Method and system for synchronously analyzing long-range blood pressure and electrocardio

Abstract

The invention provides a method and system for synchronously analyzing long-range blood pressure and electrocardio. The method comprises the steps that S1, signal values collected by different devices are preprocessed; s2, automatically identifying and extracting hidden features of the electrocardiosignal by using the trained deep learning model to obtain an analysis result of the electrocardiosignal; s3, various blood pressure parameters are calculated according to the blood pressure signals, and analysis results of the blood pressure signals are obtained; and S4, synchronously outputting an analysis result of the electrocardiosignal and an analysis result of the blood pressure signal. The blood pressure and electrocardiosignals collected by different devices are resampled to the sampling points with the same number, so that the consistency of the signal length is ensured, and the data processing flow is simplified; by adopting the structure of logical judgment and the deep neural network, the problem of synchronous analysis of long-range blood pressure and electrocardio is solved, high-sensitivity detection of hypertension and arrhythmia diseases is realized, and the workload of doctors is reduced.

Description

Technical Field

[0001] The present invention relates to the field of body parameter testing, and in particular to a method and system for synchronously analyzing long-term blood pressure and electrocardiogram. Background Art

[0002] Long-term blood pressure monitoring (Ambulatory Blood Pressure Monitoring, ABPM) refers to a method of regularly measuring blood pressure for 24 hours or longer in daily life environment through a portable blood pressure monitoring device. Long-term electrocardiogram (Holter Monitoring) refers to a method of recording electrocardiogram data for 24 hours or longer through a portable electrocardiogram monitoring device.

[0003] Traditional long-term blood pressure and ECG monitoring methods mainly rely on manual interpretation, which has the problems of strong subjectivity, low efficiency, and difficulty in achieving continuous monitoring. In recent years, with the development of artificial intelligence and deep learning technologies, it has become possible to use these technologies to automatically process and analyze physiological signals. However, existing automatic analysis systems can often only process ECG or blood pressure data separately, and lack the ability to analyze both simultaneously, resulting in incomplete and inaccurate evaluation results.

[0004] In a Chinese patent document with publication number CN117503086A, a blood pressure monitoring method based on dynamic blood pressure monitoring data feature extraction is disclosed, including the following steps: obtaining dynamic blood pressure monitoring data; extracting multi-dimensional features of the dynamic blood pressure monitoring data, the multi-dimensional features including time domain features, frequency domain features, nonlinear dynamic features and correlation features; constructing a classification model, training the classification model based on the multi-dimensional features; and classifying and monitoring the dynamic blood pressure monitoring data based on the trained classification model. Although this patent document simplifies the amount of data generated by dynamic blood pressure monitoring, it still performs a separate test on blood pressure and still cannot solve the above-mentioned problem. Summary of the invention

[0005] In view of the defects in the prior art, the object of the present invention is to provide a method and system for synchronously analyzing long-term blood pressure and electrocardiogram.

[0006] A method for synchronously analyzing long-term blood pressure and electrocardiogram provided by the present invention comprises:

[0007] Step S1: pre-processing the signal values ​​collected by different devices;

[0008] The signal values ​​include blood pressure signals and electrocardiogram signals;

[0009] The preprocessing includes resampling the signal value and unifying the number of sampling points;

[0010] Step S2: using the trained deep learning model to automatically identify and extract hidden features of the ECG signal to obtain analysis results of the ECG signal;

[0011] Step S3: Calculate various blood pressure parameters according to the blood pressure signal to obtain the analysis result of the blood pressure signal;

[0012] The blood pressure parameters include blood pressure load value and nocturnal blood pressure drop percentage;

[0013] Step S4: Synchronously output the analysis results of the ECG signal and the analysis results of the blood pressure signal.

[0014] Preferably, step S2 includes using CNN to form a feature extractor, automatically mining hidden features of the ECG or heartbeat segments in end-to-end learning, obtaining a valid feature expression of the input data, and obtaining an analysis result of the ECG signal.

[0015] Preferably, step S3 includes the following sub-steps:

[0016] Step S3.1: Recording dynamic blood pressure for 24 hours, so that the blood pressure measurement values ​​in different time periods reach corresponding preset numbers; the blood pressure measurement values ​​in different time periods include daytime measurement values ​​and nighttime measurement values;

[0017] Step S3.2: Calculate the blood pressure load value according to the percentage of the number of times the blood pressure measurement value is greater than the preset value to the total number of blood pressure measurement values.

[0018] Preferably, the calculation formula for the percentage of nocturnal blood pressure decrease includes:

[0019]

[0020] Preferably, the analysis result of the blood pressure signal includes:

[0021] Judging blood pressure patterns based on the percentage of nighttime blood pressure drop:

[0022] If the percentage of nocturnal blood pressure drop is 10% to 20%, the blood pressure pattern is a dipper;

[0023] If the percentage of nocturnal blood pressure drop is less than 10%, the blood pressure pattern is non-dipper;

[0024] If the percentage of nocturnal blood pressure drop is greater than 20%, the blood pressure pattern is super-dipper;

[0025] If the percentage of nocturnal blood pressure drop is less than 0%, the blood pressure pattern is a reverse dipper;

[0026] Judging whether it is hypertension based on the blood pressure load value:

[0027] If the blood pressure load value is between 5% and 15%, the blood pressure is normal; if the blood pressure load value is greater than 50%, it is hypertension.

[0028] A system for synchronously analyzing long-term blood pressure and electrocardiogram provided by the present invention comprises:

[0029] Module M1: pre-process the signal values ​​collected by different devices;

[0030] The signal values ​​include blood pressure signals and electrocardiogram signals;

[0031] The preprocessing includes resampling the signal value and unifying the number of sampling points;

[0032] Module M2: Use the trained deep learning model to automatically identify and extract the hidden features of the ECG signal to obtain the analysis results of the ECG signal;

[0033] Module M3: Calculate various blood pressure parameters according to the blood pressure signal to obtain the analysis result of the blood pressure signal;

[0034] The blood pressure parameters include blood pressure load value and nocturnal blood pressure drop percentage;

[0035] Module M4: Synchronously outputs the analysis results of the ECG signal and the blood pressure signal.

[0036] Preferably, the module M2 includes using CNN to form a feature extractor, automatically mining hidden features of ECG or heartbeat segments in end-to-end learning, obtaining effective feature expressions of input data, and obtaining analysis results of ECG signals.

[0037] Preferably, the module M3 includes the following submodules:

[0038] Module M3.1: Recording dynamic blood pressure for 24 hours, so that the blood pressure measurement values ​​in different time periods reach the corresponding preset numbers; the blood pressure measurement values ​​in different time periods include daytime measurement values ​​and nighttime measurement values;

[0039] Module M3.2: Calculate the blood pressure load value based on the percentage of the number of times the blood pressure measurement value is greater than the preset value to the total number of blood pressure measurement values.

[0040] Preferably, the calculation formula for the percentage of nocturnal blood pressure decrease includes:

[0041]

[0042] Preferably, the analysis result of the blood pressure signal includes:

[0043] Judging blood pressure patterns based on the percentage of nighttime blood pressure drop:

[0044] If the percentage of nocturnal blood pressure drop is 10% to 20%, the blood pressure pattern is a dipper;

[0045] If the percentage of nocturnal blood pressure drop is less than 10%, the blood pressure pattern is non-dipper;

[0046] If the percentage of nocturnal blood pressure drop is greater than 20%, the blood pressure pattern is super-dipper;

[0047] If the percentage of nocturnal blood pressure drop is less than 0%, the blood pressure pattern is a reverse dipper;

[0048] Judging whether it is hypertension based on the blood pressure load value:

[0049] If the blood pressure load value is between 5% and 15%, the blood pressure is normal; if the blood pressure load value is greater than 50%, it is hypertension.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] 1. The present invention adopts a convolutional neural network ECG deep feature extractor, which can automatically mine the hidden features of ECG or heartbeat segments, obtain the effective feature expression of input data and output ECG diagnosis results, which greatly improves the accuracy and reliability of ECG analysis and reduces human errors.

[0052] 2. The present invention develops a statistical long-term blood pressure judgment method, which can automatically output the user's blood pressure health status based on the 24-hour dynamic blood pressure monitoring results. It not only takes into account the absolute value of blood pressure, but also evaluates the diurnal variation pattern of blood pressure, providing a more comprehensive blood pressure health assessment.

[0053] 3. The present invention ensures the consistency of signal length and simplifies the data processing process by resampling the blood pressure and ECG signals collected by different devices to a uniform number of sampling points. At the same time, by adopting the structure of logical judgment and deep neural network, it solves the problem of synchronous analysis of long-term blood pressure and ECG, realizes high-sensitivity detection of hypertension and arrhythmia diseases, reduces the workload of doctors, and improves the user experience and satisfaction.

[0054] Other beneficial effects of the present invention will be explained in the specific implementation manner through the introduction of specific technical features and technical solutions. Through the introduction of these technical features and technical solutions, those skilled in the art should be able to understand the beneficial technical effects brought about by the technical features and technical solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0056] Figure 1The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0057] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0058] Reference Figure 1 As shown, a method for synchronously analyzing long-term blood pressure and electrocardiogram comprises:

[0059] Step S1: pre-processing the signal values ​​collected by different devices;

[0060] The signal values ​​include blood pressure signals and electrocardiogram signals;

[0061] The preprocessing includes resampling the signal value and unifying the number of sampling points;

[0062] Step S2: Use the trained deep learning model to automatically identify and extract hidden features of the ECG signal to obtain the analysis results of the ECG signal; for example, use CNN to form a feature extractor to automatically mine the hidden features of the ECG or heartbeat segments in end-to-end learning, obtain the effective feature expression of the input data, and obtain the analysis results of the ECG signal.

[0063] Step S3: Calculate various blood pressure parameters according to the blood pressure signal to obtain the analysis result of the blood pressure signal;

[0064] The blood pressure parameters include blood pressure load value and nocturnal blood pressure drop percentage;

[0065] Step S3.1: Recording dynamic blood pressure for 24 hours, so that the blood pressure measurement values ​​in different time periods reach corresponding preset numbers; the blood pressure measurement values ​​in different time periods include daytime measurement values ​​and nighttime measurement values;

[0066] Step S3.2: Calculate the blood pressure load value according to the percentage of the number of times the blood pressure measurement value is greater than the preset value to the total number of blood pressure measurement values.

[0067] The formula for calculating the percentage of nocturnal blood pressure drop includes:

[0068]

[0069] Judging blood pressure patterns based on the percentage of nighttime blood pressure drop:

[0070] If the percentage of nocturnal blood pressure drop is 10% to 20%, the blood pressure pattern is a dipper;

[0071] If the percentage of nocturnal blood pressure drop is less than 10%, the blood pressure pattern is non-dipper;

[0072] If the percentage of nocturnal blood pressure drop is greater than 20%, the blood pressure pattern is super-dipper;

[0073] If the percentage of nocturnal blood pressure drop is less than 0%, the blood pressure pattern is a reverse dipper;

[0074] Judging whether it is hypertension based on the blood pressure load value:

[0075] If the blood pressure load value is between 5% and 15%, the blood pressure is normal; if the blood pressure load value is greater than 50%, it is hypertension.

[0076] Step S4: Synchronously output the analysis results of the ECG signal and the analysis results of the blood pressure signal.

[0077] The present invention ensures the consistency of signal length and simplifies the data processing process by resampling the blood pressure and ECG signals collected by different devices to a uniform number of sampling points. At the same time, by adopting the structure of logical judgment and deep neural network, it solves the problem of synchronous analysis of long-term blood pressure and ECG, realizes high-sensitivity detection of hypertension and arrhythmia diseases, reduces the workload of doctors, and improves the user experience and satisfaction.

[0078] The above is a basic embodiment of the present invention. The technical solution of the present invention is further described below through a preferred embodiment.

[0079] Example 1

[0080] Reference Figure 1 As shown, a method for synchronously analyzing long-term blood pressure and electrocardiogram, the steps are as follows:

[0081] Step 1: Since the sampling frequencies of different devices are different, in order to ensure the consistency of signal length, all records are resampled to 4096 sampling points;

[0082] Step 2: The feature extractor is composed of CNN, which can automatically mine the hidden features of ECG or heartbeat segments in end-to-end learning, obtain the effective feature expression of the input data, and obtain the ECG diagnosis result:

[0083] X i+1 =F(X i ,W i ,b i ) (1)

[0084]

[0085] Where X is the input, F(·) is a nonlinear function used to extract deep features containing multiple convolutional layers of different sizes. bi and Wi are the bias and weight parameters of the i-th layer, respectively.

[0086] Step 3: Record 24-hour ambulatory blood pressure;

[0087] Valid blood pressure readings during a 24-hour period should account for more than 70% of the total monitoring times, and there should be at least one blood pressure reading every hour, with daytime blood pressure readings ≥20 and nighttime blood pressure readings ≥7.

[0088] Step 4: Calculate the blood pressure load, which is the percentage of blood pressure readings greater than normal to the total number of measurements.

[0089] In this embodiment, the percentage of the number of times when the daytime average value is greater than 140 / 90 mmHg and the nighttime average value is greater than 120 / 80 mmHg in the total number of measurements is calculated.

[0090] Step 5: Calculate the percentage of nocturnal blood pressure drop using the following formula:

[0091] [(Daytime average value - Nighttime average value) / Daytime average value] × 100%.

[0092] Step 6: Determine the blood pressure pattern based on the percentage of blood pressure drop at night. Specifically:

[0093] If the percentage of nocturnal blood pressure drop is 10% to 20%, the blood pressure pattern is judged to be dipper type;

[0094] If the percentage of nocturnal blood pressure drop is less than 10%, the blood pressure pattern is judged to be non-dipper type;

[0095] If the percentage of nocturnal blood pressure drop is greater than 20%, the blood pressure pattern is judged to be super-dipper type;

[0096] If the percentage of nocturnal blood pressure drop is less than 0%, the blood pressure pattern is judged to be a reverse dipper.

[0097] If the blood pressure load value is between 5% and 15%, the blood pressure is judged to be normal; if the blood pressure load value is >50%, it is judged to be hypertension.

[0098] Step 7: Based on the blood pressure and ECG data, output the blood pressure and ECG conclusions simultaneously.

[0099] The present invention adopts a convolutional neural network ECG deep feature extractor, which can automatically mine the hidden features of ECG or heartbeat segments, obtain the effective feature expression of input data and output ECG diagnosis results, which greatly improves the accuracy and reliability of ECG analysis and reduces human errors.

[0100] The present invention also provides a system for synchronously analyzing long-term blood pressure and electrocardiogram. The system for synchronously analyzing long-term blood pressure and electrocardiogram can be implemented by executing the process steps of the method for synchronously analyzing long-term blood pressure and electrocardiogram, that is, technical personnel in this field can understand the method for synchronously analyzing long-term blood pressure and electrocardiogram as a preferred implementation of the system for synchronously analyzing long-term blood pressure and electrocardiogram.

[0101] Specifically, a system for synchronously analyzing long-term blood pressure and electrocardiogram includes:

[0102] Module M1: pre-process the signal values ​​collected by different devices;

[0103] The signal values ​​include blood pressure signals and electrocardiogram signals;

[0104] The preprocessing includes resampling the signal value and unifying the number of sampling points;

[0105] Module M2: Use the trained deep learning model to automatically identify and extract the hidden features of the ECG signal to obtain the analysis results of the ECG signal;

[0106] Module M3: Calculate various blood pressure parameters according to the blood pressure signal to obtain the analysis result of the blood pressure signal;

[0107] The blood pressure parameters include blood pressure load value and nocturnal blood pressure drop percentage;

[0108] Module M4: Synchronously outputs the analysis results of the ECG signal and the blood pressure signal.

[0109] The module M2 includes using CNN to form a feature extractor, automatically mining the hidden features of the ECG or heartbeat segments in end-to-end learning, obtaining the effective feature expression of the input data, and obtaining the analysis results of the ECG signal.

[0110] The module M3 includes the following submodules:

[0111] Module M3.1: Recording dynamic blood pressure for 24 hours, so that the blood pressure measurement values ​​in different time periods reach the corresponding preset numbers; the blood pressure measurement values ​​in different time periods include daytime measurement values ​​and nighttime measurement values;

[0112] Module M3.2: Calculate the blood pressure load value based on the percentage of the number of times the blood pressure measurement value is greater than the preset value to the total number of blood pressure measurement values.

[0113] The calculation formula of the nighttime blood pressure drop percentage includes:

[0114]

[0115] The analysis results of the blood pressure signal include:

[0116] Judging blood pressure patterns based on the percentage of nighttime blood pressure drop:

[0117] If the percentage of nocturnal blood pressure drop is 10% to 20%, the blood pressure pattern is a dipper;

[0118] If the percentage of nocturnal blood pressure drop is less than 10%, the blood pressure pattern is non-dipper;

[0119] If the percentage of nocturnal blood pressure drop is greater than 20%, the blood pressure pattern is super-dipper;

[0120] If the percentage of nocturnal blood pressure drop is less than 0%, the blood pressure pattern is a reverse dipper;

[0121] Judging whether it is hypertension based on the blood pressure load value:

[0122] If the blood pressure load value is between 5% and 15%, the blood pressure is normal; if the blood pressure load value is greater than 50%, it is hypertension.

[0123] Those skilled in the art know that, in addition to realizing the system and its various devices, modules, and units provided by the present invention in a purely computer-readable program code, it is entirely possible to realize the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a hardware component, and the devices, modules, and units included therein for realizing various functions can also be regarded as structures within the hardware component; the devices, modules, and units for realizing various functions can also be regarded as both software modules for realizing the method and structures within the hardware component.

[0124] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A method for synchronously analyzing long-term blood pressure and electrocardiogram, characterized in that: include: Step S1: pre-processing the signal values ​​collected by different devices; The signal values ​​include blood pressure signals and electrocardiogram signals; The preprocessing includes resampling the signal value and unifying the number of sampling points; Step S2: using the trained deep learning model to automatically identify and extract hidden features of the ECG signal to obtain analysis results of the ECG signal; Step S3: Calculate various blood pressure parameters according to the blood pressure signal to obtain the analysis result of the blood pressure signal; The blood pressure parameters include blood pressure load value and nocturnal blood pressure drop percentage; Step S4: Synchronously output the analysis results of the ECG signal and the analysis results of the blood pressure signal.

2. The method for synchronously analyzing long-term blood pressure and electrocardiogram according to claim 1, characterized in that: The step S2 includes using CNN to form a feature extractor, automatically mining the hidden features of the ECG or heartbeat segments in end-to-end learning, obtaining a valid feature expression of the input data, and obtaining the analysis result of the ECG signal.

3. The method for synchronously analyzing long-term blood pressure and electrocardiogram according to claim 1, characterized in that: The step S3 comprises the following sub-steps: Step S3.1: Recording dynamic blood pressure for 24 hours, so that the blood pressure measurement values ​​in different time periods reach corresponding preset numbers; the blood pressure measurement values ​​in different time periods include daytime measurement values ​​and nighttime measurement values; Step S3.2: Calculate the blood pressure load value according to the percentage of the number of times the blood pressure measurement value is greater than the preset value to the total number of blood pressure measurement values.

4. The method for synchronously analyzing long-term blood pressure and electrocardiogram according to claim 3, characterized in that: The calculation formula of the nighttime blood pressure drop percentage includes:

5. The method for synchronously analyzing long-term blood pressure and electrocardiogram according to claim 4, characterized in that: The analysis results of the blood pressure signal include: Judging blood pressure patterns based on the percentage of nighttime blood pressure drop: If the percentage of nocturnal blood pressure drop is 10% to 20%, the blood pressure pattern is a dipper; If the percentage of nocturnal blood pressure drop is less than 10%, the blood pressure pattern is non-dipper; If the percentage of nocturnal blood pressure drop is greater than 20%, the blood pressure pattern is super-dipper; If the percentage of nocturnal blood pressure drop is less than 0%, the blood pressure pattern is a reverse dipper; Judging whether it is hypertension based on the blood pressure load value: If the blood pressure load value is between 5% and 15%, the blood pressure is normal; if the blood pressure load value is greater than 50%, it is hypertension.

6. A system for synchronously analyzing long-term blood pressure and electrocardiogram, characterized in that: include: Module M1: pre-process the signal values ​​collected by different devices; The signal values ​​include blood pressure signals and electrocardiogram signals; The preprocessing includes resampling the signal value and unifying the number of sampling points; Module M2: Use the trained deep learning model to automatically identify and extract the hidden features of the ECG signal to obtain the analysis results of the ECG signal; Module M3: Calculate various blood pressure parameters according to the blood pressure signal to obtain the analysis result of the blood pressure signal; The blood pressure parameters include blood pressure load value and nocturnal blood pressure drop percentage; Module M4: Synchronously outputs the analysis results of the ECG signal and the blood pressure signal.

7. The system for synchronously analyzing long-term blood pressure and electrocardiogram according to claim 6, characterized in that: The module M2 includes using CNN to form a feature extractor, automatically mining the hidden features of the ECG or heartbeat segments in end-to-end learning, obtaining the effective feature expression of the input data, and obtaining the analysis results of the ECG signal.

8. The system for synchronously analyzing long-term blood pressure and electrocardiogram according to claim 6, characterized in that: The module M3 includes the following submodules: Module M3.1: Recording dynamic blood pressure for 24 hours, so that the blood pressure measurement values ​​in different time periods reach the corresponding preset numbers; the blood pressure measurement values ​​in different time periods include daytime measurement values ​​and nighttime measurement values; Module M3.2: Calculate the blood pressure load value based on the percentage of the number of times the blood pressure measurement value is greater than the preset value to the total number of blood pressure measurement values.

9. The system for synchronously analyzing long-term blood pressure and electrocardiogram according to claim 8, characterized in that: The calculation formula of the nighttime blood pressure drop percentage includes:

10. The system for synchronously analyzing long-term blood pressure and electrocardiogram according to claim 9, characterized in that: The analysis results of the blood pressure signal include: Judging blood pressure patterns based on the percentage of nighttime blood pressure drop: If the percentage of nocturnal blood pressure drop is 10% to 20%, the blood pressure pattern is a dipper; If the percentage of nocturnal blood pressure drop is less than 10%, the blood pressure pattern is non-dipper; If the percentage of nocturnal blood pressure drop is greater than 20%, the blood pressure pattern is super-dipper; If the percentage of nocturnal blood pressure drop is less than 0%, the blood pressure pattern is a reverse dipper; Judging whether it is hypertension based on the blood pressure load value: If the blood pressure load value is between 5% and 15%, the blood pressure is normal; if the blood pressure load value is greater than 50%, it is hypertension.

Images