A pressure value determination system, method, device and readable storage medium

Abstract

The present application discloses a pressure value determination system, method, device and readable storage medium, the system comprising: an acquisition module, a first preprocessing module, a second preprocessing module, a neural network module, for inputting envelope difference waveform data and pressure change trend waveform data into a neural network model, so that the neural network model outputs the diastolic pressure and systolic pressure of the pressure data to be measured; wherein the neural network model is specifically a neural network model obtained by training and learning using pressure training samples, and the pressure training samples include: envelope difference waveform samples and pressure change trend waveform samples corresponding to the pressure data samples; diastolic pressure samples and their diastolic pressure moments on the pressure change trend waveform samples; systolic pressure samples and their systolic pressure moments on the pressure change trend waveform samples. The present application can accurately determine the values ​​and positions of diastolic pressure and systolic pressure according to the pressure data to be measured, and has the advantages of self-adaptation, self-organization and real-time learning.

Description

Technical Field

[0001] The present invention relates to the field of digital signal processing, and in particular to a pressure value determination system, method, device and readable storage medium. Background Art

[0002] Among many blood pressure measurement methods, intermittent blood pressure measurement is widely used because of its advantages of simple operation, short measurement time, and the ability to periodically record the user's blood pressure measurement values. The intermittent blood pressure measurement method is mainly based on the principle of Koch sound method, which blocks arterial blood flow by inflating the cuff, and then determines the diastolic and systolic blood pressure by monitoring the sound of the arterial blood vessels. Electronic blood pressure monitors that apply this principle have broad application prospects because of their simple operation and low threshold for use.

[0003] However, existing electronic blood pressure monitors generally choose a fixed proportional coefficient method or an inflection point method when determining diastolic and systolic blood pressure. However, different cases require different proportional coefficients, and the accuracy of a single fixed ratio is poor. At the same time, some test samples do not have obvious inflection points, and the inflection point method cannot be used well to determine the pressure value. Therefore, there is currently no pressure testing solution with high accuracy.

[0004] Therefore, how to provide a solution to the above technical problems is a problem that technicians in this field need to solve. Summary of the invention

[0005] In view of this, the object of the present invention is to provide an accurate and reliable pressure value determination system, method, device and readable storage medium. The specific scheme is as follows:

[0006] A pressure value determination system, comprising:

[0007] An acquisition module, used for acquiring pressure data to be measured;

[0008] A first preprocessing module is used to separate the pressure data to be measured to obtain separated pulse waveform data and pressure change trend waveform data;

[0009] A second preprocessing module, used to determine the upper envelope and the lower envelope of the pulse waveform data, and obtain envelope difference waveform data of the upper envelope and the lower envelope;

[0010] A neural network module, used for inputting the envelope difference waveform data and the pressure change trend waveform data into a neural network model, so that the neural network model outputs the diastolic pressure and the systolic pressure of the pressure data to be measured;

[0011] The neural network model is specifically a neural network model trained and learned using stress training samples, and the stress training samples include:

[0012] Envelope difference waveform samples and pressure change trend waveform samples corresponding to the pressure data samples;

[0013] Diastolic pressure samples and their diastolic pressure moments on the pressure change trend waveform samples;

[0014] The systolic pressure sample and the systolic pressure moment on the pressure change trend waveform sample.

[0015] Preferably, the pressure value determination system further includes a training sample acquisition module for acquiring the pressure training sample, and the training sample acquisition module includes:

[0016] An initial sample acquisition unit, used to acquire the diastolic pressure sample and the systolic pressure sample obtained by the sample providing subject through manual auscultation, and record the pressure data sample at the same time;

[0017] A sample processing unit, used for filtering and separating the pressure data samples to obtain the envelope difference waveform samples and the pressure change trend waveform samples;

[0018] The time determination unit is used to determine the diastolic pressure time of the diastolic pressure sample on the pressure change trend waveform sample, and is also used to determine the systolic pressure time of the systolic pressure sample on the pressure change trend waveform sample.

[0019] Preferably, the time determination unit is further used for:

[0020] Determine a first moment of a valid pulse starting point and a second moment of a maximum pulse in the envelope difference waveform sample;

[0021] The envelope difference waveform sample, the pressure change trend waveform sample, the first moment, the second moment, the diastolic pressure moment and the systolic pressure moment are used as the pressure training samples.

[0022] Preferably, the neural network model is specifically a neural network classifier model.

[0023] Preferably, the pressure value determination system further includes a training sample acquisition module for acquiring the pressure training sample, and the training sample acquisition module includes:

[0024] An initial sample acquisition unit, used to acquire the diastolic pressure sample and the systolic pressure sample obtained by the sample providing subject through manual auscultation, and record the pressure data sample at the same time;

[0025] A sample processing unit, used for filtering and separating the pressure data samples to obtain the envelope difference waveform samples and the pressure change trend waveform samples;

[0026] a time determination unit, used to determine a first time of a valid pulse starting point and a second time of a maximum pulse in the envelope difference waveform sample, and also used to determine the diastolic pressure time of the diastolic pressure sample on the pressure change trend waveform sample, and also used to determine the systolic pressure time of the systolic pressure sample on the pressure change trend waveform sample;

[0027] The interval determination unit is used to partition the envelope difference waveform samples and the pressure change trend waveform samples according to preset intervals and take the first moment as the starting point to determine the diastolic pressure moment interval and the systolic pressure moment interval corresponding to the diastolic pressure moment and the systolic pressure moment respectively.

[0028] Preferably, the preset interval is specifically a ratio of the duration from the first moment to the second moment to the number of preset intervals;

[0029] The interval determination unit uses the envelope difference waveform sample, the pressure change trend waveform sample, the preset interval, the diastolic pressure time interval, and the systolic pressure time interval as the pressure training samples.

[0030] Preferably, the neural network module comprises:

[0031] A first determining unit, used to determine a current diastolic pressure time interval and a current systolic pressure time interval of the envelope difference waveform data;

[0032] A second determination unit is used to determine the midpoint of the diastolic pressure interval of the current diastolic pressure time interval, and determine the midpoint of the systolic pressure interval of the current systolic pressure time interval;

[0033] The third determination unit is used to determine the diastolic pressure corresponding to the midpoint of the diastolic pressure interval and the systolic pressure corresponding to the midpoint of the systolic pressure interval, respectively, based on the pressure change trend waveform data, and output the diastolic pressure and the systolic pressure.

[0034] Accordingly, the present application also discloses a method for determining a pressure value, comprising:

[0035] Obtain the pressure data to be measured;

[0036] Separating the pressure data to be measured to obtain separated pulse waveform data and pressure change trend waveform data;

[0037] Determine the upper envelope and the lower envelope of the pulse waveform data, and obtain envelope difference waveform data of the upper envelope and the lower envelope;

[0038] Inputting the envelope difference waveform data and the pressure change trend waveform data into a neural network model, so that the neural network model outputs the diastolic pressure and the systolic pressure of the pressure data to be measured;

[0039] The neural network model is specifically a neural network model trained and learned using stress training samples, and the stress training samples include:

[0040] Envelope difference waveform samples and pressure change trend waveform samples corresponding to the pressure data samples;

[0041] Diastolic pressure samples and their diastolic pressure moments on the pressure change trend waveform samples;

[0042] The systolic pressure sample and the systolic pressure moment on the pressure change trend waveform sample.

[0043] Accordingly, the present application also discloses a pressure value determination device, comprising:

[0044] A pressure sensor, used for generating pressure data to be measured;

[0045] Memory for storing computer programs;

[0046] A processor is used to implement the steps of the pressure value determination method as described above when executing the computer program.

[0047] Correspondingly, the present application also discloses a readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the pressure value determination method described above are implemented.

[0048] The present application discloses a pressure value determination system, including: an acquisition module, a first preprocessing module, a second preprocessing module, and a neural network module, which is used to input the envelope difference waveform data and the pressure change trend waveform data into a neural network model, so that the neural network model outputs the diastolic pressure and systolic pressure of the pressure data to be measured; wherein the neural network model is specifically a neural network model trained and learned using pressure training samples, and the pressure training samples include: envelope difference waveform samples and pressure change trend waveform samples corresponding to the pressure data samples; diastolic pressure samples and their diastolic pressure moments on the pressure change trend waveform samples; systolic pressure samples and their systolic pressure moments on the pressure change trend waveform samples. The present application trains and learns the neural network model through pressure training samples, thereby determining the complex relationship between diastolic pressure, systolic pressure and pressure change trend waveform, so that the system can accurately judge the values ​​and positions of diastolic pressure and systolic pressure according to the pressure data to be measured, and has the advantages of self-adaptation, self-organization and real-time learning. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0050] Figure 1 This is a structural distribution diagram of a pressure value determination system in an embodiment of the present invention:

[0051] Figure 2 A schematic diagram of processing pressure data to be measured in an embodiment of the present invention;

[0052] Figure 3 A structural distribution diagram of a specific pressure value determination system in an embodiment of the present invention;

[0053] Figure 4 A structural distribution diagram of another specific pressure value determination system in an embodiment of the present invention;

[0054] Figure 5 A schematic diagram of processing a pressure data sample in an embodiment of the present invention;

[0055] Figure 6 It is a structural distribution diagram of a method for determining a pressure value in an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] Existing electronic blood pressure monitors generally choose the fixed proportional coefficient method or the inflection point method when determining diastolic and systolic blood pressure. However, different cases require different proportional coefficients, and the accuracy of a single fixed ratio is poor. At the same time, some test samples do not have obvious inflection points, and the inflection point method cannot be used well to determine the pressure value. Therefore, there is currently no pressure testing solution with high accuracy.

[0058] The present application trains and learns the neural network model through pressure training samples, thereby determining the complex relationship between diastolic pressure, systolic pressure and pressure change trend waveform, so that the present application can accurately determine the value and position of diastolic pressure and systolic pressure according to the pressure data to be measured, and has the advantages of self-adaptation, self-organization and real-time learning.

[0059] The embodiment of the present invention discloses a pressure value determination system, see Figure 1 As shown, including:

[0060] Acquisition module 1, used to acquire pressure data to be measured;

[0061] The first preprocessing module 2 is used to separate the pressure data to be measured to obtain separated pulse waveform data and pressure change trend waveform data;

[0062] The second preprocessing module 3 is used to determine the upper envelope and the lower envelope of the pulse waveform data, and obtain the envelope difference waveform data of the upper envelope and the lower envelope;

[0063] A neural network module 4, used for inputting the envelope difference waveform data and the pressure change trend waveform data into the neural network model, so that the neural network model outputs the diastolic pressure and the systolic pressure of the pressure data to be measured;

[0064] The neural network model is specifically a neural network model trained and learned using stress training samples, and the stress training samples include:

[0065] Envelope difference waveform samples and pressure change trend waveform samples corresponding to the pressure data samples;

[0066] Diastolic pressure samples and their diastolic pressure moments on the pressure change trend waveform samples;

[0067] Systolic blood pressure samples and their systolic blood pressure moments on the pressure change trend waveform samples.

[0068] It can be understood that this embodiment is based on a deflation measurement process, which includes: first closing the pump valve, inflating the cuff to block the blood flow in the upper arm to an almost completely blocked state, and then opening the deflation valve to slowly deflate, and the acquisition module 1 obtains the pressure data to be measured collected by the pressure sensor during the deflation process.

[0069] Further, such as Figure 2 As shown, the first preprocessing module 2 separates the pressure data to be measured, and obtains the separated pulse waveform data 102 and the pressure change trend waveform data 101 through a threshold algorithm, and the filtering process is mainly for the pulse waveform data 102; the second preprocessing module 3 determines the upper and lower envelopes and the envelope difference waveform data 103 of the pulse waveform data 102 through methods such as rank-order filter; finally, the envelope difference waveform data 103 and the pressure change trend waveform data 101 are input into the neural network model, and the trained neural network model can output the accurate diastolic pressure and systolic pressure of the pressure data to be measured.

[0070] It can be understood that the neural network model in this embodiment implies the corresponding relationship between the diastolic pressure Dia. on the pressure change trend waveform data 101 and the moment of the envelope difference waveform data 103, and the same applies to the systolic pressure Sys. Therefore, in order to ensure the accuracy of the neural network model, the pressure training samples used in this embodiment include: the envelope difference waveform samples and the pressure change trend waveform samples corresponding to the pressure data samples; the diastolic pressure samples and their diastolic pressure moments on the pressure change trend waveform samples; the systolic pressure samples and their systolic pressure moments on the pressure change trend waveform samples. Here, the systolic pressure samples and diastolic pressure samples are usually determined by manual auscultation, and the results of manual auscultation are the most accurate compared to other methods. The systolic pressure moment on the pressure change trend waveform sample is determined using the numerical value of the systolic pressure sample. Similarly, the diastolic pressure moment riding on the pressure change trend waveform sample is determined using the numerical value of the diastolic pressure sample.

[0071] This application trains and learns the neural network model through pressure training samples, thereby determining the complex relationship between diastolic pressure, systolic pressure and pressure change trend waveform, so that the system can accurately determine the value and position of diastolic pressure and systolic pressure according to the pressure data to be tested. Compared with the traditional fixed coefficient method and inflection point method, this system overcomes the defects of traditional solutions in processing intuitive and unstructured information, and has the advantages of self-adaptation, self-organization and real-time learning.

[0072] The embodiment of the present invention discloses a specific pressure value determination system. Compared with the previous embodiment, this embodiment further illustrates and optimizes the technical solution. Figure 3 As shown:

[0073] The pressure value determination system further includes a training sample acquisition module 5 for acquiring pressure training samples. The training sample acquisition module 5 includes:

[0074] An initial sample acquisition unit 51 is used to acquire a diastolic pressure sample and a systolic pressure sample obtained by a sample providing subject through manual auscultation, and record a pressure data sample at the same time;

[0075] The sample processing unit 52 is used to filter and separate the pressure data samples to obtain envelope difference waveform samples and pressure change trend waveform samples;

[0076] The time determination unit 53 is used to determine the diastolic pressure time of the diastolic pressure sample on the pressure change trend waveform sample, and is also used to determine the systolic pressure time of the systolic pressure sample on the pressure change trend waveform sample.

[0077] Specifically, the initial sample acquisition unit 51 works in the deflation measurement process, records the pressure data samples in the process, and determines the diastolic pressure and the systolic pressure in the process as the diastolic pressure sample and the systolic pressure sample, respectively.

[0078] It can be understood that the processing actions of the sample processing unit 52 are equivalent to the first preprocessing module 2 and the second preprocessing module 3 .

[0079] Furthermore, the time determination unit 53 is further configured to:

[0080] Determine a first moment of a valid pulse starting point and a second moment of a maximum pulse in the envelope difference waveform sample;

[0081] The envelope difference waveform samples, the pressure change trend waveform samples, the first moment, the second moment, the diastolic pressure moment and the systolic pressure moment are used as pressure training samples.

[0082] It can be understood that the most important feature points in the envelope difference waveform samples and envelope difference waveform data are the effective pulse starting point and the maximum pulse. Figure 2 For example, the effective pulse starting point E is the peak of the first effective wave of the pulse waveform data 102, the first moment corresponds to point A, the maximum pulse corresponds to the maximum value point max of the envelope difference waveform data 103, and the second moment corresponds to point B. Figure 2 The medium pressure data is taken as a pressure data sample. In the pressure training sample obtained by the final moment determination unit 53, the first moment is the moment value of point A, the second moment is the moment value of point B, the systolic pressure moment is the moment value of point Ps corresponding to Sys., and the diastolic pressure moment is the moment value of point Pd corresponding to Dia.

[0083] It can be seen that in this embodiment, the corresponding relationship between the diastolic pressure and the systolic pressure in the pressure change trend waveform data and the envelope difference waveform data in the neural network model is further defined as the relationship between the first moment, the second moment, the diastolic pressure moment and the systolic pressure moment. Select multiple pressure training samples to train the neural network model. The trained neural network model can obtain the corresponding accurate diastolic pressure moment and systolic pressure moment according to the input envelope difference waveform data and pressure change trend waveform data, and further find the pressure values ​​corresponding to the diastolic pressure moment and systolic pressure moment on the current pressure change trend waveform data to determine the corresponding diastolic pressure and systolic pressure.

[0084] The embodiment of the present invention discloses a specific pressure value determination system. Compared with the previous embodiment, this embodiment further illustrates and optimizes the technical solution. Figure 4 As shown:

[0085] Furthermore, during actual training selection, the neural network model may be selected to be specifically a neural network classifier model.

[0086] When training and using the neural network classifier model, it is necessary to segment samples or data by intervals, so the pressure value determination system also includes a training sample acquisition module 5 for acquiring pressure training samples. The training sample acquisition module 5 includes:

[0087] An initial sample acquisition unit 51 is used to acquire a diastolic pressure sample and a systolic pressure sample obtained by a sample providing subject through manual auscultation, and record a pressure data sample at the same time;

[0088] The sample processing unit 52 is used to filter and separate the pressure data samples to obtain envelope difference waveform samples and pressure change trend waveform samples;

[0089] a time determination unit 53, for determining a first time of a valid pulse starting point and a second time of a maximum pulse in an envelope difference waveform sample, for determining a diastolic pressure time of a diastolic pressure sample on a pressure change trend waveform sample, and for determining a systolic pressure time of a systolic pressure sample on a pressure change trend waveform sample;

[0090] The interval determination unit 54 is used to partition the envelope difference waveform samples and the pressure change trend waveform samples according to preset intervals and starting from the first moment, and determine the diastolic pressure moment interval and the systolic pressure moment interval corresponding to the diastolic pressure moment and the systolic pressure moment respectively.

[0091] Furthermore, the preset interval is specifically a ratio of the duration from the first moment to the second moment to the number of preset intervals;

[0092] The interval determination unit 54 uses the envelope difference waveform sample, the pressure change trend waveform sample, the preset interval, the diastolic pressure time interval, and the systolic pressure time interval as pressure training samples.

[0093] by Figure 5 Take the pressure data sample processing as an example, where the pressure change trend waveform sample is 101, the pulse waveform sample is 102, the pressure change trend waveform sample is 103, the effective pulse starting point is point E, the corresponding first moment is point A, the maximum pulse is point Max, the corresponding second moment is point B, the corresponding pressure value is H, and the distance between points AB is L. A proportional slope c0=L / H can be obtained.

[0094] The determination of systolic pressure and diastolic pressure has a similar idea, but they are not closely related when determined. Therefore, this embodiment first takes the neural network model training of systolic pressure as an example for explanation:

[0095] The preset interval is determined to be SP=L / n, where n is the number of preset intervals, which is actually the number of preset intervals from the effective pulse starting point to the maximum pulse. It can be set according to the target accuracy and can be set to 20. Then the interval numbers between AB are 1, 2, ..., 20. After partitioning, the interval determination unit 54 can determine the diastolic pressure time interval, such as Figure 2 The systolic blood pressure time interval ns of the systolic blood pressure Sys. is 5.

[0096] In order to reduce the amount of data and normalize the data, the interval interception point of the envelope difference waveform sample is selected and multiplied by the coefficient -100 to obtain the following Figure 2 As shown in a1-a20, a neural network model of systolic pressure is designed to train the systolic pressure time interval. The model parameters are [a1, a2, ..., a20, c0], and the model is established:

[0097] [a1, a2,…, a20, c0]--Wns--ns;

[0098] Similarly, partitioning continues to be performed toward the right side of point B at the same preset intervals until the partition includes the diastolic pressure moment point Pd. Figure 2 The diastolic pressure time interval nd of the middle diastolic pressure Dia. is 6.

[0099] In order to reduce the amount of data and normalize the data, the interval interception point of the envelope difference waveform sample is selected and multiplied by the coefficient -100 to obtain the following Figure 2 As shown in b1-b6, a neural network model of diastolic pressure is designed to train the diastolic pressure time interval. The model parameters are [b1, b2, ..., b6, c0], and the model is established:

[0100] [b1, b2,…, b6, c0]--Wnd--nd.

[0101] The above two models are trained, and finally the trained neural network model is applied in the neural network module 4.

[0102] Furthermore, the neural network module 4 includes:

[0103] A first determining unit 41 is used to determine a current diastolic pressure time interval and a current systolic pressure time interval of the envelope difference waveform data;

[0104] A second determining unit 42, configured to determine a diastolic pressure interval midpoint of a current diastolic pressure time interval, and determine a systolic pressure interval midpoint of a current systolic pressure time interval;

[0105] The third determination unit 43 is used to determine the diastolic pressure corresponding to the midpoint of the diastolic pressure interval and the systolic pressure corresponding to the midpoint of the systolic pressure interval on the pressure change trend waveform data, and output the diastolic pressure and the systolic pressure.

[0106] It can be understood that the first determination unit 41 obtains the current diastolic pressure time interval and the current systolic pressure time interval corresponding to the current pressure data to be measured through the neural network model, and then determines the midpoints of the two time intervals and their corresponding pressure values ​​on the pressure change trend waveform data.

[0107] This application trains and learns the neural network model through pressure training samples, thereby determining the complex relationship between diastolic pressure, systolic pressure and pressure change trend waveform, so that the system can accurately determine the value and position of diastolic pressure and systolic pressure according to the pressure data to be tested. Compared with the traditional fixed coefficient method and inflection point method, this system overcomes the defects of traditional solutions in processing intuitive and unstructured information, and has the advantages of self-adaptation, self-organization and real-time learning.

[0108] Accordingly, the present application also discloses a method for determining a pressure value, see Figure 6 As shown, including:

[0109] S1: Obtain the pressure data to be measured;

[0110] S2: filtering and separating the pressure data to be measured to obtain separated pulse waveform data and pressure change trend waveform data;

[0111] S3: determining an upper envelope and a lower envelope of the pulse waveform data, and obtaining envelope difference waveform data of the upper envelope and the lower envelope;

[0112] S4: inputting the envelope difference waveform data and the pressure change trend waveform data into a neural network model, so that the neural network model outputs the diastolic pressure and the systolic pressure of the pressure data to be measured;

[0113] The neural network model is specifically a neural network model trained and learned using stress training samples, and the stress training samples include:

[0114] Envelope difference waveform samples and pressure change trend waveform samples corresponding to the pressure data samples;

[0115] Diastolic pressure samples and their diastolic pressure moments on the pressure change trend waveform samples;

[0116] The systolic pressure sample and the systolic pressure moment on the pressure change trend waveform sample.

[0117] This application trains and learns the neural network model through pressure training samples, thereby determining the complex relationship between diastolic pressure, systolic pressure and pressure change trend waveform, so that this application can accurately determine the value and position of diastolic pressure and systolic pressure according to the pressure data to be tested. Compared with the traditional fixed coefficient method and inflection point method, this application overcomes the defects of traditional solutions in processing intuitive and unstructured information, and has the advantages of self-adaptation, self-organization and real-time learning.

[0118] In some specific embodiments, the process of obtaining training samples includes:

[0119] Obtain the diastolic pressure sample and the systolic pressure sample obtained by the sample providing subject through manual auscultation, and record the pressure data sample at the same time;

[0120] Filtering and separating the pressure data samples to obtain the envelope difference waveform samples and the pressure change trend waveform samples;

[0121] The method for determining the diastolic pressure moment of the diastolic pressure sample on the pressure change trend waveform sample is also used to determine the systolic pressure moment of the systolic pressure sample on the pressure change trend waveform sample.

[0122] In some specific embodiments, the process of obtaining training samples further includes:

[0123] Determine a first moment of a valid pulse starting point and a second moment of a maximum pulse in the envelope difference waveform sample;

[0124] The envelope difference waveform sample, the pressure change trend waveform sample, the first moment, the second moment, the diastolic pressure moment and the systolic pressure moment are used as the pressure training samples.

[0125] In some specific embodiments, the neural network model is specifically a neural network classifier model.

[0126] In some specific embodiments, the process of obtaining training samples includes:

[0127] Obtain the diastolic pressure sample and the systolic pressure sample obtained by the sample providing subject through manual auscultation, and record the pressure data sample at the same time;

[0128] Filtering and separating the pressure data samples to obtain the envelope difference waveform samples and the pressure change trend waveform samples;

[0129] Determine the first moment of the effective pulse starting point and the second moment of the maximum pulse in the envelope difference waveform sample, and also determine the diastolic pressure moment of the diastolic pressure sample on the pressure change trend waveform sample, and also determine the systolic pressure moment of the systolic pressure sample on the pressure change trend waveform sample;

[0130] According to preset intervals, taking the first moment as a starting point, the envelope difference waveform samples and the pressure change trend waveform samples are partitioned to determine the diastolic pressure moment interval and the systolic pressure moment interval corresponding to the diastolic pressure moment and the systolic pressure moment respectively.

[0131] In some specific embodiments, the preset interval is specifically a ratio of the duration from the first moment to the second moment to the number of preset intervals;

[0132] The interval determination unit uses the envelope difference waveform sample, the pressure change trend waveform sample, the preset interval, the diastolic pressure time interval, and the systolic pressure time interval as the pressure training samples.

[0133] In some specific embodiments, the process of inputting the envelope difference waveform data and the pressure change trend waveform data into a neural network model so that the neural network model outputs the diastolic pressure and the systolic pressure of the pressure data to be measured includes:

[0134] Inputting the envelope difference waveform data and the pressure change trend waveform data into a neural network model so that the neural network model determines a current diastolic pressure time interval and a current systolic pressure time interval of the envelope difference waveform data;

[0135] Determine the midpoint of the diastolic pressure interval of the current diastolic pressure time interval, and determine the midpoint of the systolic pressure interval of the current systolic pressure time interval;

[0136] The diastolic pressure corresponding to the midpoint of the diastolic pressure interval and the systolic pressure corresponding to the midpoint of the systolic pressure interval are determined on the pressure change trend waveform data, and the diastolic pressure and the systolic pressure are output.

[0137] Accordingly, the present application also discloses a pressure value determination device, comprising:

[0138] A pressure sensor, used for generating pressure data to be measured;

[0139] Memory for storing computer programs;

[0140] A processor is used to implement the steps of the pressure value determination method as described above when executing the computer program.

[0141] Correspondingly, the present application also discloses a readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the pressure value determination method described above are implemented.

[0142] For details about the method for determining the pressure value, please refer to the relevant description in the above embodiment, which will not be repeated here.

[0143] Among them, the pressure value determination device and the readable storage medium in this embodiment have the same technical effects as in the above embodiments, and will not be repeated here.

[0144] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0145] The above is a detailed introduction to a pressure value determination system, method, device and readable storage medium provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for a person skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A pressure value determination system, characterized in that: include: An acquisition module, used for acquiring pressure data to be measured; A first preprocessing module is used to separate the pressure data to be measured to obtain separated pulse waveform data and pressure change trend waveform data; A second preprocessing module, used to determine the upper envelope and the lower envelope of the pulse waveform data, and obtain envelope difference waveform data of the upper envelope and the lower envelope; A neural network module, used for inputting the envelope difference waveform data and the pressure change trend waveform data into a neural network model, so that the neural network model outputs the diastolic pressure and the systolic pressure of the pressure data to be measured; The neural network model is specifically a neural network model trained and learned using stress training samples, and the stress training samples include: Envelope difference waveform samples and pressure change trend waveform samples corresponding to the pressure data samples; Diastolic pressure samples and their diastolic pressure moments on the pressure change trend waveform samples; A systolic pressure sample and its systolic pressure moment on the pressure change trend waveform sample; The pressure value determination system further includes a training sample acquisition module, which is used to acquire the pressure training sample. The training sample acquisition module includes: An initial sample acquisition unit, used to acquire the diastolic pressure sample and the systolic pressure sample obtained by the sample providing subject through manual auscultation, and record the pressure data sample at the same time; A sample processing unit, used for filtering and separating the pressure data samples to obtain the envelope difference waveform samples and the pressure change trend waveform samples; a time determination unit, used to determine the diastolic pressure time of the diastolic pressure sample on the pressure change trend waveform sample, and also used to determine the systolic pressure time of the systolic pressure sample on the pressure change trend waveform sample; The time determination unit is also used for: Determine a first moment of a valid pulse starting point and a second moment of a maximum pulse in the envelope difference waveform sample; wherein the valid pulse starting point is a peak of the first valid wave of the pulse waveform data; The envelope difference waveform sample, the pressure change trend waveform sample, the first moment, the second moment, the diastolic pressure moment and the systolic pressure moment are used as the pressure training samples.

2. The pressure value determination system according to claim 1, characterized in that: The neural network model is specifically a neural network classifier model.

3. The pressure value determination system according to claim 2, characterized in that: It also includes a training sample acquisition module, which is used to acquire the pressure training sample. The training sample acquisition module includes: An initial sample acquisition unit, used to acquire the diastolic pressure sample and the systolic pressure sample obtained by the sample providing subject through manual auscultation, and record the pressure data sample at the same time; A sample processing unit, used for filtering and separating the pressure data samples to obtain the envelope difference waveform samples and the pressure change trend waveform samples; a time determination unit, used to determine a first time of a valid pulse starting point and a second time of a maximum pulse in the envelope difference waveform sample, and also used to determine the diastolic pressure time of the diastolic pressure sample on the pressure change trend waveform sample, and also used to determine the systolic pressure time of the systolic pressure sample on the pressure change trend waveform sample; The interval determination unit is used to partition the envelope difference waveform samples and the pressure change trend waveform samples according to preset intervals and take the first moment as the starting point to determine the diastolic pressure moment interval and the systolic pressure moment interval corresponding to the diastolic pressure moment and the systolic pressure moment respectively.

4. The pressure value determination system according to claim 3, characterized in that: The preset interval is specifically a ratio of the duration from the first moment to the second moment to the number of preset intervals; The interval determination unit uses the envelope difference waveform sample, the pressure change trend waveform sample, the preset interval, the diastolic pressure time interval, and the systolic pressure time interval as the pressure training samples.

5. The pressure value determination system according to claim 4, characterized in that: The neural network module comprises: A first determining unit, used to determine a current diastolic pressure time interval and a current systolic pressure time interval of the envelope difference waveform data; A second determination unit is used to determine the midpoint of the diastolic pressure interval of the current diastolic pressure time interval, and determine the midpoint of the systolic pressure interval of the current systolic pressure time interval; The third determination unit is used to determine the diastolic pressure corresponding to the midpoint of the diastolic pressure interval and the systolic pressure corresponding to the midpoint of the systolic pressure interval, respectively, based on the pressure change trend waveform data, and output the diastolic pressure and the systolic pressure.

6. A method for determining a pressure value, characterized in that: include: Obtain the pressure data to be measured; Separating the pressure data to be measured to obtain separated pulse waveform data and pressure change trend waveform data; Determine the upper envelope and the lower envelope of the pulse waveform data, and obtain envelope difference waveform data of the upper envelope and the lower envelope; Inputting the envelope difference waveform data and the pressure change trend waveform data into a neural network model, so that the neural network model outputs the diastolic pressure and the systolic pressure of the pressure data to be measured; The neural network model is specifically a neural network model trained and learned using stress training samples, and the stress training samples include: Envelope difference waveform samples and pressure change trend waveform samples corresponding to the pressure data samples; Diastolic pressure samples and their diastolic pressure moments on the pressure change trend waveform samples; A systolic pressure sample and its systolic pressure moment on the pressure change trend waveform sample; The process of obtaining training samples includes: Obtain the diastolic pressure sample and the systolic pressure sample obtained by the sample providing subject through manual auscultation, and record the pressure data sample at the same time; Filtering and separating the pressure data samples to obtain the envelope difference waveform samples and the pressure change trend waveform samples; Determine the diastolic pressure moment of the diastolic pressure sample on the pressure change trend waveform sample, and determine the systolic pressure moment of the systolic pressure sample on the pressure change trend waveform sample; Determine a first moment of a valid pulse starting point and a second moment of a maximum pulse in the envelope difference waveform sample; wherein the valid pulse starting point is a peak of the first valid wave of the pulse waveform data; The envelope difference waveform sample, the pressure change trend waveform sample, the first moment, the second moment, the diastolic pressure moment and the systolic pressure moment are used as the pressure training samples.

7. A pressure value determination device, characterized in that: include: A pressure sensor, used for generating pressure data to be measured; Memory for storing computer programs; A processor is used to implement the steps of the pressure value determination method as claimed in claim 6 when executing the computer program.

8. A readable storage medium, characterized in that: The readable storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the pressure value determination method according to claim 6 are implemented.

Images