Multi-module collaborative rapid electrocardiogram automatic diagnosis method and system

Abstract

The invention discloses a multi-module collaborative rapid electrocardiogram automatic diagnosis method and system, and belongs to the technical field of automatic diagnosis. The method comprises the following steps: preprocessing an original ECG signal; the preprocessed ECG signals serve as input, and after the preprocessed ECG signals are sequentially processed by a lightweight convolution sub-module, a deformation convolution sub-module and a non-local attention sub-module, multi-level features containing local details and global time sequence dependency are output; fusing the multi-level features, and sending the fused multi-level features into a classifier to complete prediction of heart rhythm types to obtain prediction probabilities of all categories; a sliding window mode is adopted, the ECG signal fragments with the fixed length are sent to the diagnosis model for reasoning, the diagnosis model continuously outputs diagnosis results, and alarm or diagnosis prompt information is output in time according to the diagnosis results. According to the method, by designing a lightweight network structure, the parameter quantity and calculation complexity of a diagnosis model are effectively reduced, and real-time online ECG signal automatic diagnosis is realized on the premise of ensuring high diagnosis accuracy.

Description

Technical Field

[0001] The present application proposes a multi-module collaborative rapid electrocardiogram automatic diagnosis method and system, which belongs to the field of automatic diagnosis technology. Background Art

[0002] ECG, as a non-invasive detection method for recording cardiac electrophysiological activities, has long been an important basis for clinical diagnosis of cardiovascular diseases. Traditional ECG analysis methods mainly rely on doctors' manual observation of waveform morphology and rich clinical experience to make judgments. This method not only relies on professional knowledge, but also has low efficiency and the results are easily affected by subjective factors. To solve this problem, researchers have begun to explore automated means to process ECG signals in recent years, striving to achieve objective, fast and high-precision diagnosis. In the signal preprocessing stage, commonly used methods include high-pass filtering, morphological filtering, wavelet transform, bandpass filtering and empirical mode decomposition. These methods can effectively eliminate slow baseline drift, suppress power frequency interference, myoelectric noise and other digital superposition noise, and highlight QRS waves, P waves, T waves and other signal components that are critical to diagnosis. After entering the feature extraction and classification stage, the researchers, on the one hand, reflect heart rate variability and basic waveform characteristics by detecting indicators such as R peak position, RR interval, and PT interval; on the other hand, they use support vector machine (SVM) and random forest machine learning methods to classify and identify manually extracted features. In recent years, deep neural networks have been introduced into the field of ECG signal analysis, achieving end-to-end automatic feature extraction and classification, significantly improving the accuracy and robustness of diagnosis. But at the same time, with the continuous improvement of application requirements such as telemedicine and real-time monitoring, the requirements for system real-time and lightweight are also getting higher and higher. Although existing Holter monitoring equipment can achieve long-term continuous monitoring, it often faces back-end delay problems when processing large-scale data; and due to the limitations of computing resources and power consumption, the online processing capabilities of smart wearable devices are still insufficient.

[0003] At present, ECG signal analysis schemes can be mainly divided into two categories: methods based on traditional signal processing and manual feature extraction, and automated methods based on machine learning and deep neural networks. Traditional methods generally use technologies such as high-pass filters and morphological filters to remove baseline drift, and locate QRS waves through time domain threshold detection and other methods, while combining expert-developed rules to preliminarily identify arrhythmias. This method has a clear theoretical basis and low implementation cost, but it is not adaptable enough when dealing with complex noise interference and diversified waveforms, and it is difficult to meet the clinical needs for high-accuracy diagnosis. Another type of solution is to further reduce noise based on preprocessing, and then extract features such as R peak position, PR interval, QT interval, heart rate variability, and use SVM or random forest for classification. This type of method has achieved good results on some public data sets, but due to its reliance on cumbersome manual feature engineering, its robustness to noise and individual differences is limited. In recent years, with the continuous development of deep learning technology, researchers have applied diagnostic models such as convolutional neural networks (CNN) and long short-term memory networks (LSTM) to ECG signal analysis, and further explored the use of Transformer diagnostic models to globally model signal fragments. Although these methods have made significant progress in improving diagnostic accuracy and anti-interference capabilities, they are usually accompanied by large parameter scales and high computational complexity, making real-time deployment on wearable devices or edge computing platforms difficult.

[0004] In clinical practice, electrocardiogram (ECG) is a non-invasive detection method for recording cardiac electrophysiological activity and is widely used in the diagnosis of cardiovascular diseases. However, in the actual acquisition process, due to the influence of multiple factors such as baseline drift, electromyographic interference, power frequency noise and poor electrode contact, ECG signals often have problems such as noise mixing and waveform distortion, which makes it difficult to accurately identify key features (such as P waves, QRS complexes, T waves, etc.), which may cause misdiagnosis or missed diagnosis of arrhythmias, myocardial ischemia and other diseases. In addition, existing ECG diagnostic systems often rely on complex deep network structures or cumbersome manual feature engineering. The diagnostic model parameters are huge, the amount of calculation is large, and the real-time performance is poor, which is not suitable for deployment on wearable devices or edge computing platforms.

[0005] Traditional filtering methods have the problem of insufficient noise adaptability when facing low-frequency baseline drift, mixed multi-type noise or extreme interference, and are prone to filtering or under-filtering, resulting in the loss of key ECG features or failure to effectively suppress noise. In addition, many machine learning-based methods rely too much on manually setting thresholds or extracting time domain, frequency domain, and morphological features, and therefore have poor adaptability to different types of arrhythmias and low versatility. The reliance on manual feature engineering also increases the difficulty of algorithm updating and maintenance. At the same time, although solutions based on deep convolutional networks or Transformers have greatly improved accuracy, their network structures are complex and the number of parameters is large. They cannot run in real time in resource-constrained hardware environments, and the huge data requirements and training costs also limit the large-scale promotion of these diagnostic models in actual clinical practice. In addition, although some methods perform well in offline mode, they lack the ability to be quickly deployed and updated online when performing real-time inference and immediate alarm in remote medical monitoring or emergency. Summary of the invention

[0006] In response to the above problems, this application proposes a lightweight and fast electrocardiogram diagnosis method that takes into account multiple noise interference correction and automatic feature extraction. This method effectively reduces the number of diagnostic model parameters and computational complexity by designing a lightweight network structure, and realizes real-time online ECG signal automatic diagnosis while ensuring high diagnostic accuracy.

[0007] In order to solve the above technical problems, the technical solution adopted in this application is: In a first aspect, the present application provides a multi-module collaborative rapid electrocardiogram automatic diagnosis system, comprising: A signal preprocessing module, used for preprocessing the original ECG signal; A multi-level feature extraction module, including a lightweight convolution submodule, a deformable convolution submodule and a non-local attention submodule; used to take the preprocessed ECG signal as input, and after being processed by the lightweight convolution submodule, the deformable convolution submodule and the non-local attention submodule in sequence, output a multi-level feature containing local details and global temporal dependencies; The feature fusion and classification module is used to fuse multi-level features and send them to the classifier to complete the prediction of heart rhythm type and obtain the prediction probability of each category; The online reasoning and real-time monitoring module is used to send ECG signal segments of fixed length to the diagnosis model for reasoning using a sliding window method, so that the diagnosis model can continuously output diagnosis results and output alarms or diagnosis prompt information in a timely manner according to the diagnosis results.

[0008] As a further improvement of the present invention, the signal preprocessing module is specifically used to: perform noise reduction, filtering and amplitude normalization processing on the original ECG signal, The filtering adopts an adaptive filter to filter out low-frequency baseline drift, a band-stop filter to suppress power frequency interference, and an empirical mode decomposition method to eliminate myoelectric interference; The amplitude normalization process obtains a signal including a P wave, a QRS complex and a T wave.

[0009] As a further improvement of the present invention, the lightweight convolution submodule uses depthwise separable convolution to realize feature extraction, and the specific method is as follows:

[0010] Where x(t) represents the preprocessed ECG signal of the input, W is the weight parameter of the lightweight convolution kernel, b is the bias term, BN represents the batch normalization operation, ReLU is the activation function, and y(t) represents the final output feature sequence of the lightweight convolution submodule, which can be subsequently input into the deformable convolution submodule.

[0011] As a further improvement of the present invention, the deformable convolution submodule adopts a method of adaptively adjusting convolution sampling points, and the specific method is:

[0012] Where t represents the time index of the current convolution center point, w i is the weight parameter of the convolution kernel at index i, p i is a fixed offset, Δp i is the learnable deformation offset.

[0013] As a further improvement of the present invention, the non-local attention submodule is used to capture the long-distance dependencies across cycles or abnormal bands in the ECG signal, construct a global correlation matrix for each sampling point in the sequence, and obtain the features of the diagnostic time series segment through a re-weighted operation. The specific method is as follows:

[0014] In the formula, x i 、x j denotes the feature vector of the ECG input at index i and j, respectively, f(x i ,x j ) is a similarity function used to measure x i With x j The degree of correlation between them, g(x j ) is the feature transformation function, x j Linear mapping is performed to extract fusible information, and C(x) is the normalization factor.

[0015] As a further improvement of the present invention, the feature fusion and classification module is specifically used to use point convolution to merge and compress multi-level features in the channel dimension, retain low-frequency and high-frequency features, and fuse them in a unified dimension; then introduce a residual skip connection mechanism to reuse early shallow features of multi-level features of unified dimension for feature fusion; the final fused features enter the classification layer through a lightweight classification head, and use the Softmax function to normalize the output scores to obtain the prediction probability of each category.

[0016] As a further improvement of the present invention, the diagnosis model is compressed into a half-precision floating point number or an integer number for real-time online diagnosis on a CPU or an edge device.

[0017] In a second aspect, the present application provides a multi-module collaborative rapid electrocardiogram automatic diagnosis method, comprising: Preprocessing the raw ECG signal; The preprocessed ECG signal is used as input, and after being processed by the lightweight convolution submodule, the deformable convolution submodule, and the non-local attention submodule in sequence, the multi-level features containing local details and global temporal dependencies are output; The multi-level features are fused and sent to the classifier to complete the prediction of the heart rhythm type and obtain the prediction probability of each category; Using a sliding window method, ECG signal segments of fixed length are sent to the diagnostic model for inference, so that the diagnostic model can continuously output diagnostic results and output alarms or diagnostic prompt information in a timely manner based on the diagnostic results.

[0018] In a third aspect, the present application provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the multi-module collaborative rapid electrocardiogram automatic diagnosis method when executing the computer program.

[0019] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the multi-module collaborative rapid electrocardiogram automatic diagnosis method is implemented.

[0020] In a fifth aspect, the present application provides a computer program product, which includes computer instructions, and the computer instructions instruct a computer to execute the multi-module collaborative rapid electrocardiogram automatic diagnosis method.

[0021] The beneficial effects of this application compared to the prior art are: The system of the present application takes into account the high robustness and real-time online diagnostic accuracy in complex noise environments. First, by designing a preprocessing module combining adaptive filtering and deep networks, the low-frequency baseline drift, myoelectric interference and power frequency noise in the ECG signal are effectively suppressed, while ensuring the complete retention of key features such as P wave, QRS complex and T wave. Secondly, a multi-level feature extraction architecture is constructed, which adopts the organic fusion of lightweight convolutional network, deformation convolution and non-local attention mechanism, which can not only capture the local details of ECG signals, but also learn global timing dependencies, realize end-to-end automatic feature extraction, and avoid the limitations of artificial feature engineering. Thirdly, through module cutting, weight sharing and the combination of attention mechanism and convolution operation, the parameters and computational complexity of the diagnostic model are greatly reduced, thereby meeting the requirements of wearable devices and edge platforms for real-time reasoning. Compared with traditional methods, this technical solution shows higher sensitivity and specificity in arrhythmia detection while ensuring high diagnostic accuracy, effectively reducing the risk of missed diagnosis and misdiagnosis, and is suitable for multiple scenarios such as telemedicine and bedside monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following introduction is made to the drawings of the embodiments of the present application or the related technical solutions in the prior art. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0023] Figure 1 A block diagram of a multi-module collaborative rapid electrocardiogram automatic diagnosis system provided in this application; Figure 2 Schematic diagram of the multi-level feature extraction module. DETAILED DESCRIPTION

[0024] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and are not to be construed as limitations on the present application. For the step numbers in the following embodiments, they are only provided for the convenience of explanation, and the order between the steps is not limited in any way, and the execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0025] In the description of this application, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in this application based on the specific content of the technical solution.

[0026] This application combines adaptive filtering with lightweight deep networks, such as Figure 1 As shown, a multi-module collaborative rapid electrocardiogram automatic diagnosis system is proposed, including a signal preprocessing module, a feature extraction module, a feature fusion and classification module, and an online reasoning and real-time monitoring module.

[0027] A signal preprocessing module, used for preprocessing the original ECG signal; A multi-level feature extraction module, including a lightweight convolution submodule, a deformable convolution submodule and a non-local attention submodule; used to take the preprocessed ECG signal as input, and after being processed by the lightweight convolution submodule, the deformable convolution submodule and the non-local attention submodule in sequence, output a multi-level feature containing local details and global temporal dependencies; The feature fusion and classification module is used to fuse multi-level features and send them to the classifier to complete the prediction of heart rhythm type and obtain the prediction probability of each category; The online reasoning and real-time monitoring module is used to send ECG signal segments of fixed length to the diagnosis model for reasoning using a sliding window method, so that the diagnosis model can continuously output diagnosis results and output alarms or diagnosis prompt information in a timely manner according to the diagnosis results.

[0028] The system can achieve multi-module collaboration. Under the premise of low parameter quantity and computing consumption, it can still realize high-precision automatic ECG diagnosis and achieve real-time online reasoning. It can run efficiently on ordinary computers, wearable devices or edge computing platforms, and meet the application needs of multiple scenarios such as telemedicine and bedside monitoring.

[0029] The lightweight and rapid electrocardiogram automatic diagnosis method designed in this application is based on four major modules: signal preprocessing module, multi-level feature extraction module, feature fusion and classification module, and online reasoning and real-time monitoring module. The specific process of the automatic diagnosis method is as follows: First, the original ECG signal is subjected to noise reduction, filtering and amplitude normalization through the signal preprocessing module, thereby effectively eliminating baseline drift, electromyographic interference and power frequency noise, ensuring that key signal components such as P wave, QRS complex and T wave are fully preserved; Secondly, the preprocessed ECG signal is sent to the multi-level feature extraction module, which uses a lightweight convolutional neural network combined with deformation convolution and non-local attention mechanism to construct local detail features and global temporal dependencies respectively, and realize the automatic extraction of multi-level features of the signal; Next, the feature fusion and classification module effectively integrates the features extracted at each level, reduces the dimension of the features through 1×1 convolution, and uses the attention weighting mechanism to highlight the key bands, and finally completes the accurate classification of arrhythmia types; Finally, the online reasoning and real-time monitoring module designs a dedicated reasoning pipeline for continuous streaming ECG signals, adopts a sliding window strategy and diagnostic model compression and quantization technology to realize real-time online diagnosis of the system on a general-purpose CPU or edge device, and outputs alarm or diagnostic prompt information in a timely manner.

[0030] Each module is described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0031] In the signal preprocessing module, this application first uses an adaptive filter to filter out the common low-frequency baseline drift problem in ECG signals to ensure that low-frequency components such as P waves and T waves can be fully retained. In order to suppress power frequency interference, a band-stop filter is set. At the same time, for electromyographic interference, the empirical mode decomposition (EMD) technology is combined to further reduce high-frequency noise. After the above filtering process, the signal is amplitude normalized to balance the voltage range differences between different measuring devices and different patients, providing a unified scale for the subsequent neural network input.

[0032] like Figure 2 As shown in the figure, the structure diagram of the multi-level feature extraction module. Combining deformable convolution with attention mechanism, based on lightweight convolution, it takes into account both local and global information extraction.

[0033] Figure 2 In the proposed method, the preprocessed ECG sequence is first input into the lightweight convolution module, and the deep convolution and point convolution are used to efficiently extract features from each channel. Then, the deformable convolution module is used to adaptively sample the key bands on the time axis using the learnable offset. Then, the non-local attention module is used to capture the long-distance dependencies across cycles and abnormal bands based on global similarity calculation and weighted summation. Finally, the output is a multi-level feature representation, which provides refined feature support for subsequent classification.

[0034] This module receives the preprocessed ECG sequence x(t) as input, and after being processed by the lightweight convolution submodule, the deformable convolution submodule, and the non-local attention submodule in sequence, it finally outputs a multi-level feature representation containing local details and global temporal dependencies, providing effective support for subsequent feature fusion and classification. First, in the lightweight convolution submodule, in order to reduce the number of parameters and computational complexity, this application draws on the design of lightweight networks such as MobileNet and uses deep separable convolution to achieve feature extraction. This module can be described by the following formula:

[0035] Where x(t) represents the preprocessed ECG signal of the input, W is the weight parameter of the lightweight convolution kernel, b is the bias term, BN represents the batch normalization operation, ReLU is the activation function, and y(t) represents the final output feature sequence of the lightweight convolution submodule, which can be subsequently input into the deformable convolution submodule.

[0036] Next, the deformable convolution submodule proposes a method to adaptively adjust the convolution sampling points based on the morphological differences of ECG signals in different patients or physiological states. This module learns a set of offsets to enable the conventional convolution kernel to be fine-tuned in the time domain, thereby more flexibly capturing the features of key locations. Its calculation formula can be expressed as:

[0037] Where t represents the time index of the current convolution center point, w i is the weight parameter of the convolution kernel at index i, p i is a fixed offset, Δp i It is a learnable deformation offset, which is used to make fine adjustments based on the original fixed offset pi, so as to adapt to different waveform characteristics and individual differences.

[0038] Finally, the non-local attention submodule is used to capture the long-distance dependencies across cycles or abnormal bands in the ECG signal. This module builds a global correlation matrix for each sampling point in the sequence, and highlights the features of the most critical time segment for diagnosis through re-weighting operations. The formula is described as follows:

[0039] In the formula, x i 、x j denotes the feature vector of the ECG input at index i and j, respectively, f(x i ,x j ) is a similarity function used to measure x i With x j The degree of correlation between them, g(x j ) is the feature transformation function, x j Linear mapping is performed to extract fusible information, and C(x) is the normalization factor.

[0040] Through this non-local operation, global information can be effectively integrated and the diagnostic model's ability to capture ECG periodicity and abnormal states can be enhanced.

[0041] In the feature fusion and classification module, after the multi-level features are extracted, these features need to be fused and sent to the classifier to complete the prediction of the heart rhythm type. First, point convolution is used to merge and compress the features output by each submodule in the channel dimension, so as to retain both low-frequency and high-frequency features and unify the dimensions of the output of each module to ensure full fusion of information. In addition, in order to prevent information loss due to the deepening of the network layers, a residual skip connection mechanism is introduced to reuse early shallow features to enhance the overall feature representation capability. After the feature fusion is completed, the final fused features will enter the classification layer through the lightweight classification head, and the output scores will be normalized using the Softmax function to obtain the prediction probability of each category.

[0042] In the online reasoning and real-time monitoring module, first of all, for clinical or remote monitoring application scenarios, since ECG signals are often transmitted in the form of streaming data, a sliding window method is used to input ECG segments of fixed length into the diagnosis model for reasoning, and continuously output diagnosis results.

[0043] Preferably, in order to meet the actual deployment requirements, the diagnostic model can be further compressed or quantized to FP16 or even INT8 precision, so as to achieve fast and efficient operation on CPU, GPU or dedicated AI acceleration chip. When abnormal conditions such as atrial fibrillation and ventricular premature beats are detected, the system will quickly send an alarm signal to the monitor or smart terminal, and upload the detailed diagnostic results to the cloud or doctor's workstation for further decision support.

[0044] As an example, this application proposes to significantly improve the running speed and energy efficiency of the model on a variety of hardware platforms by compressing the model from the original high precision (such as FP32) to half-precision floating point numbers (FP16) or even integer numbers (INT8) without significantly sacrificing the model accuracy. The diagnostic model of this application effectively reduces the storage requirements and computational complexity of the model by flexibly selecting FP16 or INT8 precision, and improves the running efficiency and compatibility of the model on a variety of hardware platforms. At the same time, through sophisticated quantization-aware training and hardware acceleration optimization strategies, it is ensured that the quantization model achieves significant performance improvements while maintaining high accuracy, providing strong technical support for the actual deployment of the diagnostic model.

[0045] In summary, this application proposes a multi-level feature representation module that combines deformable convolution and non-local attention mechanism to perform multi-level feature extraction on ECG signals. Specifically, deformable convolution is used to adaptively capture the offset of key positions of the waveform, thereby constructing a local detail representation; at the same time, the non-local attention mechanism is used to calculate the long-distance dependencies between positions in the signal to form global timing information. This multi-level representation ensures the precise identification of various types of arrhythmias and improves the accuracy of diagnosis.

[0046] The lightweight convolution backbone network structure that constitutes the core of multi-level feature extraction in this application adopts deep separable convolution technology and lightweight network design, which significantly reduces the number of diagnostic model parameters and computational complexity, thereby achieving real-time online reasoning. This design optimizes the utilization of hardware resources, enabling the system to run stably on wearable devices and edge computing platforms, meeting the needs of real-time application scenarios such as telemedicine and bedside monitoring.

[0047] A residual skip connection mechanism and a feature fusion module were introduced. Based on the feature reuse and fusion strategy, this strategy not only effectively alleviated the problem of gradient vanishing in deep networks, but also enhanced the overall feature expression capability by fusing the redundant information extracted at each level, thereby further improving the classification and discrimination performance of various arrhythmias.

[0048] Through testing, this application has obvious advantages in noise robustness. It combines adaptive filtering and deformable convolution technology, which can effectively deal with baseline drift and various noise interferences, while suppressing noise while ensuring that key ECG features are not over-filtered. Compared with pure Transformer or large deep convolutional networks, this application adopts a lightweight convolution design, which greatly reduces the parameter scale of the convolution kernel, thereby achieving real-time operation on general-purpose CPUs or mobile devices, meeting the needs of telemedicine and bedside monitoring. In addition, this application realizes end-to-end automatic feature extraction, avoiding reliance on cumbersome manual threshold settings and manual feature engineering, so that the system can identify multiple types of arrhythmias and has stronger versatility and scalability.

[0049] Furthermore, the present application was tested on the public data set PTB-XL arrhythmia database. Experimental results show that in multi-category detection tasks such as atrial fibrillation and ventricular premature beats, the accuracy of this method can reach 98.67%, and it is higher than most existing comparison algorithms in terms of sensitivity and specificity. In the simulation experiment of adding baseline drift and power frequency interference noise, the present application can still maintain a high classification accuracy, and also shows a more stable anti-interference ability than the existing methods under extreme interference conditions.

[0050] Furthermore, the present application can adapt to different application requirements through module replacement, diagnostic model optimization and expansion. The deformable convolution submodule can be replaced by other deformable or dynamic convolution structures, and the non-local attention mechanism can be replaced by the Transformer encoder to achieve global dependency modeling. At the same time, the filtering method of the preprocessing module can be adjusted according to needs. In order to reduce the scale of the diagnostic model and the inference delay, knowledge distillation or low-bit quantization can be used to improve deployment efficiency while maintaining accuracy.

[0051] The third object of the embodiment of the present application is to provide an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned multi-module collaborative rapid electrocardiogram automatic diagnosis method when executing the computer program. It also includes a communication interface and a bus.

[0052] The fourth objective of an embodiment of the present application is to provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned multi-module collaborative rapid electrocardiogram automatic diagnosis method is implemented.

[0053] The fifth objective of the embodiment of the present application is to provide a computer program product, which includes computer instructions, and the computer instructions instruct the computer to execute the above-mentioned multi-module collaborative rapid electrocardiogram automatic diagnosis method.

[0054] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0055] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0056] The present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, readable storage medium, optical storage, etc.) containing computer-usable program codes.

[0057] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0058] Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present application.

[0059] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present application rather than to limit it. Although the present application has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present application can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present application should be included in the scope of protection of the claims of the present application.

Claims

1. A multi-module collaborative rapid electrocardiogram automatic diagnosis system, characterized in that: include: A signal preprocessing module, used for preprocessing the original ECG signal; Multi-level feature extraction module, including lightweight convolution submodule, deformable convolution submodule and non-local attention submodule; It is used to take the preprocessed ECG signal as input, and after being processed by the lightweight convolution submodule, the deformation convolution submodule and the non-local attention submodule in sequence, it outputs multi-level features containing local details and global temporal dependencies; The feature fusion and classification module is used to fuse multi-level features and send them to the classifier to complete the prediction of heart rhythm type and obtain the prediction probability of each category; The online reasoning and real-time monitoring module is used to send ECG signal segments of fixed length to the diagnosis model for reasoning using a sliding window method, so that the diagnosis model can continuously output diagnosis results and output alarms or diagnosis prompt information in a timely manner according to the diagnosis results.

2. A multi-module collaborative rapid electrocardiogram automatic diagnosis system according to claim 1, characterized in that: The signal preprocessing module is specifically used to: perform noise reduction, filtering and amplitude normalization processing on the original ECG signal, The filtering adopts an adaptive filter to filter out low-frequency baseline drift, a band-stop filter to suppress power frequency interference, and an empirical mode decomposition method to eliminate myoelectric interference; The amplitude normalization process obtains a signal including a P wave, a QRS complex and a T wave.

3. The multi-module collaborative rapid electrocardiogram automatic diagnosis system according to claim 1, characterized in that: The lightweight convolution submodule uses depth-separable convolution to achieve feature extraction. The specific method is as follows: Where x(t) represents the preprocessed ECG signal of the input, W is the weight parameter of the lightweight convolution kernel, b is the bias term, BN represents the batch normalization operation, ReLU is the activation function, and y(t) represents the final output feature sequence of the lightweight convolution submodule, which can be subsequently input into the deformable convolution submodule.

4. The multi-module collaborative rapid electrocardiogram automatic diagnosis system according to claim 1, characterized in that: The deformation convolution submodule adopts a method of adaptively adjusting the convolution sampling points, and the specific method is as follows: Where t represents the time index of the current convolution center point, w i is the weight parameter of the convolution kernel at index i, p i is a fixed offset, Δp i is the learnable deformation offset.

5. The multi-module collaborative rapid electrocardiogram automatic diagnosis system according to claim 1, characterized in that: The non-local attention submodule is used to capture the long-distance dependencies across cycles or abnormal bands in the ECG signal, build a global correlation matrix for each sampling point in the sequence, and obtain the features of the diagnostic time series segment through re-weighting operations. The specific method is as follows: In the formula, x i 、x j denotes the feature vector of the ECG input at index i and j, respectively, f(x i ,x j ) is a similarity function used to measure x i With x j The correlation between g(x j ) is the feature transformation function, x j Linear mapping is performed to extract fusible information, and C(x) is the normalization factor.

6. The multi-module collaborative rapid electrocardiogram automatic diagnosis system according to claim 1, characterized in that: The feature fusion and classification module is specifically used to use point convolution to merge and compress multi-level features in the channel dimension, retain low-frequency and high-frequency features, and fuse them in a unified dimension; then introduce a residual skip connection mechanism to reuse early shallow features for multi-level features of unified dimension to perform feature fusion; the final fused features enter the classification layer through a lightweight classification head, and the output scores are normalized using the Softmax function to obtain the prediction probability of each category.

7. The multi-module collaborative rapid electrocardiogram automatic diagnosis system according to claim 1, characterized in that: The diagnostic model is compressed into half-precision floating point numbers or integers for real-time online diagnosis on the CPU or edge devices.

8. A multi-module collaborative rapid electrocardiogram automatic diagnosis method, characterized in that: include: Preprocessing the raw ECG signal; The preprocessed ECG signal is used as input, and after being processed by the lightweight convolution submodule, the deformable convolution submodule, and the non-local attention submodule in sequence, the multi-level features containing local details and global temporal dependencies are output; The multi-level features are fused and sent to the classifier to complete the prediction of the heart rhythm type and obtain the prediction probability of each category; Using a sliding window method, ECG signal segments of fixed length are sent to the diagnostic model for inference, so that the diagnostic model can continuously output diagnostic results and output alarms or diagnostic prompt information in a timely manner based on the diagnostic results.

9. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the multi-module collaborative rapid electrocardiogram automatic diagnosis method as claimed in claim 8 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the multi-module collaborative rapid electrocardiogram automatic diagnosis method of claim 8 is implemented.

Images