characteristic information of the first electrocardiogram signal based on the variation parameter of the first electrocardiogram signal, the baseline drift index, the first number, and the second number.

4. The method according to claim 3, characterized in that The method further includes: Performing slicing processing on the first electrocardiogram signal to obtain a plurality of signal segments with equal lengths; Determining the number of the first wave groups in each of the signal segments as a third number; Determining a variation parameter corresponding to each of the signal segments based on the third number corresponding to each of the signal segments and first morphological information of the first wave group; Determining the variation parameters corresponding to each of the signal segments as the variation parameter of the first electrocardiogram signal.

5. The method according to claim 4, wherein The first morphological information includes: amplitudes of each wave in the first wave group; the first wave group includes: a first type of wave, a second type of wave, and a third type of wave; The determining the variation parameter corresponding to each of the signal segments based on the third number corresponding to each of the signal segments and the first morphological information of the first wave group includes: Determining a first amplitude difference based on the amplitude of the first type of wave and the amplitude of the second type of wave; Determining a second amplitude difference based on the amplitude of the second type of wave and the amplitude of the third type of wave; Determining a first amplitude of the first wave group based on the first amplitude difference and the second amplitude difference; Determining an amplitude average value based on each of the first amplitudes and the third number; Calculating an amplitude abnormality parameter corresponding to the signal segment based on each of the first amplitudes and the amplitude average value.

6. The method according to claim 5, wherein The first morphological information includes: widths between each wave in the first wave group; The determining the variation parameter corresponding to each of the signal segments based on the third number corresponding to each of the signal segments and the first morphological information of the first wave group includes: Determining a first width between the first type of wave and the third type of wave; Determine the average value of the first width based on each of the first widths and the third quantity; Calculate the first width anomaly parameter corresponding to the signal segment based on each of the first widths and the average value of the first width; and Determine the second width between the second type of waves in two adjacent first wave groups; Determine the average value of the second width based on each of the second widths and the third quantity; Calculate the second width anomaly parameter corresponding to the signal segment based on each of the second widths and the average value of the second width.

7. The method according to claim 6, wherein The method further includes: Determine the first wave group with the first amplitude not within the first preset amplitude range as the first wave group in an abnormal state; Determine the first wave group with the first width not within the preset width range as the first wave group in the abnormal state; Wherein, the first preset amplitude range is related to the average value of the amplitudes, and the preset width range is related to the average value of the first width.

8. The method according to claim 5, characterized in that, The method further includes: Determine the second morphological information of each candidate wave in each of the signal segments except the second wave group; wherein, the second morphological information includes the second amplitude of each candidate wave; Determine the candidate wave with the second amplitude not within the second preset amplitude range as the abnormal wave; Wherein, the second preset amplitude range is related to the average value of the amplitudes.

9. The method according to claim 5, characterized in that, The determining the baseline drift index of the first electrocardiogram signal includes: Determine the peak points of each wave in the first electrocardiogram signal; Calculate the rising edge amplitude and the falling edge amplitude corresponding to each of the peak points; Determine the baseline drift index based on the rising edge amplitude, the falling edge amplitude, and the average value of the amplitudes.

10. The method according to claim 9, wherein The method further includes: Determine the third amplitude difference corresponding to each of the peak points based on the rising edge amplitude and the falling edge amplitude corresponding to each of the peak points; The determining the baseline drift index based on the rising edge amplitude, the falling edge amplitude, and the average value of the amplitudes includes: When it is detected that a plurality of consecutive or adjacent third amplitude differences are greater than a preset threshold, select the largest third amplitude difference as the baseline drift threshold; Calculate the baseline drift index based on the baseline drift threshold and the average value of the amplitudes.

11. The method according to any one of claims 1 to 10, characterized in that, The method further includes: Perform filtering processing on the second electrocardiogram signal to obtain a third electrocardiogram signal, a low-frequency noise signal, and a high-frequency noise signal; Perform normalization processing on the data in the third electrocardiogram signal to obtain the first electrocardiogram signal.

12. The method according to claim 11, wherein The method further includes: Determine the third amplitude of the high-frequency noise signal based on the peak and valley values of each wave in the high-frequency noise signal; Determine the fourth amplitude of the low-frequency noise signal based on the peak and valley values of each wave in the low-frequency noise signal; The determining the characteristic information of the first electrocardiogram signal based on the variation parameter of the first electrocardiogram signal, the baseline drift index, and the number of abnormal waves includes: Determine the variation parameter, the baseline drift index, the number of abnormal waves, the third amplitude, and the fourth amplitude as the characteristic information of the electrocardiogram signal.

13. An electrocardiogram signal quality assessment device, characterized in that, The device includes: A first acquisition module, configured to acquire a first electrocardiogram signal; A first determination module, configured to determine a baseline drift index of the first electrocardiogram signal and the number of target abnormal waves in the first electrocardiogram signal; A second determination module, configured to determine feature information of the first electrocardiogram signal based on a variation parameter of the first electrocardiogram signal, the baseline drift index, and the number of abnormal waves; A second acquisition module, configured to perform at least one classification process on the feature information to obtain at least one classification result; A third acquisition module, configured to obtain a target evaluation result corresponding to the first electrocardiogram signal based on each of the classification results.

14. An electronic device, characterized in that, Comprising: A processor; A memory for storing instructions executable by the processor; Wherein, the processor is configured to: when executing the executable command, implement the steps in the electrocardiogram signal quality evaluation method according to any one of claims 1 to 12.

15. A non-transitory computer-readable storage medium, when instructions in the storage medium are executed by a processor of an electronic device, implement the steps in the electrocardiogram signal quality evaluation method according to any one of claims 1 to 12.

Images