Electrocardiosignal processing method, electrocardiograph, electronic equipment and storage medium

Abstract

The invention relates to the technical field of signal processing, and discloses an electrocardiosignal processing method, an electrocardiograph, electronic equipment and a storage medium, the method comprises the following steps: acquiring an original electrocardiosignal collected in an exercise load test process; performing denoising processing on the original electrocardiosignal to generate a denoised electrocardiosignal; and responding to a display starting instruction triggered by a user, and contrasting and displaying the original electrocardiosignal and the de-noised electrocardiosignal. The de-noised electrocardiosignals are displayed to the user, the user is assisted in effectively recognizing electrocardiograph features in the electrocardiosignals, and subsequent accurate analysis is facilitated; moreover, the user can observe the waveform form of the electrocardiosignal before and after denoising in a contrast display mode, and can autonomously select whether to carry out subsequent analysis and diagnosis based on the original electrocardiosignal or the denoised electrocardiosignal after denoising, so that the accuracy of analyzing the exercise electrocardiosignal is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of signal processing, and particularly to a method for processing electrocardiogram signals, an electrocardiograph, an electronic device, and a storage medium. Background Art

[0002] The electrocardiogram exercise test, also known as the electrocardiogram exercise stress test, is a test method that increases the heart load through a certain amount of exercise and observes the electrocardiogram changes to determine the heart health status.

[0003] Since the subject is not stationary during the electrocardiogram exercise stress test but remains in a moving state during the test, noise interference is likely to exist in the collected exercise electrocardiogram signals. The noise interference in the exercise electrocardiogram signals makes it difficult to effectively read the electrocardiogram, and the noise interference in the exercise electrocardiogram signals greatly affects the accuracy of the electrocardiogram analysis system for electrocardiogram feature detection and parameter calculation. Summary of the Invention

[0004] In view of this, the present invention provides a method for processing electrocardiogram signals, an electrocardiograph, an electronic device, and a storage medium to reduce the noise interference of electrocardiogram signals.

[0005] In a first aspect, the present invention provides a method for processing electrocardiogram signals, including:

[0006] Obtaining the original electrocardiogram signals collected during the exercise stress test;

[0007] Performing denoising processing on the original electrocardiogram signals to generate denoised electrocardiogram signals;

[0008] In response to the display start instruction triggered by the user, displaying the original electrocardiogram signals and the denoised electrocardiogram signals in comparison.

[0009] In some optional embodiments, the performing denoising processing on the original electrocardiogram signals to generate denoised electrocardiogram signals includes:

[0010] Inputting the original electrocardiogram signals within a preset time period into a denoising model at each interval of the preset time period; the denoising model is an encoder-decoder network model for outputting a sequence of the same length as the input signal;

[0011] Performing encoder-decoder processing on the original electrocardiogram signals within the preset time period according to the denoising model to generate the denoised electrocardiogram signals within the preset time period.

[0012] In some alternative embodiments, the denoising model includes an encoder, a decoder, and an output layer; the encoder includes n encoding units, and the decoder includes n decoding units; the encoding unit includes a first convolutional part and a pooling layer; the decoding unit includes an upsampling layer and a second convolutional part; both the first convolutional part and the second convolutional part include at least one convolutional layer;

[0013] The first convolutional part of the i-th encoding unit is configured to perform convolutional processing on the input data and output it to the pooling layer of the i-th encoding unit;

[0014] The pooling layer of the i-th encoding unit is configured to perform pooling processing on the input data to generate pooled data;

[0015] Wherein, the input data of the first convolutional part of the first encoding unit is the original electrocardiogram signal within the preset time period, and the input data of the first convolutional part of the remaining encoding units is the pooled data generated by the previous encoding unit;

[0016] The upsampling layer of the j-th decoding unit is configured to perform upsampling processing on the output data of the (j - 1)-th decoding unit to generate upsampled data, and fuse the upsampled data with the pooled data of the same size output by one of the encoding units, and input the fused data into the second convolutional part of the j-th decoding unit;

[0017] The second convolutional part of the j-th decoding unit is configured to perform convolutional processing on the input data and output;

[0018] The output layer is configured to process the output data of the last decoding unit based on an activation function and output the denoised electrocardiogram signal within the preset time period.

[0019] In some alternative embodiments, the method further includes:

[0020] When displaying the original electrocardiogram signal, delaying the display of the original electrocardiogram signal, and the delay duration matches the preset time period.

[0021] In some alternative embodiments, the step of contrastively displaying the original electrocardiogram signal and the denoised electrocardiogram signal includes:

[0022] Stopping updating the display of the original electrocardiogram signal;

[0023] Replacing the denoised electrocardiogram signal with the original electrocardiogram signal and displaying the denoised electrocardiogram signal in at least a partial display area for displaying the original electrocardiogram signal.

[0024] In some alternative embodiments, after acquiring the original electrocardiogram signal collected during the exercise stress test, the method further includes:

[0025] Analyze the original electrocardiogram (ECG) signal within the current time period to determine the quality level of the original ECG signal;

[0026] Display the quality level in real time on the interface.

[0027] In some alternative embodiments, the method further includes:

[0028] Detect the type of characteristic wave to which the signal points in the target ECG signal belong; the target ECG signal is the original ECG signal or the denoised ECG signal;

[0029] When the target ECG signal is being displayed, synchronously mark the type of characteristic wave to which the signal points in the target ECG signal belong.

[0030] In some alternative embodiments, the synchronously marking the type of characteristic wave to which the signal points in the target ECG signal belong includes:

[0031] Mark signal segments of different characteristic wave types with different colors; the signal segments include multiple signal points.

[0032] In some alternative embodiments, the method further includes: in response to a close display instruction triggered by the user, stop displaying the denoised ECG signal.

[0033] In a second aspect, the present invention provides an electrocardiograph, including: a processing device and a display device;

[0034] The processing device is configured to:

[0035] Obtain the original ECG signal collected during the exercise stress test;

[0036] Perform denoising processing on the original ECG signal to generate a denoised ECG signal;

[0037] In response to an open display instruction triggered by the user, control the display device to display the original ECG signal and the denoised ECG signal in a side-by-side manner.

[0038] In a third aspect, the present invention provides an electronic device, including: a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to perform the method for processing an ECG signal according to the first aspect or any corresponding embodiment thereof.

[0039] In a fourth aspect, the present invention provides a computer-readable storage medium, on which computer instructions are stored. The computer instructions are used to cause a computer to execute the method for processing an ECG signal according to the first aspect or any corresponding embodiment thereof.

[0040] The present invention performs denoising processing on the original electrocardiogram (ECG) signals collected during an exercise stress test, and can generate denoised ECG signals with less noise interference. Moreover, the user can input a display start instruction based on their own needs, so that the original ECG signals and the denoised ECG signals can be displayed for comparison. By displaying the denoised ECG signals to the user, the noise interference in the original ECG signals can be reduced, assisting the user to effectively identify the ECG features therein, which is convenient for subsequent accurate analysis. Furthermore, the method of comparative display enables the user to observe the waveform patterns of the ECG signals before and after denoising, and the user can independently choose to perform subsequent analysis and diagnosis based on the original ECG signals or the denoised ECG signals, which is beneficial to improving the accuracy of the analysis of exercise ECG signals.

[0041] Moreover, the ECG features therein can be more accurately extracted based on the denoising model, and precise denoising can be achieved. Also, the pooled data of the encoding unit can be fused with the upsampled data of the decoding unit at the corresponding stage. Through this skip connection architecture, the decoding unit is allowed to learn the relevant features lost in the pooling of the encoding unit at each stage. Therefore, the convolutional layer can remove the noise of the signal more precisely and can generate more accurate denoised ECG signals. Automatically marking the characteristic waves of the ECG signals realizes the automatic positioning of the characteristic waves of each heartbeat in the exercise electrocardiogram and achieves visual identification. Marking the characteristic waves of the exercise electrocardiogram on the electrocardiogram can assist the user in quickly finding the positions of the characteristic waves, thereby improving the user's efficiency in reading electrocardiograms. Marking with different colors can conveniently mark all signal points, and can visually display the positions of each characteristic wave to the user, with good visualization effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the related art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the related art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0043] Figure 1 is a schematic flowchart of a method for processing ECG signals according to an embodiment of the present invention;

[0044] Figure 2 is a schematic flowchart of another method for processing ECG signals according to an embodiment of the present invention;

[0045] Figure 3 is a schematic structural diagram of a denoising model according to an embodiment of the present invention;

[0046] Figure 4It is a schematic flowchart of a method for processing another electrocardiogram (ECG) signal according to an embodiment of the present invention;

[0047] Figure 5 It is a schematic diagram of the process for processing an ECG signal according to an embodiment of the present invention;

[0048] Figure 6 It is a structural block diagram of an apparatus for processing an ECG signal according to an embodiment of the present invention;

[0049] Figure 7 It is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed implementation manners

[0050] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0051] According to an embodiment of the present invention, an embodiment of a method for processing an ECG signal is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0052] In this embodiment, a method for processing an ECG signal is provided, which can be used in devices capable of collecting and displaying ECG signals, such as electrocardiographs and the like. Figure 1 It is a flowchart of a method for processing an ECG signal according to an embodiment of the present invention. As Figure 1 shown, the process includes the following steps.

[0053] Step S101, obtain the original ECG signal collected during the exercise stress test.

[0054] In this embodiment, during the exercise stress test of the subject, the ECG signal of the subject during exercise can be collected through leads and the like, that is, the exercise ECG signal. For convenience of description, the collected exercise ECG signal is referred to as the original ECG signal. Among them, since there is likely to be noise interference in the collected exercise ECG signal at this time, that is, there is a high probability of noise interference in the original ECG signal.

[0055] The treadmill exercise test is currently the most widely used exercise stress test method, and signal acquisition can be performed using the treadmill exercise test. For example, the subject can walk or run on a moving treadmill. According to the selected exercise protocol, the treadmill speed and slope can be set to adjust the load until the subject's heart rate reaches the submaximal level, and on this basis, electrocardiogram signals can be collected to obtain the corresponding original electrocardiogram signals.

[0056] Step S102: Denoise the original electrocardiogram signals to generate denoised electrocardiogram signals.

[0057] In this embodiment, after obtaining the original electrocardiogram signals, the original electrocardiogram signals can be denoised to remove the noise therein to generate exercise electrocardiogram signals with less noise interference, that is, denoised electrocardiogram signals. For example, the noise in the original electrocardiogram signals can be filtered based on signal filtering to generate corresponding denoised electrocardiogram signals.

[0058] Step S103: In response to the display start instruction triggered by the user, display the original electrocardiogram signals and the denoised electrocardiogram signals in a comparative manner.

[0059] In this embodiment, during the exercise stress test, an interactive function is also provided, and the user can actively select whether to display the denoised electrocardiogram signals. For example, if the currently collected original electrocardiogram signals have a lot of noise, resulting in disordered waveforms of the original electrocardiogram signals and making it difficult for the user to effectively extract electrocardiogram features, the user can actively select to display the denoised electrocardiogram signals at this time.

[0060] Specifically, the user can trigger a display start instruction. After obtaining this display start instruction, the denoised electrocardiogram signals can be displayed to the user. For example, the electrocardiograph used to implement this method is provided with a button for starting denoised display. This button can be a physical button or a virtual button set in the software interface. The user can trigger the display start instruction by operating this button; alternatively, the user can also input a display start instruction based on a voice command. This embodiment does not limit this.

[0061] Among them, when the display start instruction is obtained, the original electrocardiogram signals and the denoised electrocardiogram signals are displayed in a comparative manner, so that the original electrocardiogram signals and the denoised electrocardiogram signals can be compared with each other. In addition to viewing the denoised electrocardiogram signals, the user can compare the two to manually judge whether the current denoising effect is ideal, and then determine whether the health status of the subject can be analyzed based on the denoised electrocardiogram signals in the future. If the current denoising effect is good, the analysis can be performed based on the denoised electrocardiogram signals; since it is difficult to denoise exercise electrocardiogram signals, if the current denoising effect is poor, for example, normal electrocardiogram features are misidentified as noise and filtered out, the analysis can be performed based on the original electrocardiogram signals.

[0062] For example, the area for displaying the electrocardiogram (ECG) signal can be partitioned, such as into upper and lower regions. One region displays the original ECG signal, and the other region displays the denoised ECG signal to achieve a comparative display.

[0063] It should be noted that this embodiment does not limit the timing of generating the denoised ECG signal. Specifically, the original ECG signal can be denoised after receiving the display start instruction to generate and display the denoised ECG signal; or, when the original ECG signal is acquired, it can be denoised in real time, so that when the display start instruction is received, the denoised ECG signal can be immediately displayed to the user, reducing the user's waiting time.

[0064] The ECG signal processing method provided in this embodiment denoises the original ECG signal collected during the exercise stress test, and can generate a denoised ECG signal with less noise interference. Moreover, the user can input a display start instruction based on their own needs, so that the original ECG signal and the denoised ECG signal can be displayed comparatively. By displaying the denoised ECG signal to the user, the noise interference in the original ECG signal can be reduced, assisting the user to effectively identify the ECG features therein, facilitating subsequent accurate analysis. And the comparative display method enables the user to observe the waveform patterns of the ECG signals before and after denoising, and the user can independently choose to perform subsequent analysis and diagnosis based on the original ECG signal or the denoised ECG signal, which is beneficial to improving the accuracy of the analysis of exercise ECG signals.

[0065] In this embodiment, an ECG signal processing method is provided, which can be used in devices capable of collecting and displaying ECG signals, such as electrocardiographs. Figure 2 is a flowchart of the ECG signal processing method according to an embodiment of the present invention, as Figure 2 shown, and this process includes the following steps.

[0066] Step S201: Acquire the original ECG signal collected during the exercise stress test.

[0067] For details, please refer to Figure 1 Step S101 of the embodiment shown, which will not be elaborated here.

[0068] In some optional embodiments, after the above step S201 "Acquire the original ECG signal collected during the exercise stress test", the quality of the ECG signal can also be analyzed. Specifically, the method can further include the following steps A1 to A2.

[0069] Step A1: Analyze the original ECG signal within the current time period to determine the quality level of the original ECG signal.

[0070] In this embodiment, after the original electrocardiogram (ECG) signal is acquired, the quality discrimination and prompt of the ECG signal can also be performed. Specifically, during the acquisition of the exercise ECG signal, the original ECG signal can be acquired in real time; and the quality of a segment of the original ECG signal within the current time period can be analyzed to determine the quality level of the current time period. The current time period can be the time period closest to the current moment; for example, the quality level of the original ECG signal in the current time period can be determined every certain time length (such as 1 second).

[0071] Among them, the quality of the original ECG signal within the current time period can be automatically classified by an automated method such as artificial intelligence, and different classification categories correspond to different quality levels; for example, the classification categories can include three quality levels: good, medium, and poor signal quality. "Good" signal quality means that the original ECG signal within the current time period is basically free of signal noise interference, and the electrocardiogram can be clearly observed. "Medium" signal quality means that the original ECG signal within the current time period contains a certain amount of signal noise interference, which has a certain interference effect on the analysis of the electrocardiogram. "Poor" signal quality means that the original ECG signal within the current time period has a large amount of noise interference and it is difficult to effectively read the electrocardiogram.

[0072] For example, a deep learning convolutional neural network model can be constructed, and the quality level can be determined based on this convolutional neural network model. Specifically, the overall structure of the convolutional neural network model includes multiple convolutional layers, pooling layers, activation layers, fully connected layers, and an output layer; the convolutional neural network model can automatically extract and fit the high-dimensional features of the ECG signal through the convolutional calculations in the convolutional layers, and finally return the signal quality classification probability value through the output layer, and select the category with the highest probability value, and convert the classification probability value into the corresponding classification category, that is, the quality level.

[0073] Specifically, the original ECG signal of the current time period is input into the convolutional neural network model, and the extraction of ECG features is realized through the calculations of multiple convolutional layers therein, and finally the quality level of the signal is output through the output layer of the convolutional neural network model to achieve classification.

[0074] Step A2, display the quality level in the interface in real time.

[0075] In this embodiment, after determining the quality level of the original electrocardiogram (ECG) signal, the quality level can be displayed in real time on the interface to prompt the user whether there is excessive noise in the original ECG signal collected during the current time period. For example, in addition to the area for displaying the ECG waveform, the entire display area of the electrocardiograph also includes other display areas, and the quality level can be newly displayed in these other display areas, such as high, medium, and low quality levels. Or, when the number of quality level classifications is small, different colors can also be used to represent different quality levels; for example, displaying color blocks corresponding to the quality level, or displaying the original ECG signal in the color corresponding to the quality level, etc. This embodiment does not limit the way of displaying the quality level.

[0076] By displaying the quality level of the original ECG signal to the user, it can assist the user in considering whether to select the original ECG signal for subsequent analysis; and, based on the current quality of the ECG signal, the movement state of the measured person can also be adjusted. For example, if the quality level is low, that is, the quality of the ECG signal is poor, the speed and slope of the treadmill can be reduced, or the measured person can be instructed to adjust the movement posture and state in order to reduce noise interference.

[0077] Step S202: Denoise the original ECG signal to generate a denoised ECG signal.

[0078] Since the signal noise generated during exercise is relatively complex and has aliasing, it is difficult to effectively denoise by filtering. In this embodiment, a neural network model for denoising exercise ECG signals, that is, a denoising model, is constructed, and denoising is achieved based on this denoising model. Specifically, the above step S202, "Denoise the original ECG signal to generate a denoised ECG signal", includes the following steps S2021 to S2022.

[0079] Step S2021: At every preset time interval, input the original ECG signal within the preset time interval into the denoising model; the denoising model is an encoder-decoder network model for outputting a sequence of the same length as the input signal.

[0080] Step S2022: According to the denoising model, perform encoding and decoding processing on the original ECG signal within the preset time interval to generate a denoised ECG signal within the preset time interval.

[0081] In this embodiment, the constructed denoising model is an encoder-decoder network model, that is, the denoising model includes an encoder and a decoder, and signal denoising is achieved by performing encoding and decoding processing on the input exercise ECG signal. And, to ensure that the generated denoised ECG signal is consistent with the original ECG signal, the input signal and the output signal of this denoising model are sequences of the same length.

[0082] Among them, to ensure the real-time performance of denoising, the corresponding original electrocardiogram (ECG) signal is denoised at regular intervals of a preset time period. For example, if the preset time period is 1 s, then every 1 s, the latest obtained original ECG signal within the current 1 s can be input into the denoising model to generate the latest denoised ECG signal in real time; correspondingly, the denoising model also outputs a denoised ECG signal with a duration of 1 s, which is of the same length as the input original ECG signal.

[0083] In this embodiment, the encoder of the denoising model is used to encode the input original ECG signal to map the data of the original ECG signal and extract the ECG features and noise features therein; the decoder is used to reconstruct the data based on the features extracted by the encoder to try to restore the noise-free ECG signal, so the ECG signal output by the encoder can be used as the denoised ECG signal.

[0084] Optionally, as shown in Figure 3 the denoising model includes an encoder 310, a decoder 320, and an output layer 330; the encoder 310 includes n encoding units 311, and the decoder 320 includes n decoding units 321; the encoding unit 311 includes a first convolutional part and a pooling layer; the decoding unit 321 includes an upsampling layer and a second convolutional part; both the first convolutional part and the second convolutional part include at least one convolutional layer. Figure 3 Taking n = 4 as an example, and both the first convolutional part and the second convolutional part include two convolutional layers.

[0085] Among them, the first convolutional part of the i-th encoding unit is used to perform convolutional processing on the input data and output it to the pooling layer of the i-th encoding unit; the pooling layer of the i-th encoding unit is used to perform pooling processing on the input data to generate pooled data; among them, the input data of the first convolutional part of the first encoding unit is the original ECG signal within the preset time period, and the input data of the first convolutional part of the remaining encoding units is the pooled data generated by the previous encoding unit. It can be understood that i = 1, 2,..., n.

[0086] And, the upsampling layer of the j-th decoding unit is used to perform upsampling processing on the output data of the (j - 1)-th decoding unit to generate upsampled data, and fuse the upsampled data with the pooled data of the same size output by one of the encoding units, and input the fused data into the second convolutional part of the j-th decoding unit; the second convolutional part of the j-th decoding unit is used to perform convolutional processing on the input data and output. It can be understood that j = 1, 2,..., n.

[0087] The output layer 330 is used to process the output data of the last decoding unit based on the activation function and output the denoised ECG signal within the preset time period.

[0088] Specifically, the sequence numbers of the encoding units and decoding units are defined according to the sequence of encoding and decoding. The i-th encoding unit is the unit that needs to perform encoding in the i-th position during the encoding process. Correspondingly, the j-th decoding unit is the unit that needs to perform decoding in the j-th position during the decoding process. Taking Figure 3 as an example, among the four encoding units 311 in the encoder 310, they are the 1st encoding unit, the 2nd encoding unit, the 3rd encoding unit, and the 4th encoding unit in sequence from left to right (from top to bottom). Among the four decoding units 321 in the decoder 320, they are the 1st decoding unit, the 2nd decoding unit, the 3rd decoding unit, and the 4th decoding unit in sequence from left to right (from bottom to top).

[0089] Among them, the input data of the 1st encoding unit is the original electrocardiogram signal. After the convolution processing of its first convolution part and the pooling processing of the pooling layer, the corresponding pooling data Pool-1 can be generated. The pooling layer can specifically be a max pooling layer, an average pooling layer, etc. Through the pooling processing, downsampling of the input data can be achieved. For example, the pooling data Pool-1 is the data generated by downsampling the original electrocardiogram signal.

[0090] After that, for the remaining encoding units, such as Figure 3 the 2nd encoding unit, the 3rd encoding unit, and the 4th encoding unit, their input data are all the pooling data output by the previous encoding unit. As Figure 3 shown, the input data of the 2nd encoding unit is the pooling data Pool-1, and its output is the pooling data Pool-2; the input data of the 3rd encoding unit is the pooling data Pool-2, and its output is the pooling data Pool-3; the input data of the 4th encoding unit is the pooling data Pool-3, and its output is the pooling data Pool-4. The working principles of the convolution layer and the pooling layer therein are similar to those of the 1st encoding unit and will not be elaborated here.

[0091] For the decoder 320, since there is no 0th decoding unit, the input data of the 1st decoding unit can be empty, that is, the corresponding upsampling data is empty. For example, the upsampling data has a certain size, but all elements therein are 0. And the upsampling data of the 1st decoding unit has the same size as the pooling data output by one of the encoding units. Generally, it has the same size as the pooling data output by the last encoding unit. As Figure 3As shown, the upsampled data of the first decoding unit has the same size as the pooled data Pool-4 output by the last encoding unit (i.e., the fourth encoding unit). Therefore, the first decoding unit can fuse the empty upsampled data with the pooled data Pool-4 to generate the fused data. After that, the second convolutional part of the first decoding unit can perform convolutional processing on the fused data and output the corresponding convolutional processing result Cov-1.

[0092] Alternatively, the encoder 310 may further include an (n + 1)-th encoding unit that performs convolutional processing and pooling processing on the pooled data output by the n-th encoding unit and outputs the corresponding pooled data. And the pooled data is input to the first encoding unit. That is, the input data of the first encoding unit is the pooled data output by the (n + 1)-th encoding unit. It performs upsampling on the pooled data to generate upsampled data and fuses the upsampled data with the pooled data output by the n-th encoding unit.

[0093] Among them, the upsampled data and the pooled data with the same size as the upsampled data can be fused by means of splicing.

[0094] The input data of the second decoding unit is the convolutional processing result Cov-1. Its upsampling layer performs upsampling on the convolutional processing result Cov-1 to generate the corresponding upsampled data. And the upsampled data has the same size as the pooled data Pool-3 output by the third encoding unit. The two are fused and, after being processed by the second convolutional part, can output the corresponding convolutional processing result Cov-2.

[0095] For the remaining decoding units, such as Figure 3 the third decoding unit and the fourth decoding unit in Figure 3 their input data is the convolutional processing result output by their previous decoding unit. As

[0096] shown, the input data of the third decoding unit is the convolutional processing result Cov-2. The upsampled data generated after its upsampling is fused with the pooled data Pool-2 output by the second encoding unit and finally outputs the convolutional processing result Cov-3. The input data of the fourth decoding unit is the convolutional processing result Cov-3. The upsampled data generated after its upsampling is fused with the pooled data Pool-1 output by the first encoding unit and finally outputs the convolutional processing result Cov-4.

[0096] And the fourth decoding unit is the last decoding unit. The convolutional processing result Cov-4 output by it is input to the output layer 330. The output layer 330 is provided with an activation function, such as the Softmax function, etc. Based on the processing of the convolutional processing result Cov-4 by the output layer 330, the denoised signal, that is, the denoised electrocardiogram signal, is finally output.

[0097] Among them, the convolutional layers in the first convolutional part and the second convolutional part, in addition to being able to perform convolutional processing, may also include batch normalization (BN) processing and activation processing, such as activation processing based on the ReLU function, etc. The structure of this convolutional layer is not limited in this embodiment.

[0098] After generating the denoised electrocardiogram signal, the denoised electrocardiogram signal can be stored so that it can be displayed to the user after receiving the display start instruction.

[0099] In this embodiment, the pooled data of the encoding unit 311 can be fused with the upsampled data of the decoding unit 321 at the corresponding stage, thereby forming a skip connection. Through this skip connection architecture, at each stage, the decoding unit 321 is allowed to learn the relevant features lost in the pooling of the encoding unit 311. Therefore, the convolutional layer can remove the noise of the signal more accurately and can generate a more accurate denoised electrocardiogram signal.

[0100] Step S203, in response to the display start instruction triggered by the user, display the original electrocardiogram signal and the denoised electrocardiogram signal in comparison.

[0101] Among them, for details, please refer to Figure 1 Step S103 of the embodiment shown, which will not be elaborated here.

[0102] Optionally, since the input of the denoising model is the original electrocardiogram signal of a period of time (i.e., the current period), there is a corresponding delay in the generated denoised electrocardiogram signal; in order to be able to display in comparison better, the method further includes: when displaying the original electrocardiogram signal, delaying the display of the original electrocardiogram signal, and the duration of the delayed display matches the preset period.

[0103] In this embodiment, the displayed original electrocardiogram signal is the electrocardiogram signal collected some time ago, and there is a certain delay. Specifically, after the original electrocardiogram signal is collected, it is not displayed immediately, but after this preset period, the original electrocardiogram signal is displayed again. Since this preset period is generally not very long, for example, this preset period is 1 s, and delaying the display of the original electrocardiogram signal by 1 s will not affect subsequent analysis.

[0104] By adopting the method of delaying the display of the original electrocardiogram signal, when the display start instruction input by the user is obtained, although there is a delay in the denoised electrocardiogram signal, the delayed display of the original electrocardiogram signal can be synchronized with the delayed denoised electrocardiogram signal, so that it can be better displayed in comparison.

[0105] In some alternative embodiments, due to the limited size of the display screen of the electrocardiograph and the fact that multiple leads simultaneously collect electrocardiogram signals during an exercise stress test, that is, there are multi-channel raw electrocardiogram signals. When performing comparison display, if zoning is carried out, it will affect the display effect of the multi-channel electrocardiogram signals. In this embodiment, during comparison display, taking advantage of the characteristic that the overall change of the electrocardiogram signal is generally small within a period of time, sequential comparison display is adopted, so that the display area can be fully used to display the raw electrocardiogram signal or the denoised electrocardiogram signal, ensuring the display effect.

[0106] Specifically, the above step S203, "comparison display of the raw electrocardiogram signal and the denoised electrocardiogram signal", may specifically include the following steps B1 to B2.

[0107] Step B1, stop updating the display of the raw electrocardiogram signal.

[0108] Step B2, replace the denoised electrocardiogram signal with the raw electrocardiogram signal, and display the denoised electrocardiogram signal in at least part of the display area used to display the raw electrocardiogram signal.

[0109] In this embodiment, after the raw electrocardiogram signal is obtained, the raw electrocardiogram signal can be displayed, and the raw electrocardiogram signal is displayed in a real-time update manner. Limited by the size of the display screen, the display area of the electrocardiograph can generally only display the electrocardiogram signal for a period of time (for example, 10 s), and the signal points in the raw electrocardiogram signal are sequentially displayed in the order from left to right to achieve signal update; when the display position moves to the rightmost end of the display area, the next signal point is displayed at the leftmost segment of the display area, covering the previously displayed signal point. Therefore, the signal points in the displayed raw electrocardiogram signal can be continuously displayed in the display area for a period of time.

[0110] If during the display of the raw electrocardiogram signal, the user inputs an instruction to start display, stop updating the display of the raw electrocardiogram signal, that is, no longer update the raw electrocardiogram signal. And, replace the raw electrocardiogram signal with the denoised electrocardiogram signal, and display the corresponding denoised electrocardiogram signal at the position originally used to display the raw electrocardiogram signal.

[0111] Specifically, if the raw electrocardiogram signal collected at the i-th moment is currently displayed, if the comparison display function needs to be enabled later, then in the display area where the raw electrocardiogram signal collected at the (i + 1)-th moment should originally be displayed, display the denoised electrocardiogram signal generated by denoising the raw electrocardiogram signal collected at the (i + 1)-th moment, so as to realize the replacement of the raw electrocardiogram signal with the denoised electrocardiogram signal. And, since the previously displayed raw electrocardiogram signal can be kept displayed in the display area for a period of time, the previously displayed raw electrocardiogram signal can be compared with the subsequently displayed denoised electrocardiogram signal to achieve comparison display.

[0112] Optionally, when the user does not need to display the denoised electrocardiogram (ECG) signal, the denoising function can also be turned off. Specifically, the method further includes: in response to the close display instruction triggered by the user, stopping the display of the denoised ECG signal.

[0113] Similar to the above-mentioned trigger to turn on the display instruction, the user can also input the close display instruction by means such as pressing a button or voice. For example, if the user finds that the denoising effect is poor after turning on the denoising display, or there is less current noise and the denoising effect is not obvious, the user can input the close display instruction based on their own needs, so as to stop displaying the denoised ECG signal and turn off the denoising.

[0114] It can be understood that the denoised ECG signal can also be stopped from being displayed based on a similar processing method to the above-mentioned steps B1 to B2, so as to normally display the original ECG signal again, which will not be elaborated here.

[0115] The ECG signal processing method provided in this embodiment can more accurately extract the ECG features based on the denoising model, and can achieve precise denoising; moreover, the pooled data of the encoding unit can be fused with the upsampled data of the decoding unit at the corresponding stage. Through this skip connection architecture, the decoding unit is allowed to learn the relevant features lost in the pooling of the encoding unit at each stage. Therefore, the convolutional layer is more precise in removing the noise of the signal and can generate a more accurate denoised ECG signal.

[0116] In this embodiment, an ECG signal processing method is provided, which can be used in devices capable of collecting and displaying ECG signals, such as electrocardiographs. Figure 4 is a flowchart of the ECG signal processing method according to an embodiment of the present invention, as Figure 4 shown, and this process includes the following steps.

[0117] Step S401, obtain the original ECG signal collected during the exercise stress test.

[0118] Among them, for details, please refer to Figure 1 step S101 of the embodiment shown, which will not be elaborated here.

[0119] Step S402, analyze the original ECG signal within the current time period to determine the quality level of the original ECG signal.

[0120] Among them, for details, please refer to the relevant description of the above step A1, which will not be elaborated here.

[0121] Step S403, display the quality level in real time on the interface.

[0122] Among them, for details, please refer to the relevant description of the above step A2, which will not be elaborated here.

[0123] Step S404: Denoise the original electrocardiogram (ECG) signal to generate a denoised ECG signal.

[0124] For details, please refer to Figure 1 Step S102 of the embodiment shown, which will not be elaborated here.

[0125] Step S405: In response to the display start instruction triggered by the user, display the original ECG signal and the denoised ECG signal in comparison.

[0126] For details, please refer to Figure 1 Step S103 of the embodiment shown, which will not be elaborated here.

[0127] Step S406: Detect the type of characteristic wave to which the signal point in the target ECG signal belongs; the target ECG signal is the original ECG signal or the denoised ECG signal.

[0128] The cardiac excitation conducts and spreads in the atria and ventricles, respectively exciting the atrial muscle and the ventricular muscle. After a certain period of time, the atrioventricular muscle returns to the resting state before excitation. These repetitive electrical changes are collected and recorded by an electrocardiograph to form a unique ECG waveform, namely the electrocardiogram. One cardiac cycle in the electrocardiogram is called a heartbeat; the wave band of the heartbeat reflects the electrical activity of the heart, and the waveform pattern and law of the heartbeat signal also have important reference significance clinically.

[0129] The basic waveforms of the electrocardiogram are named with letters: P wave - atrial depolarization, QRS complex - ventricular depolarization, ST segment, T wave - ventricular repolarization. The heartbeat, as the basic unit of the electrocardiogram, is a signal segment containing P wave, QRS wave, T wave, etc. Among them, the P wave represents the depolarization of the entire atrium; the QRS complex represents the spread of excitation in the ventricular muscle and causes ventricular depolarization; the ST segment and T wave represent the process of the ventricular muscle recovering from the electrically excited state to the electrically resting state, that is, ventricular repolarization.

[0130] In this embodiment, these basic waveforms are called characteristic waves. Correspondingly, these characteristic waves can be divided into multiple types such as P wave, QRS wave, T wave, etc., that is, characteristic wave types. Moreover, the ECG signal contains multiple signal points. Generally, each characteristic point corresponds to a characteristic wave; the signal points that do not belong to any characteristic wave can be classified as background, that is, the specific characteristic wave types can include: P wave, QRS wave, T wave, background, etc.

[0131] It is possible to determine the type of characteristic wave to which each characteristic point in the collected original ECG signal belongs, and it is also possible to determine the type of characteristic wave to which each characteristic point in the denoised ECG signal belongs. For the convenience of description, the target ECG signal is used to represent the original ECG signal or the denoised ECG signal. For example, since noise may also affect the accuracy of identifying the characteristic wave type, the characteristic wave type identification can be performed only on the denoised ECG signal.

[0132] For example, an encoder-decoder convolutional neural network structure can be used to implement characteristic wave detection for ECG signals to determine the characteristic wave type to which each signal point belongs. Specifically, the output layer of the neural network structure can output a sequence of the same length as the input signal, and the output sequence is the characteristic wave detection result of the input original ECG signal. The detection result is the predicted probability value of whether each signal point belongs to the four categories of P wave, QRS wave, T wave, and background. According to the category with the largest predicted probability value, it is predicted that the signal point belongs to which type of P wave, QRS wave, T wave, and background, and then the result of which characteristic wave all the signal points in the entire original ECG signal belong to can be obtained; and the starting point position and the ending point position of all the P waves, QRS waves, and T waves that appear in the original ECG signal can be obtained, so as to obtain the position information of each characteristic wave in the entire original ECG signal.

[0133] Step S407, when displaying the target ECG signal, synchronously mark the characteristic wave type to which the signal point in the target ECG signal belongs.

[0134] In this embodiment, if the target ECG signal is an original ECG signal, when the original ECG signal needs to be displayed, the characteristic wave type to which the signal point in the original ECG signal belongs can be marked while displaying the original ECG signal. Similarly, when the denoised ECG signal needs to be displayed, the characteristic wave type to which the signal point belongs can also be marked synchronously.

[0135] Among them, one processing process of ECG signal can be seen in Figure 5 As shown. Figure 5 As shown, the electrocardiograph can collect the original ECG signal, which may contain a lot of noise interference; perform quality analysis on it, determine the corresponding quality level and display it, such as which quality level is high, medium, or low; denoise the original ECG signal in real time to form a denoised ECG signal, and display the denoised ECG signal when receiving a display start instruction. In addition, the characteristic wave type of each signal point in the denoised ECG signal can be determined to display which specific characteristic wave it belongs to when displaying the denoised ECG signal.

[0136] Alternatively, since the characteristic wave contains more signal points, only some of the signal points may be marked. Alternatively, since the characteristic wave is a waveform, which is a signal segment in the ECG signal, the signal segment may be marked as a whole, thereby marking all signal points in the signal segment.

[0137] Specifically, the above step S407 "synchronously marking the characteristic wave type to which the signal point in the target ECG signal belongs" may include: marking signal segments of different characteristic wave types with different colors; the signal segment includes multiple signal points.

[0138] In this embodiment, based on the type of characteristic wave to which each identified signal point belongs, the starting position of each type of characteristic wave can be determined. For example, the starting point positions and ending point positions of all the P waves, QRS waves, and T waves that appear in the original electrocardiogram (ECG) signal can be determined. A signal segment between these starting positions, that is, a signal segment belonging to the same characteristic wave, can be marked with a color. For example, the P wave is represented by red, the QRS wave is represented by purple, the T wave is represented by blue, and the background is represented by green, etc.

[0139] The method for processing the ECG signal provided in this embodiment can automatically mark the characteristic waves of the ECG signal, achieve automatic positioning of the characteristic waves of each cardiac beat in the exercise electrocardiogram, and achieve visual identification; marking the characteristic waves of the exercise electrocardiogram on the electrocardiogram can assist the user in quickly finding the positions of the characteristic waves, thereby improving the user's efficiency in reading electrocardiograms. Marking with different colors can conveniently mark all signal points and can visually show the user the positions of each characteristic wave, with a good visual effect.

[0140] In this embodiment, an electrocardiograph is also provided. The electrocardiograph includes a processing device and a display device. Among them, the processing device can be a processing chip with processing functions, and the display device can be a display screen.

[0141] The processing device is used to execute the method for processing the ECG signal provided in the above embodiment to achieve noise reduction of the ECG signal, etc.; and, the processing device controls the display device to display the corresponding ECG signal. Specifically, the processing device is configured to: acquire the original ECG signal collected during the exercise stress test; perform noise reduction processing on the original ECG signal to generate a denoised ECG signal; in response to the display start instruction triggered by the user, control the display device to display the original ECG signal and the denoised ECG signal in comparison.

[0142] The processes of the processing device for realizing quality analysis, characteristic wave detection, etc. are similar to the principles of the above method embodiments and will not be elaborated here.

[0143] In this embodiment, a processing device for ECG signals is also provided. The device is used to implement the above embodiments and preferred implementation manners, and those that have been described will not be repeated. As used hereinafter, the term "module" can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0144] This embodiment provides a processing device for ECG signals, as Figure 6 shown, including:

[0145] An acquisition module 601, configured to acquire the original electrocardiogram signal collected during the exercise stress test;

[0146] A denoising module 602, configured to perform denoising processing on the original electrocardiogram signal to generate a denoised electrocardiogram signal;

[0147] A display module 603, configured to, in response to an enabling display instruction triggered by a user, display the original electrocardiogram signal and the denoised electrocardiogram signal in comparison.

[0148] In some alternative embodiments, the denoising module 602 performs denoising processing on the original electrocardiogram signal to generate a denoised electrocardiogram signal, including:

[0149] At every preset time period, input the original electrocardiogram signal within the preset time period into a denoising model; the denoising model is an encoder-decoder network model for outputting a sequence of the same length as the input signal;

[0150] Perform encoder-decoder processing on the original electrocardiogram signal within the preset time period according to the denoising model to generate the denoised electrocardiogram signal within the preset time period.

[0151] In some alternative embodiments, the denoising model includes an encoder, a decoder, and an output layer; the encoder includes n encoding units, and the decoder includes n decoding units; the encoding unit includes a first convolutional part and a pooling layer; the decoding unit includes an upsampling layer and a second convolutional part; both the first convolutional part and the second convolutional part include at least one convolutional layer;

[0152] The first convolutional part of the i-th encoding unit is configured to perform convolutional processing on the input data and output it to the pooling layer of the i-th encoding unit;

[0153] The pooling layer of the i-th encoding unit is configured to perform pooling processing on the input data to generate pooled data;

[0154] Wherein, the input data of the first convolutional part of the first encoding unit is the original electrocardiogram signal within the preset time period, and the input data of the first convolutional part of the remaining encoding units is the pooled data generated by the previous encoding unit;

[0155] The upsampling layer of the j-th decoding unit is configured to perform upsampling processing on the output data of the (j - 1)-th decoding unit to generate upsampled data, and fuse the upsampled data and the pooled data of the same size output by one of the encoding units, and input the fused data into the second convolutional part of the j-th decoding unit;

[0156] The second convolutional part of the j-th decoding unit is configured to perform convolutional processing on the input data and output;

[0157] The output layer is used to process the output data of the last decoding unit based on an activation function, and output the denoised electrocardiogram signal within the preset time period.

[0158] In some alternative embodiments, the display module 603 is further configured to:

[0159] When displaying the original electrocardiogram signal, delay the display of the original electrocardiogram signal, and the delay duration matches the preset time period.

[0160] In some alternative embodiments, the display module 603 displays the original electrocardiogram signal and the denoised electrocardiogram signal in comparison, including:

[0161] Stop updating the display of the original electrocardiogram signal;

[0162] Replace the original electrocardiogram signal with the denoised electrocardiogram signal, and display the denoised electrocardiogram signal in at least part of the display area for displaying the original electrocardiogram signal.

[0163] In some alternative embodiments, the device further includes a quality determination module, configured to:

[0164] After the acquisition module 601 acquires the original electrocardiogram signal collected during the exercise stress test, analyze the original electrocardiogram signal within the current time period to determine the quality level of the original electrocardiogram signal;

[0165] The display module 603 is further configured to display the quality level in real time in the interface.

[0166] In some alternative embodiments, the device further includes a characteristic wave detection module, configured to: detect the type of characteristic wave to which a signal point in the target electrocardiogram signal belongs; the target electrocardiogram signal is the original electrocardiogram signal or the denoised electrocardiogram signal;

[0167] The display module 603 is further configured to, when displaying the target electrocardiogram signal, synchronously mark the type of characteristic wave to which the signal points in the target electrocardiogram signal belong.

[0168] In some alternative embodiments, the display module 603 synchronously marks the type of characteristic wave to which the signal points in the target electrocardiogram signal belong, including:

[0169] Mark signal segments of different characteristic wave types with different colors; the signal segments include multiple signal points.

[0170] In some alternative embodiments, the display module 603 is further configured to: in response to a close display instruction triggered by a user, stop displaying the denoised electrocardiogram signal.

[0171] The further function descriptions of the above-mentioned modules and units are the same as those in the corresponding embodiments above, and will not be repeated here.

[0172] In this embodiment, the electrocardiogram signal processing device is presented in the form of functional units. Here, the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and a memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0173] An embodiment of the present invention further provides an electronic device having the above-mentioned Figure 6 shown electrocardiogram signal processing device.

[0174] Please refer to Figure 7 , Figure 7 which is a schematic structural diagram of an electronic device provided by an optional embodiment of the present invention. As Figure 7 shown, the electronic device includes: one or more processors 10, a memory 20, and an interface for connecting various components, including a high-speed interface and a low-speed interface. Each component communicates with each other using different buses and can be installed on a common main board or installed in other ways as needed. The processor can process instructions executed within the electronic device, including instructions stored in the memory or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple electronic devices can be connected, and each device provides some necessary operations (such as an array of servers, a set of blade servers, or a multi-processor system). Figure 7 In

[0175] Processor 10 can be a central processing unit, a network processor, or a combination thereof. Among them, processor 10 can further include a hardware chip. The above-mentioned hardware chip can be an application specific integrated circuit, a programmable logic device, or a combination thereof. The above-mentioned programmable logic device can be a complex programmable logic device, a field programmable gate array, a general array logic, or any combination thereof.

[0176] Among them, the memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0177] The memory 20 may include a program storage area and a data storage area. The program storage area may store an operating system and application programs required for at least one function. The data storage area may store data created according to the use of the electronic device and the like. In addition, the memory 20 may include a high-speed random access memory and may also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some alternative embodiments, the memory 20 may optionally include a memory remotely disposed relative to the processor 10, and these remote memories may be connected to the electronic device through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0178] The memory 20 may include a volatile memory, such as a random access memory. The memory may also include a non-volatile memory, such as a flash memory, a hard disk, or a solid-state drive. The memory 20 may also include a combination of the above types of memories.

[0179] The electronic device further includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30, and the output device 40 may be connected through a bus or other means. Figure 7 Taking the connection through the bus as an example.

[0180] The input device 30 may receive input digital or character information and generate key signal inputs related to the user settings and function controls of the electronic device, such as a touch screen, a keypad, a mouse, a trackpad, a touchpad, a pointing stick, one or more mouse buttons, a trackball, a joystick, etc. The output device 40 may include a display device, an auxiliary lighting device (e.g., an LED), and a haptic feedback device (e.g., a vibration motor), etc. The above-mentioned display device includes but is not limited to a liquid crystal display, a light-emitting diode, a display, and a plasma display. In some alternative embodiments, the display device may be a touch screen.

[0181] Embodiments of the present invention also provide a computer-readable storage medium. The methods according to the embodiments of the present invention can be implemented in hardware, firmware, or be implemented as computer code that can be recorded on a storage medium, or be implemented by downloading over a network and originally stored in a remote storage medium or a non-transitory machine-readable storage medium and will be stored in a local storage medium, so that the methods described herein can be stored in such software processes on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code, and when the software or computer code is accessed and executed by the computer, the processor, or the hardware, the methods shown in the above embodiments are implemented.

[0182] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for processing electrocardiogram signals, characterized in that, The method includes: Obtaining the original electrocardiogram signal collected during the exercise stress test; Performing denoising processing on the original electrocardiogram signal to generate a denoised electrocardiogram signal; In response to the user-triggered display start instruction, displaying the original electrocardiogram signal and the denoised electrocardiogram signal for comparison.

2. The method according to claim 1, characterized in that, The performing denoising processing on the original electrocardiogram signal to generate a denoised electrocardiogram signal includes: Every preset time period, inputting the original electrocardiogram signal within the preset time period into a denoising model; the denoising model is an encoder-decoder network model for outputting a sequence of the same length as the input signal; Performing encoding and decoding processing on the original electrocardiogram signal within the preset time period according to the denoising model to generate the denoised electrocardiogram signal within the preset time period.

3. The method according to claim 2, characterized in that, The denoising model includes an encoder, a decoder, and an output layer; the encoder includes n encoding units, and the decoder includes n decoding units; the encoding unit includes a first convolutional part and a pooling layer; the decoding unit includes an upsampling layer and a second convolutional part; both the first convolutional part and the second convolutional part include at least one convolutional layer; The first convolutional part of the i-th encoding unit is used to perform convolutional processing on the input data and output it to the pooling layer of the i-th encoding unit; The pooling layer of the i-th encoding unit is used to perform pooling processing on the input data to generate pooled data; Among them, the input data of the first convolutional part of the first encoding unit is the original electrocardiogram signal within the preset time period, and the input data of the first convolutional part of the remaining encoding units is the pooled data generated by the previous encoding unit; The upsampling layer of the j-th decoding unit is used to perform upsampling processing on the output data of the (j - 1)-th decoding unit to generate upsampled data, and fuse the upsampled data and the pooled data of the same size output by one of the encoding units, and input the fused data into the second convolutional part of the j-th decoding unit; The second convolutional part of the j-th decoding unit is used to perform convolutional processing on the input data and output; The output layer is used to process the output data of the last decoding unit based on an activation function and output the denoised electrocardiogram signal within the preset time period.

4. The method according to claim 2, characterized in that, It further includes: When displaying the original electrocardiogram signal, delaying the display of the original electrocardiogram signal, and the delay duration matches the preset time period.

5. The method according to claim 1, characterized in that, The displaying the original electrocardiogram signal and the denoised electrocardiogram signal for comparison includes: Stopping updating the display of the original electrocardiogram signal; Replacing the original electrocardiogram signal with the denoised electrocardiogram signal and displaying the denoised electrocardiogram signal in at least part of the display area for displaying the original electrocardiogram signal.

6. The method according to claim 1, characterized in that, After obtaining the original electrocardiogram signal collected during the exercise stress test, it further includes: Analyzing the original electrocardiogram signal within the current time period to determine the quality level of the original electrocardiogram signal; Realtime displaying the quality level in the interface.

7. The method according to claim 1, characterized in that, It further includes: Detecting the type of characteristic wave to which the signal point in the target electrocardiogram signal belongs; the target electrocardiogram signal is the original electrocardiogram signal or the denoised electrocardiogram signal; When displaying the target electrocardiogram signal, synchronously mark the type of characteristic wave to which the signal points in the target electrocardiogram signal belong.

8. The method according to claim 7, characterized in that, The synchronously marking the type of characteristic wave to which the signal points in the target electrocardiogram signal belong includes: Marking signal segments of different characteristic wave types with different colors; the signal segments include multiple signal points.

9. The method according to claim 1, characterized in that, It further includes: In response to a display-off instruction triggered by the user, stop displaying the denoised electrocardiogram signal.

10. An electrocardiograph, characterized in that, It includes: A processing device and a display device; The processing device is configured to: Obtain the original electrocardiogram signal collected during the exercise stress test; Perform denoising processing on the original electrocardiogram signal to generate a denoised electrocardiogram signal; In response to a display-on instruction triggered by the user, control the display device to contrastively display the original electrocardiogram signal and the denoised electrocardiogram signal.

11. An electronic device, characterized in that, It includes: A memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the electrocardiogram signal processing method according to any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that, Computer instructions are stored on the computer-readable storage medium, and the computer instructions are used to cause a computer to execute the electrocardiogram signal processing method according to any one of claims 1 to 9.

Images