Method for monitoring cardiac transport and thoracotomy based on high spatio-temporal resolution electrical mapping

Abstract

The present invention relates to the technical field of data processing, and more particularly to a method for monitoring cardiac transport and thoracotomy based on high spatio-temporal resolution electrical mapping. The method includes: Step S1: Generate an electrical mapping image and the correlation between mapping electrodes and obtain a plurality of target waveforms of electrical mapping signals; Step S2: Analyze the ECG signal in real time, perform denoising processing, and obtain a plurality of target waveforms; Step S3: Divide each target waveform into a plurality of sub-waveforms according to the waveform type and input them into the first model to obtain a first output result. When it is abnormal, execute Step S4. When it is normal, use a plurality of second normal waveforms with continuous time before the first normal waveform as a waveform sequence to input into the second model and obtain a second output result; Step S4: Calculate and analyze to obtain the abnormal position, generate a dynamic image, and mark it in the electrical mapping image. The present invention solves the problem of insufficient cardiac monitoring accuracy and improves the cardiac monitoring accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and more particularly to a method for monitoring cardiac transport and thoracotomy based on high spatio-temporal resolution electrical mapping. Background Art

[0002] With the development of medical electronics, high spatio-temporal resolution electrical mapping monitoring methods are widely used clinically, especially during and after cardiac transplantation transport and cardiothoracic surgery. Similar prior arts include a Chinese patent with publication number CN109688904A, which proposes a system and method for detecting arrhythmia electrocardiogram signals, including a plurality of threshold heart rates for detecting arrhythmia ECG segments and a plurality of rate-based sensitivity levels, where heart rates with higher clinical relevance are assigned rate-based sensitivity levels with higher sensitivity. The ECG signal is monitored by a medical device, and the monitored ECG signal is processed using the plurality of threshold heart rates and the plurality of rate-based sensitivity levels to detect and capture arrhythmia ECG segments. In addition, a similar prior art is a US patent with publication number US20140330145A1, which discloses a method for automatically determining local activation times in a multi-channel electrocardiogram signal including a plurality of cardiac channels. The method includes: storing the cardiac channel signals; calculating a first LAT value at a plurality of mapping channel positions using ventricular, reference, and mapping channels; monitoring the quality of at least one ventricular, reference, and mapping channel; if the quality of the monitored cardiac channel is below a standard, replacing the sub-standard channel with another channel among the plurality of channels with a quality higher than the standard; and calculating a second LAT value based on the replaced cardiac channel. The above two patents have solved the problem of cardiac monitoring, but have not considered the problem of combined monitoring of epicardial electrical mapping and ECG signals, with insufficient accuracy and unable to meet the requirements of high-precision monitoring. The present invention not only can intuitively obtain the abnormal position but also can further improve the monitoring accuracy by adopting high-resolution epicardial electrical mapping and combined monitoring with ECG signals. Summary of the Invention

[0003] In order to better solve the above problems, the present invention provides a method for monitoring cardiac transport and thoracotomy based on high spatio-temporal resolution electrical mapping, and the method includes the following steps:

[0004] Step S1: Collect electrical mapping signals of each mapping electrode in a monitoring object at a first preset period through an electrical mapping unit, and obtain physiological information of the monitoring object through calculation and analysis based on the plurality of electrical mapping signals. Generate an electrical mapping image and the correlation between the mapping electrodes based on the physiological information and the position information of each mapping electrode, where the mapping electrodes are arranged on the outer membrane of the monitoring object;

[0005] Step S2: The ECG signal acquisition unit collects the ECG signal of the monitoring object in real time, analyzes the ECG signal in real time to obtain an analysis result, and when the analysis result is abnormal, denoises the ECG signal and obtains multiple target waveforms based on the ECG signal;

[0006] Step S3: Divide each of the target waveforms into multiple sub-waveforms according to the waveform type, input the sub-waveforms into the first model to obtain a first output result. When the first output result is abnormal, execute Step S4. When the first output result is normal, obtain a first normal waveform according to the first output result, and also use N consecutive second normal waveforms before the first normal waveform as a waveform sequence to input into the second model and obtain a second output result;

[0007] Step S4: When the first output result is abnormal, obtain all abnormal electrodes according to the first output result, and obtain the abnormal position based on the electrical mapping signals of all the abnormal electrodes. When the second output result is abnormal, obtain the target electrodes according to the second output result, generate a dynamic image according to the electrical mapping signals of the target electrodes, and mark the abnormal position and the dynamic image in the electrical mapping image.

[0008] As a preferred technical solution, the mapping electrodes in the electrical mapping unit are at least 32 channels, and the distance between the electrode points in each mapping electrode is less than or equal to 4 mm.

[0009] As a preferred technical solution, Step S1 includes:

[0010] Step S11: The electrical mapping unit obtains the electrical mapping signals of each mapping electrode at a first preset period, and obtains the conduction information of the monitoring object according to the set positions of each mapping electrode and the corresponding electrical mapping signals. The conduction information includes excitation points, conduction directions, conduction speeds, depolarization dispersion and repolarization dispersion, conduction phases and spectral characteristics;

[0011] Step S12: Obtain the electrical mapping image based on the position information of each mapping electrode and the conduction information. Also, extract the electrical mapping waveforms by heartbeat for each electrical mapping signal, calculate the conduction feature points on each electrical mapping waveform, and obtain the waveform relationship and conduction relationship based on the time-space sequence of the electrical mapping waveforms and the conduction feature points. Among them, the correlation relationship includes the waveform relationship and the conduction relationship.

[0012] As a preferred technical solution, Step S2 includes:

[0013] Step S21: The ECG signal acquisition unit acquires the ECG signal of the monitoring object in real time, and extracts a plurality of first waveforms from the ECG signal, and the first waveform is an R wave;

[0014] Step S22: Obtain the maximum value of each first waveform and the maximum potential fluctuation value of the corresponding reference potential range, calculate the first difference between the maximum value and the first threshold, and the second difference between the maximum potential fluctuation value and the second threshold. When both the first difference and the second difference corresponding to each first waveform are within the corresponding set range, the analysis result corresponding to the ECG signal is normal; otherwise, the analysis result corresponding to the ECG signal is abnormal. The ECG signal is used as a noise waveform, the noise waveform is denoised, and the denoised ECG signal and the ECG signal when the analysis result is normal are used as the target waveform.

[0015] As a preferred technical solution, the step S3 includes:

[0016] Step S31: Divide each target waveform into a plurality of sub-waveforms according to the waveform type, input the plurality of sub-waveforms into the first model, and obtain the first output result. The first model is a machine learning model trained by first learning data. The first learning data includes the sub-waveforms corresponding to the ECG signals of historical heart disease patients stored in the database and the corresponding disease types, and also includes the sub-waveforms corresponding to the ECG signals of normal people;

[0017] Step S32: When the first output result is abnormal, obtain the disease types corresponding to the input plurality of sub-waveforms through the first output result, and execute step S4;

[0018] Step S33: When the first output result is normal, obtain the target waveforms corresponding to the input plurality of sub-waveforms, use the target waveform as the first normal waveform, and obtain N time-continuous second normal waveforms before the first normal waveform. The first normal waveform and the N second normal waveforms are used as a waveform sequence to be input into the second model, and a second output result is obtained. The second model is a machine learning model trained by second learning data.

[0019] As a preferred technical solution, the second learning data includes historical abnormal electro-mapping data stored in the database and the historical ECG signal sequence within a preset time period before the acquisition time of the historical abnormal electro-mapping data, and also includes the disease types corresponding to the historical abnormal electro-mapping data.

[0020] As a preferred technical solution, the step S4 includes the following steps:

[0021] Step S41: When the first output result is abnormal, obtain the first acquisition time of the target waveform corresponding to the first output result, obtain the electro-mapping signals collected by each of the mapping electrodes at the second acquisition time closest to the first acquisition time, extract electro-mapping waveforms from each of the electro-mapping signals according to heartbeats, calculate conduction feature points of each electro-mapping waveform, obtain the electrode conduction relationship according to the temporal and spatial order of the electro-mapping waveforms and conduction feature points of the electrode and other electrodes around the electrode, compare the electrode conduction relationship with the relevant relationship corresponding to the mapping electrode. When the two are consistent, the mapping electrode is a normal electrode; when the two are inconsistent, the mapping electrode is an abnormal electrode, and obtain the position of the abnormal electrode closest to the excitation point position and the positions of other abnormal electrodes, and generate an abnormal position based on the abnormal electrode position and the positions of other abnormal electrodes and mark it in the electro-mapping image;

[0022] Step S42: When the second output result is abnormal, obtain the target position in the electro-mapping image according to the historical abnormal electro-mapping data in the second output result, obtain the corresponding target electrode based on the target position, reduce the acquisition period of the electro-mapping unit for the target electrode to a second preset period, and save the electro-mapping signals collected by the target electrode, and generate a dynamic image of the electro-mapping signals of the saved target electrode within the corresponding range of the electro-mapping image and mark it in the electro-mapping image;

[0023] Step S43: The user takes corresponding measures based on the abnormal position and the electro-mapping signals at the abnormal position, and recognizes the abnormality in real time according to the dynamic image.

[0024] As a preferred technical solution, step S4 further includes:

[0025] When the second output result is normal, increase the acquisition period of the electro-mapping unit to a third preset period, where the third preset period is less than the first preset period, and the first preset period is less than the second preset period.

[0026] As a preferred technical solution, the mapping electrode has flexibility and biocompatibility.

[0027] The present invention also provides a cardiac transport and open-chest surgery monitoring system based on high spatio-temporal resolution electro-mapping. The system is used to implement the above method, and the system includes:

[0028] An electrical mapping unit is configured to collect electrical mapping signals of each mapping electrode in a monitoring object at a first preset period, and obtain physiological information of the monitoring object through computational analysis based on a plurality of the electrical mapping signals. An electrical measurement image and a correlation relationship between the mapping electrodes are generated based on the physiological information and position information of each mapping electrode, wherein the mapping electrodes are disposed on the outer membrane of the monitoring object;

[0029] An ECG signal acquisition unit is configured to collect an ECG signal of a monitoring object in real time, obtain an analysis result by analyzing the ECG signal in real time, perform denoising processing on the ECG signal when the analysis result is abnormal, and obtain a plurality of target waveforms based on the ECG signal;

[0030] A discrimination and prediction unit is configured to divide each of the target waveforms into a plurality of sub-waveforms according to waveform types, input the sub-waveforms into a first model to obtain a first output result. When the first output result is abnormal, step S4 is executed. When the first output result is normal, a first normal waveform is obtained according to the first output result. Further, N consecutive second normal waveforms before the first normal waveform are used as a waveform sequence and input into a second model, and a second output result is obtained;

[0031] A display generation unit is configured to, when the first output result is abnormal, obtain all abnormal electrodes according to the first output result, and obtain an abnormal position based on the electrical mapping signals of all the abnormal electrodes. When the second output result is abnormal, obtain target electrodes according to the second output result, generate a dynamic image based on the electrical mapping signals of the target electrodes, and mark the abnormal position and the dynamic image in the mapping image.

[0032] Compared with the prior art, the beneficial effects of the present invention are at least as follows:

[0033] The present invention obtains the above-mentioned electro-mapping signals of each of the above-mentioned mapping electrodes at a first preset period through the above-mentioned electro-mapping unit, obtains the above-mentioned electro-mapping image according to the position information and the above-mentioned conduction information of each of the above-mentioned mapping electrodes, and obtains the correlation relationship between the above-mentioned mapping electrodes. Among them, the mapping electrodes are arranged on the epicardium of the heart, the above-mentioned electro-mapping unit is a high-resolution multi-channel mapping, and can monitor each position of the heart simultaneously, laying a foundation for accurately and quickly obtaining the abnormal state and abnormal position of the above-mentioned monitoring object. The above-mentioned ECG signal acquisition unit is used to collect the above-mentioned ECG signals of the above-mentioned monitoring object in real time, obtain the target waveform, input the multiple above-mentioned sub-waveforms corresponding to the above-mentioned target waveform into the above-mentioned first model, and obtain the above-mentioned first output result, and judge whether the above-mentioned target waveform is normal. When the above-mentioned first output result is abnormal, compare each of the above-mentioned electro-mapping signals collected at the second acquisition time with the four adjacent electro-mapping signals around it, obtain the comparison result, obtain the above-mentioned electrode conduction relationship based on the above-mentioned comparison result, compare it with the corresponding above-mentioned electrode correlation relationship, and obtain the abnormal electrode position closest to the excitation point and other abnormal electrode positions, and then obtain the abnormal position. When the above-mentioned first output result is normal, input the above-mentioned first normal waveform and the previous N time-continuous second normal waveforms into the above-mentioned second model, obtain the second output result. When the above-mentioned second output result is abnormal, obtain the above-mentioned target position and target electrode through the above-mentioned second output result, and by reducing the sampling period of the above-mentioned electro-mapping unit, thereby obtaining more refined and accurate electro-mapping signals of the above-mentioned target electrode, and also saving the electro-mapping signals of the above-mentioned target electrode and generating a dynamic image, which is marked in the above-mentioned electro-mapping image. Through the mutual cooperation of the above-mentioned technical solutions, more abundant monitoring information is obtained, which is convenient for users to quickly and accurately identify the abnormal position and take corresponding measures. Description of the Drawings

[0034] Figure 1 is a flowchart of a method for monitoring cardiac transport and thoracotomy based on high spatio-temporal resolution electro-mapping according to the present invention;

[0035] Figure 2 is a structural diagram of a system for monitoring cardiac transport and thoracotomy based on high spatio-temporal resolution electro-mapping according to the present invention. Detailed Embodiments

[0036] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0037] The present invention provides a method for monitoring cardiac transport and thoracotomy based on high spatio-temporal resolution electro-mapping, as Figure 1 shown, the method includes the following steps:

[0038] Step S1: The electrophysiological mapping unit collects the electrophysiological mapping signals of each mapping electrode in the monitoring object at a first preset period, and obtains the physiological information of the monitoring object through calculation and analysis based on the multiple electrophysiological mapping signals. An electrophysiological mapping image and the correlation between the mapping electrodes are generated based on the physiological information and the position information of each mapping electrode. Among them, the mapping electrodes are arranged on the epicardium of the monitoring object;

[0039] Specifically, during the transportation of a heart transplant or in a cardiothoracic open-chest surgery, it is necessary to accurately obtain the health status of the heart in real time. The ECG signal can continuously record the electrocardiogram of the patient and provide rich diagnostic information, but its specificity is insufficient, and sometimes it is impossible to accurately judge the type and location of arrhythmia. High-time-resolution recording can calculate various time-frequency parameters such as high-frequency amplitude, high-frequency relative power, QRS duration, depolarization duration, asymmetry, and morphological variation, so as to accurately map the electrical activity of the heart, effectively capture minute abnormalities, and locate the position of abnormal electrical activity. Through more detailed electrophysiological parameters, the health status of the heart during the transportation of a heart transplant or before and after a cardiothoracic open-chest surgery can be better evaluated, which is convenient for medical staff to adjust the treatment plan in time, thus better protecting the heart, reducing the postoperative risk of the patient, improving the postoperative survival rate and recovery quality of the patient, especially applicable to complex arrhythmia situations where the electrocardiogram cannot make a clear diagnosis, and there are problems of large data acquisition volume and calculation volume. Therefore, a method of combining the above high spatio-temporal resolution electrophysiological mapping technology and the ECG signal is adopted to improve the diagnostic accuracy and treatment effect. Specifically, the electrophysiological mapping unit obtains the electrophysiological mapping signals of each of the above mapping electrodes at the first preset period, that is, the potential information and conduction time of the above mapping electrodes, and obtains the above conduction information of the monitoring object through the electrophysiological mapping signals of the mapping electrodes. An electrophysiological mapping image is obtained according to the position information and the above conduction information of each of the above mapping electrodes, and the electrophysiological mapping signals of each of the above mapping electrodes are compared with the electrophysiological mapping signals of its adjacent adjacent mapping electrodes to obtain the correlation between the above mapping electrodes. At the same time, since the above mapping electrodes are located on the epicardium of the heart, the abnormal position can be directly located through the abnormal electrodes. Through the above technical solution, not only the electrophysiological mapping image can be obtained, but also the correlation between each of the above mapping electrodes and its adjacent adjacent mapping electrodes can be obtained, laying a foundation for accurately and quickly obtaining the abnormal state and abnormal position of the monitoring object.

[0040] Step S2: The ECG signal acquisition unit collects the ECG signal of the monitoring object in real time, obtains an analysis result by analyzing the ECG signal in real time, and when the analysis result is abnormal, performs denoising processing on the ECG signal and obtains a plurality of target waveforms based on the ECG signal;

[0041] Specifically, the above ECG signal acquisition unit collects the above ECG signal of the above monitoring object in real time, extracts the above first waveform, i.e., the R wave, in the above ECG signal, obtains the maximum value of the first waveform and the maximum fluctuation value of the corresponding reference potential interval, determines whether there is noise mixed in the above ECG signal based on the maximum fluctuation potential of the above reference voltage interval and the maximum value of the above R wave, performs denoising processing on the noise waveform, and uses the above ECG signal without noise and the above ECG signal after denoising processing as the above target waveform. Through the above technical solution, the accurate above target waveform can be obtained, laying a foundation for further accurately identifying abnormalities through the above target waveform.

[0042] Step S3: Divide each of the above target waveforms into multiple sub-waveforms according to the waveform type, input the sub-waveforms into the first model, obtain the first output result. When the first output result is abnormal, execute Step S4. When the first output result is normal, obtain the first normal waveform according to the first output result, and also use N consecutive second normal waveforms before the first normal waveform as a waveform sequence to input into the second model, and obtain the second output result;

[0043] Specifically, by dividing the above target waveform into multiple sub-waveforms according to the waveform type, inputting the multiple above sub-waveforms corresponding to the above target waveform into the above first model, and obtaining the above first output result, the above first model is a machine learning model trained with the sub-waveforms corresponding to the ECG signals of historical heart disease patients and the corresponding disease types, as well as the sub-waveforms corresponding to the ECG signals of normal people as training data. Through the above first output result, it can be determined whether the above target waveform is normal. Since only whether a person is sick can be judged through the ECG signal, but the specific location information of the disease cannot be determined, and the high spatio-temporal resolution electrical mapping technology can just well solve this problem. Therefore, when the above first output result is abnormal, by executing the above Step S4, the specific abnormal location of the disease and more disease information can be determined through the electrical mapping technology. When the above first output result is normal, some heart diseases that are not obvious in the ECG signal cannot be identified. Therefore, the above target waveform corresponding to the first output result is used as the first normal waveform, and the above first normal waveform and N consecutive second normal waveforms before are input into the above second model to obtain the second output result. Through the above technical solution, the monitoring status of the above monitoring object can be judged more accurately.

[0044] Step S4: When the first output result is abnormal, obtain all abnormal electrodes according to the first output result, and obtain the abnormal position based on the electrophysiological mapping signals of all the abnormal electrodes. When the second output result is abnormal, obtain the target electrode according to the second output result, and generate a dynamic image according to the electrophysiological mapping signal of the target electrode, and mark the abnormal position and the dynamic image in the electrophysiological mapping image.

[0045] Specifically, when the first output result is abnormal, each electrophysiological mapping signal collected at the second acquisition time closest to the first acquisition time of the target waveform corresponding to the first output result is used as an analysis object. Each electrophysiological mapping signal is compared with the four adjacent electrophysiological mapping signals around it to obtain a comparison result. Based on the comparison result, the electrode conduction relationship is obtained, and the electrode conduction relationship is compared with the corresponding electrode correlation relationship, and the position of the abnormal electrode closest to the excitation point is obtained. At the same time, the positions of other abnormal electrodes and the abnormal position are also obtained, so that medical staff can take corresponding measures according to the abnormal position and the electrophysiological mapping signals at the abnormal position. When the second output result is abnormal, the target position and the target electrode are obtained through the mapping position target position in the historical abnormal electrophysiological mapping data in the second output result, and by reducing the sampling period of the electrophysiological mapping unit, a more accurate and detailed electrophysiological mapping signal of the target electrode can be obtained. The electrophysiological mapping signal of the target electrode is also saved, and a dynamic image is generated and marked in the electrophysiological mapping image, so as to facilitate the user to identify the abnormal position and take corresponding measures. Through the above technical solution, when monitoring the transplantation, transportation or surgery of the monitored object, the real-time state of the monitored object can be obtained in real time, and when the monitored object is abnormal or about to be abnormal, the abnormal position can be accurately and quickly determined, and corresponding measures can be taken in time.

[0046] Further, the mapping electrodes in the electrophysiological mapping unit are at least 32 channels, and the distance between the electrode points in each mapping electrode is less than or equal to 4 mm.

[0047] Specifically, electrophysiological mapping data can use multiple mapping electrodes at the same time to record data of multiple parts such as the atrium and ventricle. High spatio-temporal resolution electrophysiological mapping data should have a high sampling rate so as to record high-frequency potential information. By recording high spatio-temporal resolution electrophysiological mapping records, subtle changes in cardiac electrical activity can be captured. Electrocardiogram data is a standard cardiac monitoring tool widely used clinically, with multiple leads, and each lead records the electrical activity of the heart from different angles, providing a comprehensive view of cardiac function.

[0048] Further, the step S1 includes:

[0049] Step S11: The electrophysiological mapping unit obtains the electrophysiological mapping signals of each mapping electrode at a first preset period, and obtains the conduction information of the monitoring object according to the set positions of each mapping electrode and the corresponding electrophysiological mapping signals. The conduction information includes excitation points, conduction directions, conduction velocities, depolarization dispersion and repolarization dispersion, conduction phases, and spectral characteristics.

[0050] Step S12: Based on the position information of each mapping electrode and the conduction information, the electrophysiological mapping image is obtained. Also, the electrophysiological mapping waveforms are extracted from each electrophysiological mapping signal according to heartbeats, the conduction feature points on each electrophysiological mapping waveform are calculated, and the waveform relationship and conduction relationship are obtained based on the time and space order of the electrophysiological mapping waveforms and the conduction feature points. Among them, the correlation relationship includes the waveform relationship and the conduction relationship.

[0051] Specifically, the electrophysiological mapping unit obtains the electrophysiological mapping signals of each mapping electrode at a first preset period, that is, the electrical signals and conduction times of the mapping electrodes. Since the electrophysiological mapping unit is a high spatio-temporal resolution electrophysiological mapping unit, the conduction information of the monitoring object is obtained through the electrophysiological mapping signals of the mapping electrodes. Among them, the excitation point in the conduction information is the starting point of the electrical wave conduction of the monitoring object. Also, the electrophysiological mapping image is obtained according to the position information of each mapping electrode and the conduction information. The electrophysiological mapping waveforms are extracted from each electrophysiological mapping signal according to heartbeats, the conduction feature points on each electrophysiological mapping waveform are calculated, and the waveform relationship and conduction relationship between adjacent electrode positions are obtained based on the time and space order of the electrophysiological mapping waveforms and the conduction feature points at each mapping electrode position. The waveform relationship includes the waveform similarity and difference of the waveforms. Through the above technical solution, not only can the electrophysiological mapping image be obtained, but also the correlation relationship between each mapping electrode and the surrounding adjacent mapping electrodes can be obtained, laying a foundation for accurately and quickly obtaining the abnormal state and abnormal position of the monitoring object.

[0052] Further, step S2 includes:

[0053] Step S21: The ECG signal acquisition unit continuously acquires the ECG signals of the monitoring object in real time, and extracts multiple first waveforms from the ECG signals. The first waveform is the R wave.

[0054] Step S22: Obtain the maximum value of each of the first waveforms and the maximum potential fluctuation value of the corresponding reference potential range, calculate the first difference between the maximum value and the first threshold, and the second difference between the maximum potential fluctuation value and the second threshold. When both the first difference and the second difference corresponding to each of the first waveforms are within the corresponding set ranges, the analysis result corresponding to the ECG signal is normal; otherwise, the analysis result corresponding to the ECG signal is abnormal. Take the ECG signal as a noise waveform, perform denoising processing on the noise waveform, and use the denoised ECG signal and the ECG signal when the analysis result is normal as the target waveform.

[0055] Specifically, the above-mentioned ECG signal acquisition unit is used to collect the above-mentioned ECG signal of the above-mentioned monitoring object in real time, and extract the above-mentioned first waveform, that is, the R wave, from the above-mentioned ECG signal, and obtain the maximum value of the first waveform and the maximum fluctuation value of the corresponding reference potential range. Since the reference potential range of the R wave is an equipotential range in the ECG signal waveform and the potential change is small within one cardiac cycle, therefore, it is judged whether there is noise mixed in the above-mentioned ECG signal by the maximum fluctuation potential of the above-mentioned reference voltage range and the maximum value of the above-mentioned R wave. When at least one of the above-mentioned first difference and the above-mentioned second difference is not within the corresponding set range, that is, when the above-mentioned first difference is greater than the first set range or the above-mentioned second difference is greater than the second set range, the above-mentioned ECG signal is a noise waveform, and denoising processing is performed on the above-mentioned noise waveform. When both the above-mentioned first difference and the above-mentioned second difference are within the corresponding set ranges, that is, when the above-mentioned first difference is less than or equal to the above-mentioned first set range and the above-mentioned second difference is less than or equal to the above-mentioned second set range, the above-mentioned ECG signal has no noise mixed in, and the above-mentioned ECG signal without noise and the above-mentioned ECG signal after denoising processing are used as the above-mentioned target waveform. Through the above technical solution, an accurate above-mentioned target waveform can be obtained, laying a foundation for further accurately identifying abnormalities through the above-mentioned target waveform.

[0056] Further, the step S3 includes:

[0057] Step S31: Divide each of the target waveforms into multiple sub-waveforms according to the waveform type, input the multiple sub-waveforms into the first model, and obtain the first output result. The first model is a machine learning model trained by first learning data. The first learning data includes the sub-waveforms corresponding to the ECG signals of historical heart disease patients stored in the database and the corresponding disease types, and also includes the sub-waveforms corresponding to the ECG signals of normal people.

[0058] Step S32: When the first output result is abnormal, obtain the disease types corresponding to the multiple input sub-waveforms through the first output result, and execute step S4;

[0059] Specifically, by dividing the above-mentioned target waveform into multiple sub-waveforms according to the waveform types, where the waveform types include P waves, T waves, and QRS waves, and inputting the multiple sub-waveforms corresponding to the target waveform into the first model, and obtaining the first output result. The first model is a machine learning model trained with the sub-waveforms corresponding to the ECG signals of historical heart disease patients and the corresponding disease types, as well as the sub-waveforms corresponding to the ECG signals of normal people as training data. By inputting the sub-waveforms of the target waveform into the first model, it can be determined whether the target waveform is normal based on the first output result, that is, whether the target waveform is a diseased ECG signal. When the target waveform is abnormal, the corresponding disease type is also output. Since only whether a disease is present can be judged through the ECG signal, but the specific location information of the disease cannot be determined, however, the high spatio-temporal resolution electrical mapping technology can just well solve this problem. Therefore, by performing the above step S4, and then determining the specific diseased location and more disease information through the electrical mapping technology, so that medical staff can take measures in time to ensure that the abnormal state of the monitored object will not deteriorate further.

[0060] Step S33: When the first output result is normal, obtain the target waveform corresponding to the multiple input sub-waveforms, and regard the target waveform as the first normal waveform, and obtain N time-continuous second normal waveforms before the first normal waveform. Input the first normal waveform and the N second normal waveforms as a waveform sequence into the second model, and obtain the second output result, where the second model is a machine learning model trained with second learning data.

[0061] Specifically, although the above first output result shows that the above target waveform is normal, since the ECG signal at a single moment can reflect limited information of the above monitoring object and cannot accurately identify some diseases that are not obvious in the ECG signal, when the above first output result is normal, the corresponding above target waveform is used as the first normal waveform, and N consecutive second normal waveforms before the above first normal waveform are obtained. The above first normal waveform and N consecutive second normal waveforms are used as the above waveform sequence, and the above waveform sequence is input into the above second model to obtain a second output result. Furthermore, the health status of the above monitoring object is more accurately judged through the change information of each target waveform in the above waveform sequence over time. Among them, the above second model is a machine learning model trained with second learning data, and the above second learning data is historical abnormal electro-mapping data and a historical ECG signal sequence within a preset time period before the acquisition time of the above historical abnormal electro-mapping data, and also includes the disease type corresponding to the above historical abnormal electro-mapping data. When the above second output result is normal, it indicates that the above monitoring object is normal. When the above second output result is abnormal, it indicates that the above monitoring object may be abnormal in the future time period. Among them, when the above second output result is abnormal, the above second output result also outputs the corresponding historical abnormal mapping data and disease type. Through the above technical solution, when the above first output result is normal, the future monitoring status of the above monitoring object can be more accurately predicted through the above waveform sequence.

[0062] Further, the second learning data includes historical abnormal electro-mapping data stored in a database and a historical ECG signal sequence within a preset time period before the acquisition time of the historical abnormal electro-mapping data, and also includes the disease type corresponding to the historical abnormal electro-mapping data.

[0063] Specifically, the above second model trained by using the above second learning data can accurately predict the health status of the monitoring object in the future time period, and medical staff can understand the status of the above monitoring object in the future time period in advance, ensuring that the status parameters of the monitoring object can be accurately and timely obtained during transplantation transfer or cardiothoracic surgery.

[0064] Further, the step S4 includes:

[0065] Step S41: When the first output result is abnormal, obtain the first acquisition time of the target waveform corresponding to the first output result, obtain the electro-mapping signals collected by each of the mapping electrodes at the second acquisition time closest to the first acquisition time, extract electro-mapping waveforms from each of the electro-mapping signals according to heartbeats, calculate conduction feature points of each electro-mapping waveform, obtain the electrode conduction relationship based on the time and space sequence of the electro-mapping waveforms and conduction feature points of the electrode and other electrodes around the electrode, compare the electrode conduction relationship with the relevant relationship of the corresponding mapping electrode. When the two are consistent, the mapping electrode is a normal electrode; when the two are inconsistent, the mapping electrode is an abnormal electrode, and obtain the position of the abnormal electrode closest to the excitation point position and the positions of other abnormal electrodes. Generate an abnormal position based on the abnormal electrode position and the positions of other abnormal electrodes and mark it in the electro-mapping image;

[0066] Specifically, when the first output result is abnormal, it indicates that the monitored object may be in an abnormal state, that is, a disease state. It is necessary to obtain richer abnormal information through the electro-mapping unit in a timely manner, so as to facilitate the medical staff to accurately and timely treat. Since the electro-mapping signals collected closer to the first acquisition time of the target waveform corresponding to the first output result are more accurate, each of the electro-mapping signals collected at the second acquisition time is used as an analysis object. Among them, the second acquisition time can be before the first acquisition time or after the first acquisition time. Compare each of the electro-mapping signals with other adjacent electro-mapping signals around it to obtain a comparison result, and obtain the electrode conduction relationship based on the comparison result. The electrode conduction relationship includes the conduction sequence and the relationship of conduction speed between the electrode and the four adjacent electrodes around it, and compare the electrode conduction relationship with the corresponding relevant relationship of the electrode. When the two are consistent, the electrode is normal; when the two are inconsistent, the electrode is an abnormal electrode. Since the electrical conduction of the monitored object starts from the excitation point, the conduction at the position close to the excitation point is inaccurate, and the subsequent ones are all abnormal. Therefore, first obtain the position of the abnormal electrode closest to the excitation point, and at the same time obtain the positions of other abnormal electrodes. The positions of other abnormal electrodes are related to the position of the abnormal electrode and are marked as the first abnormal position in the electro-mapping image, so that the medical staff can take corresponding measures according to the abnormal position and the electro-mapping signals at the abnormal position. Through the above technical solution, the abnormal position and the electro-mapping signals at the abnormal position can be quickly and accurately obtained, which is convenient for the medical staff to take corresponding measures in a timely manner to prevent the state of the monitored object from deteriorating further.

[0067] Step S42: When the second output result is abnormal, obtain the target position in the electrogram image corresponding to the historical abnormal electrogram measurement data in the second output result, obtain the corresponding target electrode based on the target position, reduce the acquisition period of the electrogram measurement unit for the target electrode to a second preset period, save the electrogram measurement signals collected by the target electrode, generate a dynamic image within the corresponding range of the electrogram image from the saved electrogram measurement signals of the target electrode, and mark it in the electrogram image;

[0068] Step S43: The user takes corresponding measures based on the abnormal position and the electrogram measurement signals at the abnormal position, and recognizes the abnormality in real time according to the dynamic image.

[0069] Specifically, when the above second output result is abnormal, it indicates that the above monitoring object may be abnormal next. Since the above historical abnormal electrogram measurement data includes not only abnormal data but also the mapping positions corresponding to the abnormal data, the target position in the above electrogram image is obtained through the mapping position in the historical abnormal electrogram measurement data in the above second output result, and the target electrode at the above target position is also obtained based on the above target position. By reducing the sampling period of the above electrogram measurement unit, more precise and accurate electrogram measurement signals of the above target electrode are obtained, and the electrogram measurement signals of the above target electrode are saved. All the saved electrogram measurement signals of the above target electrode are used as an electrogram measurement signal sequence, and the above dynamic image is generated in the above electrogram image based on the electrogram measurement signal sequence, thereby facilitating the user to identify the abnormal position and take corresponding measures. Through the above technical solution, when performing transplantation and transportation of the monitoring object or thoracotomy, the real-time state of the monitoring object can be obtained in real time, and when the above monitoring object is abnormal or about to be abnormal, the abnormal position can be accurately and quickly determined, and measures can be taken in time.

[0070] Further, step S4 further includes:

[0071] When the second output result is normal, increase the acquisition period of the electrogram measurement unit to a third preset period, where the third preset period is less than the first preset period, and the first preset period is less than the second preset period.

[0072] Specifically, although the high spatio-temporal resolution electrical mapping technology can more accurately monitor the state of the above-mentioned monitoring object and obtain richer monitoring information of the monitoring object, however, the data volume of the high spatio-temporal resolution electrical mapping technology is relatively large, the storage occupies more resources, the data transmission pressure is also relatively large, and the power consumption of analog-to-digital conversion is also relatively large during acquisition. It is possible that the temperature of the device will rise due to heat generation. Therefore, in the initial state, the above electrical mapping unit collects electrical mapping signals at a first preset period and creates an electrical mapping image. When the second output result is normal, it indicates that the probability of the above-mentioned monitoring object being abnormal is small. In order to save the above storage resource occupancy and reduce the device temperature rise, therefore, when the probability of the monitoring object being abnormal is small, the above acquisition period is increased to the third preset period.

[0073] Further, the mapping electrode has flexibility and biocompatibility.

[0074] The present invention also provides a cardiac transport and open-chest surgery monitoring system based on high spatio-temporal resolution electrical mapping. The system is used to implement the above method, as Figure 2 shown, the system includes:

[0075] An electrical mapping unit, configured to collect electrical mapping signals of each mapping electrode in the monitoring object at a first preset period, and obtain physiological information of the monitoring object through calculation and analysis based on a plurality of the electrical mapping signals. Based on the physiological information and the position information of each mapping electrode, an electrical meter mapping image and the correlation between the mapping electrodes are generated, wherein the mapping electrodes are arranged on the outer membrane of the monitoring object;

[0076] An ECG signal acquisition unit, configured to collect the ECG signal of the monitoring object in real time, obtain an analysis result by analyzing the ECG signal in real time, when the analysis result is abnormal, perform denoising processing on the ECG signal, and obtain a plurality of target waveforms based on the ECG signal;

[0077] A discrimination and prediction unit, configured to divide each of the target waveforms into a plurality of sub-waveforms according to the waveform type, input the sub-waveforms into a first model to obtain a first output result. When the first output result is abnormal, execute step S4. When the first output result is normal, obtain a first normal waveform according to the first output result. The first normal waveform and the N consecutive second normal waveforms before it in time are used as a waveform sequence and input into a second model, and a second output result is obtained;

[0078] A display generation unit, configured to, when the first output result is abnormal, obtain all abnormal electrodes according to the first output result, obtain an abnormal position based on the electrical mapping signals of all the abnormal electrodes, when the second output result is abnormal, obtain target electrodes according to the second output result, generate a dynamic image according to the electrical mapping signals of the target electrodes, and label the abnormal position and the dynamic image in the mapping image.

[0079] In summary, the present invention obtains the electrical mapping signals of each of the above-mentioned mapping electrodes at a first preset period through the above-mentioned electrical mapping unit, obtains the above-mentioned electrical mapping image according to the position information and the above-mentioned conduction information of each of the above-mentioned mapping electrodes, and obtains the correlation between the above-mentioned mapping electrodes. Among them, the mapping electrodes are arranged on the epicardium of the heart, the above-mentioned electrical mapping unit is a high-resolution multi-channel mapping, and can monitor each position of the heart at the same time, laying a foundation for accurately and quickly obtaining the abnormal state and abnormal position of the above-mentioned monitoring object. The above-mentioned ECG signal acquisition unit collects the above-mentioned ECG signals of the above-mentioned monitoring object in real time, obtains a target waveform, inputs a plurality of the above-mentioned sub-waveforms corresponding to the above-mentioned target waveform into the above-mentioned first model, and obtains the above-mentioned first output result, and determines whether the above-mentioned target waveform is normal. When the above-mentioned first output result is abnormal, compare each of the above-mentioned electrical mapping signals collected at the second acquisition time with the four adjacent electrical mapping signals around it, obtain a comparison result, obtain the above-mentioned electrode conduction relationship based on the above-mentioned comparison result, compare it with the corresponding above-mentioned electrode correlation, and obtain the position of the abnormal electrode closest to the excitation point and the positions of other abnormal electrodes, and then obtain the abnormal position. When the above-mentioned first output result is normal, input the above-mentioned first normal waveform and the previous N consecutive second normal waveforms at different times into the above-mentioned second model, obtain a second output result. When the above-mentioned second output result is abnormal, obtain the above-mentioned target position and target electrodes through the above-mentioned second output result, and by reducing the sampling period of the above-mentioned electrical mapping unit, thereby obtaining more precise and accurate electrical mapping signals of the above-mentioned target electrodes, and also saving the electrical mapping signals of the above-mentioned target electrodes, and generating a dynamic image, which is labeled in the above-mentioned electrical mapping image. Through the mutual cooperation of the above-mentioned technical solutions, more abundant monitoring information is obtained, so as to facilitate the user to quickly and accurately identify the abnormal position and take corresponding measures.

[0080] The technical features of the above-mentioned embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

[0081] The above-mentioned embodiments only represent several implementation manners of the present invention. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent for the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent for the present invention shall be subject to the appended claims.

[0082] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for monitoring cardiac transit and thoracotomy based on high spatiotemporal resolution electrical mapping, characterized in that: The method comprises the following steps: Step S1: collecting electrical mapping signals of each mapping electrode in the monitored object at a first preset period through an electrical mapping unit, and obtaining physiological information of the monitored object through calculation and analysis based on a plurality of the electrical mapping signals, and generating an electrical mapping image and a correlation between the mapping electrodes based on the physiological information and the position information of each mapping electrode, wherein the mapping electrodes are arranged on the outer membrane of the monitored object; Step S2: collecting the ECG signal of the monitored object in real time through the ECG signal collection unit, obtaining an analysis result by real-time analysis of the ECG signal, and when the analysis result is abnormal, performing denoising on the ECG signal, and obtaining a plurality of target waveforms based on the ECG signal; Step S3: dividing each of the target waveforms into a plurality of sub-waveforms according to the waveform type, and inputting the sub-waveforms into the first model to obtain a first output result. When the first output result is abnormal, executing step S4, when the first output result is normal, obtaining a first normal waveform according to the first output result, and further inputting N second normal waveforms that are time-continuous before the first normal waveform as a waveform sequence into the second model to obtain a second output result; Step S4: When the first output result is abnormal, all abnormal electrodes are obtained according to the first output result, and the abnormal position is obtained based on the electrical mapping signals of all the abnormal electrodes; when the second output result is abnormal, the target electrode is obtained according to the second output result, and a dynamic image is generated according to the electrical mapping signals of the target electrode, and the abnormal position and the dynamic image are marked in the electrical mapping image.

2. The method according to claim 1, characterized in that The mapping electrodes in the electrical mapping unit have at least 32 channels, and the distance between electrode points in each of the mapping electrodes is less than or equal to 4 mm.

3. The method according to claim 1, characterized in that The step S1 comprises: Step S11: acquiring the electrical mapping signal of each mapping electrode at a first preset period through the electrical mapping unit, and acquiring the conduction information of the monitored object according to the setting position of each mapping electrode and the corresponding electrical mapping signal, wherein the conduction information includes the excitation point, conduction direction, conduction velocity, depolarization dispersion and repolarization dispersion, conduction phase and spectral characteristics; Step S12: acquiring the electrical mapping image based on the position information of each mapping electrode and the conduction information, and extracting the electrical mapping waveform from each electrical mapping signal according to the heartbeat, calculating the conduction feature points on each electrical mapping waveform, and acquiring the waveform relationship and the conduction relationship based on the time-space sequence of the electrical mapping waveform and the conduction feature points, wherein the correlation relationship includes the waveform relationship and the conduction relationship.

4. The method according to claim 1, characterized in that The step S2 comprises: Step S21: collecting the ECG signal of the monitored object in real time through the ECG signal collection unit, and extracting a plurality of first waveforms in the ECG signal, wherein the first waveforms are R waves and other ECG waveform features; Step S22: Obtain the maximum value of each of the first waveforms and the maximum potential fluctuation value of the corresponding reference potential interval, calculate the first difference between the maximum value and the first threshold, and the second difference between the maximum potential fluctuation value and the second threshold. When the first difference and the second difference corresponding to each of the first waveforms are within the corresponding set range, the analysis result corresponding to the ECG signal is normal. Otherwise, the analysis result corresponding to the ECG signal is abnormal. The ECG signal is taken as a noise waveform, the noise waveform is denoised, and the denoised ECG signal and the ECG signal when the analysis result is normal are taken as the target waveform.

5. The method according to claim 1, characterized in that The step S3 comprises: Step S31: dividing each of the target waveforms into a plurality of sub-waveforms according to the waveform type, and inputting the plurality of sub-waveforms into the first model, and obtaining the first output result, wherein the first model is a machine learning model trained by first learning data, and the first learning data includes sub-waveforms and corresponding disease types corresponding to ECG signals of historical heart disease patients stored in a database, and also includes sub-waveforms corresponding to ECG signals of normal people; Step S32: when the first output result is abnormal, obtaining the disease types corresponding to the input multiple sub-waveforms through the first output result, and executing the step S4; Step S33: When the first output result is normal, obtain the input multiple sub-waveforms corresponding to the target waveform, and use the target waveform as the first normal waveform, and obtain N time-continuous second normal waveforms before the first normal waveform, input the first normal waveform and N second normal waveforms as a waveform sequence into the second model, and obtain a second output result, wherein the second model is a machine learning model trained with second learning data.

6. The method according to claim 5, characterized in that The second learning data includes historical abnormal electrical mapping data stored in a database and historical ECG signal sequences within a preset time period before the acquisition time of the historical abnormal electrical mapping data, and also includes disease types corresponding to the historical abnormal electrical mapping data.

7. The method according to claim 1, characterized in that The step S4 comprises the following steps: Step S41: when the first output result is abnormal, obtain the first acquisition time of the target waveform corresponding to the first output result, obtain the electrical mapping signal collected by each mapping electrode at the second acquisition time closest to the first acquisition time, extract the electrical mapping waveform of each electrical mapping signal according to the heartbeat, calculate the conduction feature points on each electrical mapping waveform, obtain the electrode conduction relationship based on the electrical mapping waveforms and conduction feature points of the electrode and other electrodes around the electrode in time and space sequence, compare the electrode conduction relationship with the correlation relationship corresponding to the mapping electrode, when the two are consistent, the mapping electrode is a normal electrode, when the two are inconsistent, the mapping electrode is an abnormal electrode, and obtain the abnormal electrode position closest to the excitement point position and other abnormal electrode positions, generate an abnormal position based on the abnormal electrode position and the other abnormal electrode positions and mark it in the electrical mapping image; Step S42: when the second output result is abnormal, the target position corresponding to the electrical mapping image is obtained according to the historical abnormal electrical mapping data in the second output result, the corresponding target electrode is obtained based on the target position, the acquisition period of the target electrode by the electrical mapping unit is reduced to a second preset period, and the electrical mapping signal acquired by the target electrode is saved, and a dynamic image is generated within the corresponding range of the electrical mapping image with the saved electrical mapping signal of the target electrode, and the image is marked in the electrical mapping image; Step S43: The user takes corresponding measures based on the abnormal position and the electrical mapping signal at the abnormal position, and identifies the abnormality in real time according to the dynamic image.

8. The method according to claim 7, characterized in that The step S4 further comprises: When the second output result is normal, the acquisition period of the electrical labeling unit is increased to a third preset period, wherein the third preset period is smaller than the first preset period, and the first preset period is smaller than the second preset period.

9. The method according to claim 1, characterized in that: The mapping electrode is flexible and biocompatible.

10. A cardiac transport and thoracotomy monitoring system based on high spatiotemporal resolution electrical mapping, the system being used to implement the method according to any one of claims 1 to 8, characterized in that: The system comprises: An electrical mapping unit, used for collecting electrical mapping signals of each mapping electrode in the monitored object at a first preset period, and obtaining physiological information of the monitored object through calculation and analysis based on a plurality of the electrical mapping signals, and generating an electrical meter measurement image and a correlation between the mapping electrodes based on the physiological information and the position information of each mapping electrode, wherein the mapping electrodes are arranged on the outer membrane of the monitored object; An ECG signal acquisition unit is used to acquire ECG signals of a monitored object in real time, obtain analysis results by analyzing the ECG signals in real time, perform denoising on the ECG signals when the analysis results are abnormal, and obtain multiple target waveforms based on the ECG signals; a discrimination prediction unit, configured to divide each of the target waveforms into a plurality of sub-waveforms according to the waveform type, and input the sub-waveforms into a first model to obtain a first output result, and when the first output result is abnormal, execute step S4, and when the first output result is normal, obtain a first normal waveform according to the first output result, and further input N second normal waveforms that are continuous in time before the first normal waveform as a waveform sequence into a second model to obtain a second output result; A display generation unit is used to obtain all abnormal electrodes according to the first output result when the first output result is abnormal, and obtain the abnormal position based on the electrical mapping signals of all the abnormal electrodes; when the second output result is abnormal, obtain the target electrode according to the second output result, and generate a dynamic image according to the electrical mapping signals of the target electrode, and mark the abnormal position and the dynamic image in the mapping image.

Images