Supraventricular tachycardia detection method, system, storage medium, and device
By acquiring, preprocessing, and bandpass filtering electrocardiogram signals, combined with phase amplitude coupling algorithm and ROC curve evaluation, a highly efficient and accurate diagnosis of supraventricular tachycardia was achieved, solving the problems of misdiagnosis and missed diagnosis in traditional methods.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHENZHEN COMEN MEDICAL INSTR
- Filing Date
- 2023-07-25
- Publication Date
- 2026-07-10
AI Technical Summary
Existing technologies struggle to accurately distinguish supraventricular tachycardia from other arrhythmias, and traditional diagnostic methods suffer from misdiagnosis and missed diagnosis, lacking an efficient, safe, and simple diagnostic approach.
By acquiring and preprocessing electrocardiogram (ECG) signals, bandpass filtering and phase-amplitude coupling algorithms are used to determine the threshold set and the optimal discrimination threshold. ROC curves and Youden's index are used to evaluate whether the ECG signal is supraventricular tachycardia.
It improves the accuracy of identifying supraventricular tachycardia, reduces false positives and false negatives, and provides patients with a more efficient, safe, and simple detection method that is easier to accept.
Smart Images
Figure CN116869549B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of electrocardiogram signal detection technology, and in particular to methods, systems, storage media and devices for detecting supraventricular tachycardia. Background Technology
[0002] Medical devices such as pacemakers, defibrillators, or implantable cardioverter defibrillators deliver therapeutic electrical stimulation to a patient's heart via one or more medical leads and / or electrodes on the device's housing. This electrical stimulation may include signals, such as pulses or shocks for pacing, cardioversion, or defibrillation. In some cases, the device may sense the heart's inherent depolarization and control the delivery of stimulation signals to the heart based on the sensed depolarization. When an abnormal heart rhythm is detected, such as bradycardia, tachycardia, or arrhythmia, appropriate electrical stimulation signals or multiple signals may be delivered to restore or maintain a more normal heart rhythm.
[0003] Typically, medical devices detect tachycardia or atrial or ventricular fibrillation based on the interval between depolarizations, which is a function of the depolarization rate. Therefore, when a medical device detects ventricular tachycardia with an interval between ventricular depolarizations less than a first threshold and ventricular fibrillation with an interval between ventricular depolarizations less than a second threshold, along with other characteristics such as rate variability or electrocardiographic morphology, these have been used to classify or differentiate various types of arrhythmias.
[0004] In some cases, a rapid ventricular depolarization rate may be a result of sinus tachycardia or atrial arrhythmia, known as supraventricular tachycardia. Some medical devices differentiate between ventricular tachycardia and supraventricular tachycardia by comparing the rate or interval of ventricular and atrial depolarization, but this comparison may not be effective in distinguishing between ventricular tachycardia and supraventricular tachycardia in all cases.
[0005] Supraventricular tachycardia (SVT) is one of the most common arrhythmias in clinical practice. The traditional definition of SVT is tachycardia originating above the bifurcation of the His bundle. As a common cardiological disease, SVT is characterized by sudden onset and sudden termination. It can reduce the blood perfusion of the heart and blood vessels, increase the risk of hypotension, heart failure, etc., and seriously affect the patient's quality of life.
[0006] Because of its short onset, supraventricular tachycardia (SVT) currently lacks a definitive diagnostic method. Furthermore, the pathogenesis of SVT is complex, and compared to other types of arrhythmias, patients with SVT often have concealed accessory pathways, severely impacting treatment outcomes and significantly disrupting daily life. Currently, the diagnosis of SVT primarily relies on electrocardiogram (ECG) examinations, including surface ECG and Holter monitoring. While these methods are valuable, quick, non-invasive, convenient, and inexpensive, their diagnostic capabilities still require improvement and remain prone to misdiagnosis and missed diagnosis. Therefore, there is an urgent need to find a more accurate, efficient, safe, and simple examination method that is more readily accepted by patients to reduce the misdiagnosis and missed diagnosis rates of SVT. Summary of the Invention
[0007] Therefore, it is necessary to propose a method for detecting supraventricular tachycardia to address the above problems.
[0008] A method for detecting supraventricular tachycardia, the method comprising the following steps:
[0009] The raw supraventricular tachycardia (SVT) signal and the raw non-SVT signal are acquired and preprocessed to obtain the first SVT signal and the first non-SVT signal.
[0010] Bandpass filtering is performed on the first supraventricular tachycardia (SVT) signal and the first non-SVT signal to obtain a second SVT signal and a second non-SVT signal with a fixed frequency band.
[0011] The first threshold set corresponding to the second supraventricular velocities signal and the second threshold set corresponding to the second non-supraventricular velocities signal are determined based on the phase-amplitude coupling algorithm.
[0012] The optimal discrimination threshold is determined based on the first threshold set and the second threshold set;
[0013] The determination of whether the ECG signal to be detected is supraventricular tachycardia is based on the comparison between the optimal discrimination threshold and the modulation index of the ECG signal to be detected.
[0014] In the above scheme, the step of acquiring the original supraventricular tachycardia (SVT) signal and the original non-SVT signal, and preprocessing them to obtain the first SVT signal and the first non-SVT signal, specifically includes:
[0015] The original supraventricular tachycardia (SVT) signal and the original non-SVT signal are subjected to noise reduction and denoising processing, and the processed original SVT signal and the original non-SVT signal are normalized.
[0016] In the above scheme, the step of bandpass filtering the first supraventricular tachycardia (SVT) signal and the first non-SVT signal to obtain a second SVT signal and a second non-SVT signal with a fixed frequency band specifically includes:
[0017] Bandpass filtering is performed on the first supraventricular tachycardia (SVT) signal and the first non-SVT signal to extract the second SVT signal with a frequency band of 3Hz~9Hz and 10Hz~16Hz, and the second non-SVT signal with a frequency band of 3Hz~9Hz and 10Hz~16Hz, respectively.
[0018] In the above scheme, after bandpass filtering the first supraventricular tachycardia (SVT) signal and the first non-SVT signal to obtain the second SVT signal and the second non-SVT signal with a fixed frequency band, the method further includes: windowing the second SVT signal and the second non-SVT signal with the fixed frequency band.
[0019] In the above scheme, the phase-amplitude coupling algorithm specifically includes:
[0020] The high-frequency amplitudes in the low-frequency phases of the second supraventricular tachycardia signal and the second non-suprasound tachycardia signal are normalized to determine the first amplitude.
[0021] Obtain the Shannon entropy of the first amplitude;
[0022] Obtain the Kullback-Leibler distance between the second supraventricular tachycardia signal and the second non-supraventricular tachycardia signal;
[0023] Obtain a single second threshold MI value.
[0024] In the above scheme, the phase amplitude coupling algorithm normalizes the high-frequency amplitude in the low-frequency phase of the second supraventricular velocities (SNV) signal and the second non-SNV signal. The specific expression is as follows:
[0025]
[0026] Where a is the average amplitude of a single bin, k is the running index of bins, N is the total number of bins, and p is a vector with N values;
[0027] The Shannon entropy of the first amplitude is obtained by the following expression:
[0028]
[0029] Where p is a vector of the normalized average amplitude of each phase bin, and N is the total number of bins;
[0030] The Kullback-Leibler distance is obtained using the following expression:
[0031] KL(U,X)=logN-H(p)
[0032] Where U represents a uniform distribution, X represents the data distribution, and N represents the total number of bins;
[0033] The specific expression for obtaining the MI value is:
[0034]
[0035] In the formula, KL(U,X) is the Kullback-Leibler distance, and N is the total number of bins.
[0036] In the above scheme, determining the optimal discrimination threshold based on the first threshold set and the second threshold set specifically includes:
[0037] Construct an ROC curve using a first threshold set and a second threshold set;
[0038] Obtain the AUC value and Youden index from the ROC curve;
[0039] The optimal discrimination threshold is determined by combining the AUC value with the Youden index and based on the magnitude of the AUC value × (1 - Youden index).
[0040] This application also proposes a supraventricular tachycardia detection system, characterized in that the system includes: an electrocardiogram signal acquisition unit, an electrocardiogram signal processing unit, a threshold set acquisition unit, an optimal discrimination threshold determination unit, and a detection unit;
[0041] The ECG signal acquisition unit is used to acquire raw supraventricular tachycardia signals, raw non-supraventricular tachycardia signals, and ECG signals to be detected;
[0042] The ECG signal processing unit is used to preprocess and bandpass filter the original supraventricular tachycardia signal and the original non-supraventricular tachycardia signal to obtain a second supraventricular tachycardia signal and a second non-supraventricular tachycardia signal in a fixed frequency band.
[0043] The threshold set acquisition unit is used to determine the first threshold set corresponding to the second supraventricular velocities signal and the second threshold set corresponding to the second non-supraventricular velocities signal according to the phase amplitude coupling algorithm.
[0044] The optimal distinction threshold determination unit is used to determine the optimal distinction threshold based on the first threshold set and the second threshold set.
[0045] The detection unit is used to determine whether the electrocardiogram signal is supraventricular tachycardia based on the comparison between the optimal discrimination threshold and the modulation index of the electrocardiogram signal.
[0046] This application also proposes a readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the following steps:
[0047] The raw supraventricular tachycardia (SVT) signal and the raw non-SVT signal are acquired and preprocessed to obtain the first SVT signal and the first non-SVT signal.
[0048] Bandpass filtering is performed on the first supraventricular tachycardia (SVT) signal and the first non-SVT signal to obtain a second SVT signal and a second non-SVT signal with a fixed frequency band.
[0049] The first threshold set corresponding to the second supraventricular velocities signal and the second threshold set corresponding to the second non-supraventricular velocities signal are determined based on the phase-amplitude coupling algorithm.
[0050] The optimal discrimination threshold is determined based on the first threshold set and the second threshold set;
[0051] The determination of whether the ECG signal to be detected is supraventricular tachycardia is based on the comparison between the optimal discrimination threshold and the modulation index of the ECG signal to be detected.
[0052] This application also proposes a computer device, including a memory and a processor, wherein the memory stores a computer program, and the computer program is executed by the processor in the following steps:
[0053] The raw supraventricular tachycardia (SVT) signal and the raw non-SVT signal are acquired and preprocessed to obtain the first SVT signal and the first non-SVT signal.
[0054] Bandpass filtering is performed on the first supraventricular tachycardia (SVT) signal and the first non-SVT signal to obtain a second SVT signal and a second non-SVT signal with a fixed frequency band.
[0055] The first threshold set corresponding to the second supraventricular velocities signal and the second threshold set corresponding to the second non-supraventricular velocities signal are determined based on the phase-amplitude coupling algorithm.
[0056] The optimal discrimination threshold is determined based on the first threshold set and the second threshold set;
[0057] The determination of whether the ECG signal to be detected is supraventricular tachycardia is based on the comparison between the optimal discrimination threshold and the modulation index of the ECG signal to be detected.
[0058] The embodiments of the present invention have the following beneficial effects: First, the original supraventricular tachycardia (SVT) signal and the original non-SVT signal are acquired and preprocessed to obtain a first SVT signal and a first non-SVT signal; bandpass filtering is performed on the first SVT signal and the first non-SVT signal to obtain a second SVT signal and a second non-SVT signal with a fixed frequency band; then, a first threshold set corresponding to the second SVT signal and a second threshold set corresponding to the second non-SVT signal are determined according to the phase amplitude coupling algorithm; an optimal discrimination threshold is determined according to the first threshold set and the second threshold set; finally, the comparison relationship between the optimal discrimination threshold and the modulation index of the ECG signal to be detected is used to determine whether the ECG signal to be detected is supraventricular tachycardia. This detection method can improve the accuracy of supraventricular tachycardia identification, reduce false detection and false negative detection, and is a more efficient, safe and simple detection method that is more acceptable to patients. Attached Figure Description
[0059] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0060] in:
[0061] Figure 1 This is a schematic diagram of a supraventricular tachycardia detection method in one embodiment;
[0062] Figure 2 This is a flowchart illustrating the phase-amplitude coupling algorithm in one embodiment;
[0063] Figure 3 This is a flowchart illustrating the process of determining the optimal discrimination threshold based on a first threshold set and a second threshold set in one embodiment.
[0064] Figure 4 This is a flowchart illustrating the process of classifying thresholds and plotting ROC curves in one embodiment. Detailed Implementation
[0065] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0066] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention can be practiced without one or more of these details. In other instances, certain technical features well-known in the art have not been described to avoid obscuring the invention. It should be understood that the invention can be embodied in various forms and should not be construed as limited to the embodiments set forth herein. Rather, providing these embodiments will make the disclosure thorough and complete, and will fully convey the scope of the invention to those skilled in the art.
[0067] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. When used herein, the singular forms “a,” “an,” and “the” are also intended to include the plural forms, unless the context clearly indicates otherwise. The terms “comprising” and / or “including,” when used in this specification, identify the presence of said features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups. When used herein, the term “and / or” includes any and all combinations of the associated listed items.
[0068] To fully understand the present invention, a detailed structure will be presented in the following description in order to illustrate the technical solution proposed by the present invention. Optional embodiments of the present invention are described in detail below. However, in addition to these detailed descriptions, the present invention may have other embodiments.
[0069] like Figure 1 As shown, in one embodiment, a method for detecting supraventricular tachycardia is provided. This method includes steps S101 to S105, detailed below:
[0070] S101. Acquire the original supraventricular tachycardia (SVT) signal and the original non-SVT signal and preprocess them to obtain the first SVT signal and the first non-SVT signal.
[0071] In some embodiments, the preprocessing step includes: performing noise reduction and de-noising processing on the original supraventricular tachycardia (SVT) signal and the original non-SVT signal, and normalizing the processed original SVT signal and the original non-SVT signal to avoid the influence of small differences in the amplitude of the electrocardiogram signal on the results.
[0072] S102. Bandpass filter is applied to the first supraventricular tachycardia (SVT) signal and the first non-sVT signal to obtain a second SVT signal and a second non-sVT signal with a fixed frequency band.
[0073] In some embodiments, the first supraventricular tachycardia (SVT) signal and the first non-SVT signal are bandpass filtered to extract the second SVT signal with a frequency band of 3Hz-9Hz and 10Hz-16Hz, and the second non-SVT signal with a frequency band of 3Hz-9Hz and 10Hz-16Hz, respectively. Specifically, because the energy of the QRS complex wave is mainly concentrated around 8Hz-16Hz and the energy of the P wave peak is mainly concentrated around 3Hz-12Hz, the signal is passed through a bandpass filter with a center frequency of 10Hz to remove power frequency interference and noise interference, while enhancing the QRS wave and the P wave. The frequency bands where the P wave and ORS wave energy are concentrated are selected because the waveform characteristics of atrial premature beats related to supraventricular tachycardia are the premature appearance of atrial ectopic P' waves, which have a different shape from sinus P waves.
[0074] In some embodiments, after bandpass filtering the first supraventricular tachycardia (SVT) signal and the first non-sVT signal to obtain a second SVT signal and a second non-sVT signal with a fixed frequency band, the method further includes: windowing the second SVT signal and the second non-sVT signal with the fixed frequency band. Windowing helps to reduce the influence of the Gibbs phenomenon and can ensure the quality of the extracted signal results.
[0075] S103. Determine the first threshold set corresponding to the second supraventricular velocity signal and the second threshold set corresponding to the second non-supraventricular velocity signal according to the phase amplitude coupling algorithm.
[0076] Specifically, both the first threshold set and the second threshold set contain multiple MI values. The MI values in the first threshold set are helpful in distinguishing supraventricular tachycardia signals (supraventricular tachycardia signals), while the MI values in the second threshold set are helpful in distinguishing non-supraventricular tachycardia signals (non-supraventricular tachycardia signals).
[0077] like Figure 2 As shown, in some embodiments, the phase-amplitude coupling algorithm specifically includes:
[0078] S301. Normalize the high-frequency amplitude in the low-frequency phase of the second supraventricular tachycardia signal and the second non-suprasound tachycardia signal to determine the first amplitude;
[0079] S302, Obtain the Shannon entropy of the first amplitude;
[0080] Shannon entropy refers to the complexity of data information, or the degree of irregularity. It is used to measure the uncertainty of variables. The greater the uncertainty of variables, the higher the Shannon entropy.
[0081] S303, Obtain the Kullback-Leibler distance between the second supraventricular tachycardia signal and the second non-suprastraventricular tachycardia signal;
[0082] The Kullback-Leibler distance, also known as relative entropy, measures the difference between two probability distributions in the same event space, i.e., it measures the similarity between two probability distributions.
[0083] S304. Obtain a single second threshold MI value.
[0084] In some embodiments, Tort's phase-amplitude coupling algorithm can be used to divide all possible phase bins from -180 degrees to 180 degrees into arbitrarily chosen numbers. The example below uses 18 bins, each 20 degrees:
[0085] The high-frequency amplitudes in the low-frequency phase of the second supraventricular tachycardia (SVT) signal and the second non-SVT signal are normalized, and the specific expression is as follows:
[0086]
[0087] Where a is the average amplitude of a single bin, k is the running index of bins, N is the total number of bins, and p is a vector with N values;
[0088] Shannon entropy is a measure of the intrinsic information content of a variable. If the Shannon entropy is not maximum, the variable is considered redundant and predictable. The Shannon entropy is maximum when the amplitude of each phase bin is equal. The specific expression is:
[0089]
[0090] Where p is a vector of the normalized average amplitude of each phase bin, and N is the total number of bins;
[0091] Phase-amplitude coupling is defined as a distribution that significantly deviates from a uniform distribution, while the Kullback-Leibler distance is a measure of the difference between the two distributions. The Kullback-Leibler distance is specifically expressed as follows:
[0092] KL(U,X)=logN-H(p)
[0093] Where U represents a uniform distribution, X represents the data distribution, and N represents the total number of bins;
[0094] The specific expression for obtaining the MI value is:
[0095]
[0096] In the formula, KL(U,X) is the Kullback-Leibler distance, and N is the total number of bins.
[0097] This method can obtain multiple MI values. In some embodiments, ROC curves can be plotted using multiple MI values, and the AUC value can be used to select the signal with the strongest ability to distinguish between supraventricular tachycardia signals and non-supraventricular tachycardia signals.
[0098] S104. Determine the optimal discrimination threshold based on the first threshold set and the second threshold set;
[0099] like Figure 3 As shown, in some embodiments, the optimal discrimination threshold is determined based on a first threshold set and a second threshold set, specifically including:
[0100] S401. Construct an ROC curve using the first threshold set and the second threshold set;
[0101] The ROC curve (Receiver Operating Characteristic Curve) is a method used to evaluate the performance of classification models. The ROC curve expresses the relationship between the true positive rate and the false positive rate of the model at different thresholds.
[0102] In binary classification problems, the true positive rate represents the proportion of samples correctly predicted as positive out of the total number of samples that are actually positive, while the false positive rate represents the proportion of samples incorrectly predicted as positive out of the total number of samples that are actually negative.
[0103] S402. Obtain the AUC value and Youden index from the ROC curve;
[0104] AUC (Area Under the Curve) is a metric used to evaluate the performance of classification models. It is typically used to distinguish the energy difference between positive and negative samples. In signal processing, the strongest signal usually corresponds to a positive sample (i.e., a useful signal), while the weakest signal corresponds to a negative sample (i.e., a noise signal). Therefore, the AUC value can be used to distinguish the strongest signal. The performance of the model is usually evaluated by calculating the area under the curve (Area Under the Curve, AUC). The larger the AUC, the better the model performance.
[0105] S403. Combine the AUC value with the Youden index, and determine the optimal discrimination threshold based on the magnitude of AUC value × (1 - Youden index).
[0106] The Youden index is a metric used to measure the uncertainty of a classifier. The value ranges from 0 to 1. The smaller the value, the lower the uncertainty of the classifier, meaning the more reliable the classification result.
[0107] By multiplying the AUC value by (1 - Youden index), the AUC value and the Youden index can be combined to obtain a comprehensive performance evaluation index. This comprehensive index takes into account both the predictive accuracy of the classifier (AUC value) and the uncertainty of the classifier (Youden index).
[0108] The composite metric ranges from 0 to the AUC value, with a higher value indicating better classifier performance. For a classifier with low uncertainty, a higher AUC value will result in a higher composite metric. However, with high uncertainty, even a high AUC value may lead to a lower composite metric, indicating that the composite metric must consider not only the classifier's prediction accuracy but also its uncertainty.
[0109] like Figure 4 As shown, in some embodiments, the method for classifying the MI values in the first threshold set and the second threshold set and plotting ROC curves is as follows:
[0110] S410. Sort the model's prediction results from high to low probability.
[0111] S411. Select a threshold, and determine the samples whose prediction results are greater than or equal to the threshold as positive examples, and the samples whose prediction results are less than the threshold as negative examples.
[0112] S412. Calculate the true positive rate and false positive rate based on the current threshold;
[0113] S413. Select different thresholds, repeat steps S411 and S412, and record the true positive rate and false positive rate for each threshold.
[0114] S414. Plot all true positive rates and false positive rates on the coordinate axis to obtain the ROC curve.
[0115] The quality of the ROC curve directly reflects the classification performance of the model; the closer the curve is to the upper left corner, the better the model's performance.
[0116] S105. Determine whether the ECG signal to be detected is supraventricular tachycardia based on the comparison between the optimal discrimination threshold and the modulation index of the ECG signal to be detected.
[0117] In some embodiments, if the modulation index of the ECG signal to be detected is greater than or equal to the optimal discrimination threshold, the ECG signal to be detected is determined to be a supraventricular tachycardia signal; otherwise, it is a non-supraventricular tachycardia signal.
[0118] In summary, the proposed solution first acquires raw supraventricular tachycardia (SVT) and raw non-SVT signals, then preprocesses and bandpass filters them to obtain a second SVT signal and a second non-SVT signal in a fixed frequency band. Next, a phase-amplitude coupling algorithm is used to determine a first threshold set corresponding to the second SVT signal and a second threshold set corresponding to the second non-SVT signal. An optimal discrimination threshold is then determined based on the first and second threshold sets. Finally, the comparison between the optimal discrimination threshold and the modulation index of the ECG signal being detected determines whether the ECG signal being detected is supraventricular tachycardia. This detection method improves the accuracy of supraventricular tachycardia identification, reduces false positives and false negatives, and is a more efficient, safe, and simple detection method that is more easily accepted by patients.
[0119] In some embodiments, this application also proposes a classifier capable of distinguishing between supraventricular tachycardia (SVT) signals and non-SVT signals, used to determine whether the signal to be detected is an arrhythmia signal caused by atrial electrical signal disorder. The specific steps are as follows:
[0120] (1) Data collection: Collect electrocardiogram data including supraventricular tachycardia (SVT) signals and non-SVT signals. The above electrocardiogram data can come from electrocardiogram databases, research institutions or clinical trials.
[0121] (2) Feature extraction: Signal processing and machine learning algorithms are used to extract features that can represent supraventricular tachycardia and non-supraventricular tachycardia from the collected electrocardiogram data, including RR interval, heart rate, QRS waveform, ST segment, T wave and other information, as well as the MI value of the electrocardiogram signal.
[0122] (3) Feature selection: Select the most discriminative features (such as the MI value of the electrocardiogram signal). Feature selection can be performed by statistical methods or machine learning algorithms. The goal of feature selection is to minimize the number of features while retaining features that are effective for the classifier.
[0123] (4) Data partitioning: The collected data is divided into training set and test set by random partitioning and cross-validation for training and testing, and the training set is used to train the classifier and fine-tune the parameters.
[0124] (5) Classifier selection and training: Based on the features obtained from feature selection, a suitable classifier is selected for training; the classifiers include Support Vector Machine (SVM), Artificial Neural Network (ANN), Random Forest, etc.
[0125] (6) Model evaluation: The performance of the trained classifier is evaluated using the test set of evaluation metrics, including accuracy, recall, F1 score, etc.
[0126] (7) Optimization and improvement: Based on the results of model evaluation, the classifier can be optimized and improved. For example, the classifier can be optimized and improved by extracting different features, different categories of classifiers, or parameter combinations to improve the performance of the classifier.
[0127] (8) Validation and application: Apply the optimized classifier to new unknown data (data to be tested) to verify its generalization ability and practical application effect.
[0128] This application also proposes a supraventricular tachycardia detection system, which includes: an electrocardiogram signal acquisition unit, an electrocardiogram signal processing unit, a threshold set acquisition unit, an optimal discrimination threshold determination unit, and a detection unit;
[0129] The ECG signal acquisition unit is used to acquire raw supraventricular tachycardia signals, raw non-supraventricular tachycardia signals, and ECG signals to be detected.
[0130] The ECG signal processing unit is used to preprocess and bandpass filter the raw supraventricular tachycardia (SVT) signal and the raw non-SVT signal to obtain a second SVT signal and a second non-SVT signal in a fixed frequency band.
[0131] The threshold set acquisition unit is used to determine the first threshold set corresponding to the second supraventricular velocity signal and the second threshold set corresponding to the second non-supraventricular velocity signal according to the phase amplitude coupling algorithm.
[0132] The optimal discrimination threshold determination unit is used to determine the optimal discrimination threshold based on the first threshold set and the second threshold set.
[0133] The detection unit is used to determine whether the electrocardiogram signal is supraventricular tachycardia based on the comparison between the optimal discrimination threshold and the modulation index of the electrocardiogram signal.
[0134] This application also proposes a readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the following steps:
[0135] The raw supraventricular tachycardia (SVT) signal and the raw non-SVT signal are acquired and preprocessed to obtain the first SVT signal and the first non-SVT signal.
[0136] Bandpass filtering is performed on the first supraventricular tachycardia (SVT) signal and the first non-sVT signal to obtain a second SVT signal and a second non-SVT signal with a fixed frequency band.
[0137] The first threshold set corresponding to the second supraventricular velocities signal and the second threshold set corresponding to the second non-supraventricular velocities signal are determined based on the phase-amplitude coupling algorithm.
[0138] The optimal discrimination threshold is determined based on the first threshold set and the second threshold set;
[0139] The determination of whether the ECG signal to be detected is supraventricular tachycardia is based on the comparison between the optimal discrimination threshold and the modulation index of the ECG signal to be detected.
[0140] This application also proposes a computer device, including a memory and a processor, wherein the memory stores a computer program, and the computer program is executed by the processor in the following steps:
[0141] The raw supraventricular tachycardia (SVT) signal and the raw non-SVT signal are acquired and preprocessed to obtain the first SVT signal and the first non-SVT signal.
[0142] Bandpass filtering is performed on the first supraventricular tachycardia (SVT) signal and the first non-sVT signal to obtain a second SVT signal and a second non-SVT signal with a fixed frequency band.
[0143] The first threshold set corresponding to the second supraventricular velocities signal and the second threshold set corresponding to the second non-supraventricular velocities signal are determined based on the phase-amplitude coupling algorithm.
[0144] The optimal discrimination threshold is determined based on the first threshold set and the second threshold set;
[0145] The determination of whether the ECG signal to be detected is supraventricular tachycardia is based on the comparison between the optimal discrimination threshold and the modulation index of the ECG signal to be detected.
[0146] Those skilled in the art will understand that implementing all or part of the processes in the above embodiments can be accomplished by instructing related hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0147] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0148] The embodiments described above are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application's patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. The embodiments disclosed above are merely preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, equivalent variations made according to the claims of this invention are still within the scope of this invention.
Claims
1. A method for detecting supraventricular tachycardia, characterized in that, The method includes the following steps: The raw supraventricular tachycardia (SVT) signal and the raw non-SVT signal are acquired and preprocessed to obtain the first SVT signal and the first non-SVT signal. Bandpass filtering is performed on the first supraventricular tachycardia (SVT) signal and the first non-SVT signal to obtain a second SVT signal and a second non-SVT signal with a fixed frequency band. The first threshold set corresponding to the second supraventricular velocities signal and the second threshold set corresponding to the second non-supraventricular velocities signal are determined based on the phase-amplitude coupling algorithm. The phase-amplitude coupling algorithm specifically includes: normalizing the high-frequency amplitude in the low-frequency phase of the second supraventricular tachycardia (SVT) signal and the second non-SVT signal to determine a first amplitude; obtaining the Shannon entropy of the first amplitude; obtaining the Kullback-Leibler distance between the second SVT signal and the second non-SVT signal; obtaining a single second threshold MI value; traversing all second SVT signals and summarizing the obtained MI values into the first threshold set; traversing all second non-SVT signals and summarizing the obtained MI values into the second threshold set. The optimal discrimination threshold is determined based on the first threshold set and the second threshold set; The determination of whether the ECG signal to be detected is supraventricular tachycardia is based on the comparison between the optimal discrimination threshold and the modulation index of the ECG signal to be detected.
2. The method for detecting supraventricular tachycardia according to claim 1, characterized in that, The process of acquiring the original supraventricular tachycardia (SVT) signal and the original non-SVT signal, and preprocessing them to obtain the first SVT signal and the first non-SVT signal specifically includes: The original supraventricular tachycardia (SVT) signal and the original non-SVT signal are denoised, and the denoising process is then performed. The denoised original SVT signal and the original non-SVT signal are then normalized.
3. The method for detecting supraventricular tachycardia according to claim 2, characterized in that, The step of bandpass filtering the first supraventricular tachycardia (SVT) signal and the first non-SVT signal to obtain a second SVT signal and a second non-SVT signal with a fixed frequency band specifically includes: Bandpass filtering was performed on the first supraventricular tachycardia (SVT) signal and the first non-SVT signal, and frequency bands of 3 were extracted respectively. Hz ~9 Hz 10 Hz ~16 Hz The second ventricular tachycardia signal and frequency band are 3 Hz ~9 Hz 10 Hz ~16 Hz The second non-supracranial tachycardia signal.
4. The method for detecting supraventricular tachycardia according to claim 3, characterized in that, After performing bandpass filtering on the first supraventricular tachycardia (SVT) signal and the first non-SVT signal to obtain a second SVT signal and a second non-SVT signal with a fixed frequency band, the method further includes: performing windowing processing on the second SVT signal and the second non-SVT signal with the fixed frequency band.
5. The method for detecting supraventricular tachycardia according to claim 1, characterized in that, The phase-amplitude coupling algorithm normalizes the high-frequency amplitude in the low-frequency phase of the second supraventricular velocities (SNV) signal and the second non-SNV signal, with the specific expression being: in, For a single bin average amplitude, k for bins Operating index, N for bins The total number, p It is a N A vector of values; The Shannon entropy of the first amplitude is obtained by the following expression: in, p For each phase bin The vector of normalized average amplitude, N for bins The total number; Get Kullback-Leibler Distance, specifically expressed as: in, U To ensure uniform distribution, X For data distribution, N for bins The total number; Get MI The value, specifically expressed as: In the formula, KL ( U, X )for Kullback-Leibler distance, N for bins Total number, bins represents phase bins.
6. The method for detecting supraventricular tachycardia according to claim 1, characterized in that, The step of determining the optimal discrimination threshold based on the first threshold set and the second threshold set specifically includes: Constructed using the first threshold set and the second threshold set ROC curve; Get ROC In the curve AUC Value and Youden Index; Will AUC The value is combined with the Youden index, according to AU C The optimal discrimination threshold is determined by the value × (1 - Yuden index).
7. A supraventricular tachycardia detection system, characterized in that, The system includes: an electrocardiogram (ECG) signal acquisition unit, an ECG signal processing unit, a threshold set acquisition unit, an optimal discrimination threshold determination unit, and a detection unit; The ECG signal acquisition unit is used to acquire raw supraventricular tachycardia signals, raw non-supraventricular tachycardia signals, and ECG signals to be detected; The ECG signal processing unit is used to preprocess and bandpass filter the original supraventricular tachycardia signal and the original non-supraventricular tachycardia signal to obtain a second supraventricular tachycardia signal and a second non-supraventricular tachycardia signal in a fixed frequency band. The threshold set acquisition unit is used to determine a first threshold set corresponding to the second supraventricular tachycardia (SVT) signal and a second threshold set corresponding to the second non-SVT signal according to a phase-amplitude coupling algorithm. The phase-amplitude coupling algorithm specifically includes: normalizing the high-frequency amplitude in the low-frequency phase of the second SVT signal and the second non-SVT signal to determine a first amplitude; obtaining the Shannon entropy of the first amplitude; obtaining the Kullback-Leibler distance between the second SVT signal and the second non-SVT signal; obtaining a single second threshold MI value; traversing all second SVT signals and summarizing the obtained MI values into the first threshold set; traversing all second non-SVT signals and summarizing the obtained MI values into the second threshold set. The optimal distinction threshold determination unit is used to determine the optimal distinction threshold based on the first threshold set and the second threshold set. The detection unit is used to determine whether the electrocardiogram signal is supraventricular tachycardia based on the comparison between the optimal discrimination threshold and the modulation index of the electrocardiogram signal.
8. A readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of the method as claimed in any one of claims 1 to 6.
9. A computer device comprising a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 6.