A blood pressure estimation method, device, and storage medium

Abstract

The present invention relates to a blood pressure estimation method, device and storage medium, belonging to the technical field of medical devices, and solves the problems of insufficient accuracy of airbag blood pressure measurement and insufficient convenience of electrocardiogram and photoplethysmography blood pressure measurement in the prior art. It includes: Step S1, collecting the original air pressure data of the measured object, and performing data preprocessing on the original air pressure data to obtain the original mean pressure; Step S2, based on the original mean pressure, processing through a blood pressure artificial neural network module to obtain the pressure difference; Step S3, calculating the diastolic blood pressure and systolic blood pressure as the processing result based on the original mean pressure and the pressure difference, and storing and displaying the processing result.

Description

Technical Field

[0001] The present invention relates to the technical field of medical devices, and particularly relates to a blood pressure estimation method, device and storage medium. Background Art

[0002] The importance of blood pressure measurement lies in that it is one of the important indicators of normal physiological functions of the human body, plays a crucial role in maintaining the health of the cardiovascular system, and by measuring and analyzing blood pressure values, the health status of a person's heart and blood vessels can be understood.

[0003] Existing blood pressure estimation methods are based on the measurement of physiological parameters, such as electrocardiogram (ECG), photoplethysmography (PPG), etc. The blood pressure measurement devices used in these methods are highly dependent on signals and are greatly affected by factors such as skin contact, movement, and ambient light, which may lead to signal distortion or loss, thus affecting the accuracy of blood pressure measurement. In addition, existing machine learning-based blood pressure estimation methods improve the accuracy of blood pressure estimation by training a large amount of data. However, in these methods, due to complex signal processing, the computational complexity is relatively high, and the amount of calculation is large, making it difficult to be popularized in practical applications.

[0004] Chinese Patent Application for Invention, Publication No. CN111358452A, Invention Title "A Method and Device for Blood Pressure Prediction Based on Synchronous Signals", discloses a blood pressure prediction method, including synchronously collecting an electrocardiogram and a photoplethysmogram, and then performing data fusion processing on the obtained ECG signal and PPG signal, and calculating and inferring blood pressure data through a blood pressure model composed of a blood pressure convolutional neural network and a blood pressure artificial neural network. This method requires a complex electrocardiogram measurement machine, and the blood pressure model used has a large amount of calculation, resulting in the overall complexity of the method process and being unable to be integrated into low-computing-power wearable devices.

[0005] Chinese Patent Application for Invention, Publication No. CN113520357A, Invention Title "A Blood Pressure Measurement Device and Method", discloses a blood pressure measurement method, including collecting signals by a pressure sensor, and the processor performing fusion calculation on the collected signals to obtain a blood pressure value. In the processing of the collected information in this method, there is a lack of a correction process, resulting in a large error in the obtained result and it is difficult to obtain an accurate blood pressure value.

[0006] There is a need in this field for blood pressure estimation methods, devices and storage media with high accuracy, good convenience and relatively small amount of calculated data. Summary of the Invention

[0007] In order to solve the above problems, according to an embodiment of the present invention, there is provided a blood pressure estimation method, device and storage medium.

[0008] A blood pressure estimation method provided according to an embodiment of the present invention includes the following steps:

[0009] Step S1, collect the original air pressure data of the object to be measured, and perform data preprocessing on the original air pressure data to obtain the original mean pressure;

[0010] Step S2, based on the original mean pressure, process it through a blood pressure artificial neural network module to obtain the pressure difference;

[0011] Step S3, based on the original mean pressure and the pressure difference, calculate the diastolic blood pressure and systolic blood pressure as the processing result, and store and display the processing result;

[0012] Among them, the step S1 specifically includes the following steps:

[0013] Step S100, collect the original air pressure data, and the collected original air pressure data includes air pressure data and contact data;

[0014] Step S200, perform data preprocessing on the collected original air pressure data to obtain the original mean pressure as the preprocessed data;

[0015] Step S300, visitor mode judgment, judge whether the target user is in visitor mode according to the selection of the target user. If it is, judge it as visitor mode and execute step S400. Otherwise, judge it as personal mode and execute step S310;

[0016] Step S310, read the personal data and historical data of the target user;

[0017] Step S400, input the original mean pressure into the central processing unit module.

[0018] Optionally, the step S2 specifically includes the following steps:

[0019] Step S500, compare the input original mean pressure with the first threshold, and judge whether the original mean pressure exceeds the first threshold. If so, execute step S520. Otherwise, execute step S510;

[0020] Step S510, input the original mean pressure into the first blood pressure artificial neural network model, calculate and output the pressure difference, and execute step S530;

[0021] Step S520, input the original mean pressure into the second blood pressure artificial neural network model, calculate and output the pressure difference;

[0022] Step S530, input the pressure difference into the central processing unit module and store it in the data storage module.

[0023] Optionally, the step S3 specifically includes the following steps:

[0024] Step S600: Compare the obtained pressure difference with the second threshold value to determine whether the obtained pressure difference exceeds the second threshold value. If so, execute step S610; otherwise, proceed to step S700.

[0025] Step S610: Output an error, output an error message to the target user, and return to step S100.

[0026] Step S700: Calculate the estimated average pressure based on the obtained pressure difference and the original average pressure, and calibrate the estimated average pressure to obtain the calibrated estimated average pressure.

[0027] Step S800: Use the obtained calibrated estimated average pressure as the diastolic blood pressure estimate value, perform correction and calculation to obtain the diastolic blood pressure.

[0028] Step S900: Calculate the systolic blood pressure based on the obtained diastolic blood pressure and the calibrated estimated average pressure, and store, output, and display the obtained diastolic blood pressure and systolic blood pressure as the processing results.

[0029] Optionally, step S200 further includes the following steps:

[0030] Step S210: Perform a first filtering process on the collected original air pressure data to obtain an air pressure sequence and a contact sequence.

[0031] Step S220: Perform a second filtering process on the air pressure sequence to obtain dynamic pressure data and static pressure data.

[0032] Step S230: Perform Gaussian fitting on the dynamic pressure data to obtain a Gaussian curve.

[0033] Step S240: Perform static pressure linear fitting on the static pressure data to obtain a static pressure line.

[0034] Step S250: Calculate the original average pressure through the Gaussian curve.

[0035] Optionally, the first filtering process of step S210 specifically includes: performing Gaussian filtering on the air pressure data in the original air pressure data to obtain an air pressure sequence; and performing mean filtering on the contact data in the original air pressure data to obtain a contact sequence.

[0036] Optionally, the second filtering process of step S220 includes: performing Butterworth filtering on the air pressure sequence to obtain dynamic pressure data; and performing Nth-order IIR filtering on the air pressure sequence to obtain static pressure data.

[0037] Optionally, the first blood pressure artificial neural network model includes an ordinary blood pressure estimation module, and the second blood pressure artificial neural network model includes a systolic pressure estimation module and an ordinary blood pressure estimation module.

[0038] Optionally, step S800 specifically includes the following steps:

[0039] Step S810: According to the result of the visitor mode judgment, adjust the preset temporary diastolic pressure amplitude coefficient to obtain the diastolic pressure amplitude coefficient, and use the calibrated estimated mean pressure as the diastolic pressure estimated value, which is reduced by this diastolic pressure amplitude coefficient to obtain the reduced diastolic pressure estimated value;

[0040] Step S820: Perform offset correction on the reduced diastolic pressure estimated value to obtain the time of the diastolic pressure;

[0041] Step S830: Calculate the time of the diastolic pressure to obtain the diastolic pressure.

[0042] According to another embodiment of the present invention, there is provided an apparatus for performing the blood pressure estimation method, including:

[0043] An air pressure detection module, configured to detect and collect the original air pressure data of the target user, including a piezoelectric micropump, an airbag, and a pressure sensor;

[0044] A data preprocessing module, connected to the air pressure detection module, receiving the original air pressure data of the target user and performing data preprocessing to obtain the original mean pressure;

[0045] A blood pressure artificial neural network module, connected to the data preprocessing module, receiving the original mean pressure obtained after preprocessing to perform differential pressure estimation to obtain the differential pressure;

[0046] A central processing unit module, configured to perform calculation processing on the results obtained by each module and control each module;

[0047] A data storage module, configured to store the personal data of the target user and the data and results processed by each module;

[0048] A data transmission module, configured to transmit the processed data and results to an external device according to the instructions of the central processing unit module and for display.

[0049] According to another embodiment of the present invention, there is provided a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the blood pressure estimation method described above is implemented.

[0050] Compared with the prior art, the blood pressure estimation method, apparatus, and storage medium provided by the embodiments of the present invention have at least the following advantages.

[0051] 1. It effectively solves the problem of low stability of blood pressure estimation methods based on physiological parameters (such as PPG, ECG, etc.). Specifically, it solves the problem that due to individual differences and environmental factors, such as heart rate variability and vascular elasticity, these factors may lead to inaccurate blood pressure estimation, resulting in inaccurate measurement data and susceptibility to influence in existing non-airbag blood pressure detection methods.

[0052] 2. In the process of preprocessing the collected data, static pressure linear fitting and dynamic pressure Gaussian fitting are respectively used to characterize blood pressure characteristics with curve characteristic parameters, which is more accurate in calculation and has good physiological interpretability, and the finally obtained estimation results are also more accurate.

[0053] 3. The blood pressure estimation method of the present invention can collect air pressure data through a piezoelectric micropump and an airbag, which is convenient to be integrated on portable devices such as watches, and has high usability.

[0054] 4. The blood pressure estimation method of the present invention has high universality. In addition to wristband airbag measurement, it can also be used for air pressure values measured by devices such as cuff airbags, and accurate blood pressure values can be calculated.

[0055] 5. The blood pressure estimation method of the present invention is applicable to various groups of people and users. It distinguishes and calculates users and visitors through different calculation modes, and calculates blood pressure for the general population and hypertensive population respectively through different neural networks, and calculates blood pressure for target customers and visitors, or hypertensive population and general population respectively, and provides more accurate results. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. By referring to the drawings, the features and advantages of the present invention can be understood more clearly. The drawings are schematic and should not be construed as any limitation to the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0057] Figure 1 A flowchart showing the main steps of a blood pressure estimation method provided according to an embodiment of the present invention.

[0058] Figure 2 A flowchart showing a blood pressure estimation method provided according to an embodiment of the present invention.

[0059] Figure 3 A flowchart showing the data preprocessing in a blood pressure estimation method provided according to an embodiment of the present invention.

[0060] Figure 4The flowchart shows the method for correcting the estimated diastolic blood pressure in a blood pressure estimation method provided according to an embodiment of the present invention.

[0061] Figure 5 The logic block diagram shows a blood pressure estimation method provided according to an embodiment of the present invention.

[0062] Figure 6 The original air pressure data graph collected in an example of applying a blood pressure estimation method provided according to an embodiment of the present invention is shown.

[0063] Figure 7 The schematic diagram of the peak and valley of the dynamic pressure data after Butterworth filtering in an example of applying a blood pressure estimation method provided according to an embodiment of the present invention is shown.

[0064] Figure 8 The identification graph after removing pseudo peaks and pseudo valleys in an example of applying a blood pressure estimation method provided according to an embodiment of the present invention is shown.

[0065] Figure 9 The result graph after data preprocessing in an example of applying the blood pressure estimation method provided according to an embodiment of the present invention is shown.

[0066] Reference numerals:

[0067] 100, blood pressure estimation device;

[0068] 110, air pressure detection module;

[0069] 120, data preprocessing module;

[0070] 130, blood pressure artificial neural network module;

[0071] 140, data storage module;

[0072] 150, data transmission module;

[0073] 160, central processing unit module. Detailed embodiments

[0074] In order to more clearly understand the above objects, features and advantages of the present invention, the present invention will be further described in detail below with reference to the drawings and specific embodiments. It should be noted that, without conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other.

[0075] Many specific details are set forth in the following description in order to fully understand the present invention, but the present invention can also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below.

[0076] The following provides a detailed description of a blood pressure estimation method, device, and storage medium according to an embodiment of the present invention with reference to the accompanying drawings.

[0077] Example 1

[0078] As Figures 1 to 4 In Example 1 shown as follows, according to an embodiment of the present invention, a blood pressure estimation method is provided. First, the original air pressure data of the target user is obtained, then the original air pressure data is preprocessed, and then the preprocessed data is input into the blood pressure artificial neural network module to calculate the pressure difference, and finally the diastolic blood pressure and systolic blood pressure are obtained through the pressure difference.

[0079] Figure 1 And Figure 2 show the specific steps and processes of the blood pressure estimation method provided according to an embodiment of the present invention. As shown in Figure 1 And Figure 2 The flowchart shown includes the following steps.

[0080] Step S1: Collect the original air pressure data of the measured object and preprocess the original air pressure data to obtain the original mean pressure. This step S1 may specifically include the following steps S100 to S400. Optionally, the original air pressure data can be collected from the wrist part of the measured object by using a piezoelectric micropump and an airbag. Additionally, optionally, the original air pressure data can also be collected from the arm part of the measured object by using a cuff-type airbag.

[0081] Step S100: Collect the original air pressure data. Specifically, in this embodiment, the target user first wraps the blood pressure estimation device 100 around the wrist and clicks start. Then, the piezoelectric micropump is used to inflate the airbag. During the inflation process, the measured value by the pressure sensor is the original air pressure data of the target user. When starting to collect the original air pressure data, the user can also choose whether to use the guest mode. The measured value by the pressure sensor may include: the dynamic pressure and static pressure that constitute the air pressure data. And the collected information may also include the contact impedance value measured by the tactile sensor that constitutes the contact data. Figure 6 is a diagram of the air pressure data collected in an example of applying this embodiment. As shown in the figure, the curve represents the relationship between time and air pressure. In this embodiment, the pressurization of the original air pressure data is stable, and the step-up measurement method in the oscillometric method is adopted.

[0082] Step S200: Data preprocessing. Preprocess the collected original air pressure data to obtain the preprocessed data, and the preprocessed data includes the original mean pressure. Refer to Figure 3, which shows the process of the data preprocessing. This data preprocessing step can improve the universality of this method, is suitable for processing and calculating the air pressure values collected by all devices, has high accuracy, and the calculated data after processing is more accurate. As Figure 3 shown in the flowchart of, step S200 specifically includes the following steps S210 to step S250.

[0083] Step S210, perform a first filtering process on the collected original air pressure data to obtain an air pressure sequence and a contact sequence. This first filtering process includes: performing Gaussian filtering on the air pressure data in the original air pressure data to obtain an air pressure sequence; and performing mean filtering on the contact data in the original air pressure data to obtain a contact sequence. Store the obtained air pressure sequence and contact sequence in a memory (for example, flash). This contact sequence can be used to determine whether the airbag fits the user's wrist, realize the drift correction of the pulse wave curve, and achieve the avoidance of measurement errors caused by different wearing tightness.

[0084] Step S220, perform a second filtering process on the air pressure sequence to obtain dynamic pressure data and static pressure data. This second filtering process includes: performing Butterworth filtering on the air pressure sequence obtained in the above step to obtain dynamic pressure data; and performing N - order IIR (Infinite Impulse Response, recursive filter) filtering on the air pressure sequence to obtain static pressure data. Refer to Figure 7 , which shows a schematic diagram of the dynamic pressure data obtained by performing Butterworth filtering on the air pressure sequence in an example.

[0085] Step S230, perform Gaussian fitting on the dynamic pressure data to obtain a Gaussian curve. Figure 7 is a schematic diagram of the dynamic pressure data after Butterworth filtering in an example of applying a blood pressure estimation method of the present invention. As Figure 7 shown, are the suspicious points of the peaks and troughs of the identified pulse wave; Figure 8 is an identification diagram of the removed false peaks and false troughs in an example of applying a blood pressure estimation method provided according to an embodiment of the present invention. As Figure 8 shown, are the corrected peaks and troughs. Refer to Figure 9 The curve shown in shows a schematic diagram of the drawn Gaussian curve in an example, while Figure 7 and Figure 8It shows a schematic diagram of the process of performing Gaussian fitting in an example. The steps of Gaussian fitting include: First, find the maximum point (maximum value) of the set of pulse wave signals, and record its amplitude and time at this point; Given the heart rate of a person (1 - 2 times / step S) and the sampling rate, set the value between two wave peaks; Take the difference of the pulse wave signal sequence, and the extreme point positions in the pulse wave sequence can be obtained when the signs before and after are opposite; According to the maximum time of the pulse wave and the set threshold, gradually search for the extreme values from the middle to both ends of the pulse wave sequence; Correct the interference points; Obtain all the peak point positions. The above-mentioned pulse wave signal is a graph showing the relationship between the air pressure data measured by a pressure sensor and time (for example, Figure 6 as shown), and is used as dynamic pressure data.

[0086] Step S240, perform static pressure linear fitting on the static pressure data to obtain a static pressure line. Through static pressure linear fitting, a mathematical model of the linear relationship between static pressure and position variables can be established. Refer to Figure 9 the line shown in, which shows the static pressure line obtained by static pressure linear fitting in the example. This static pressure line can be used to calculate diastolic blood pressure and systolic blood pressure.

[0087] Step S250, calculate the original mean pressure through the Gaussian curve. Among the Gaussian curves obtained by performing Gaussian fitting on the dynamic pressure data, the air pressure value at the moment corresponding to the highest point is the original mean pressure.

[0088] The data preprocessing process of the above step S200 is described as follows. Perform Gaussian filtering on the air pressure data in the collected original air pressure data to obtain an air pressure sequence, and perform mean filtering on the contact data in the original air pressure data to obtain a contact sequence; Then perform Butterworth filtering on the air pressure sequence to obtain dynamic pressure data, and perform N - order IIR filtering on the air pressure sequence to obtain static pressure data; Perform Gaussian fitting on the dynamic pressure data to obtain a Gaussian curve; Perform linear fitting on the static pressure data to obtain a static pressure line; Calculate the original mean pressure through the air pressure value at the moment corresponding to the highest point of the Gaussian curve after Gaussian fitting.

[0089] Continue to refer to Figure 2 , and the blood pressure estimation method provided according to the embodiment of the present invention further includes the following steps.

[0090] Step S300, visitor mode judgment, determine whether the target user is in visitor mode. If so, it is judged as visitor mode and step S400 is executed; otherwise, it is judged as personal mode and step S310 is executed. Specifically, in this embodiment, the target user selects whether to be in visitor mode in this step. Additionally, optionally, the target user can also be reminded to select whether to use visitor mode when collecting the original air pressure data in step S100.

[0091] Step S310: Read the personal data and historical data of the target user. Specifically, in this embodiment, when the target user does not select the guest mode at the start of the test, it is defaulted to select the personal user mode, and the historical data of the target user is read in this step. This personal user mode stores the personal data and historical data of the target user, where the historical data of the target user may include the personal basic information of the target user and information such as whether the target user has hypertension. The personal basic information may include information such as gender, age, height, weight, wrist circumference, and body mass index (BMI).

[0092] Step S400: Data input. Input the original mean pressure into the central processing unit module.

[0093] Step S2: Based on the original mean pressure, process it through the blood pressure artificial neural network module to obtain the pressure difference. This step S2 specifically includes the following steps S500 to S530.

[0094] Step S500: Compare the input original mean pressure with the first threshold and determine whether the original mean pressure exceeds the first threshold. If so, execute step S520; if not, execute step S510. Specifically, in this embodiment, the first threshold can be preset comprehensively according to the blood pressure value judgment standard of statistical data and the historical data of the user.

[0095] Step S510: Import the original mean pressure into the first blood pressure artificial neural network model for calculation to obtain the pressure difference, and then execute step S530. Specifically, in this embodiment, the first blood pressure artificial neural network model performs calculation processing when the original mean pressure is less than the first threshold. This first blood pressure artificial neural network model may include a general blood pressure estimation module and can use a trained ANN neural network. The pressure difference can be used as the difference between the original mean pressure and the estimated mean pressure.

[0096] Step S520: Import the original mean pressure into the second blood pressure artificial neural network model for calculation to obtain the pressure difference. Specifically, in this embodiment, the second blood pressure artificial neural network model performs calculation processing when the original mean pressure is greater than the first threshold. This second blood pressure artificial neural network model may include a high-pressure estimation module and a general blood pressure estimation module and can use a trained ANN neural network. The pressure difference can be used as the difference between the original mean pressure and the estimated mean pressure.

[0097] The above blood pressure artificial neural network model may include 2 - 10 layers of neural networks, and each layer of neural network may include 3 - 20 neurons.

[0098] Step S530, the central processing unit module receives the pressure difference and stores it in the data storage module 140. Obtain the result of step S510 or step S520 to get the pressure difference. By comparing the original average pressure with the first threshold as described above and then selecting the method of processing by the first blood pressure artificial neural network model or the second blood pressure artificial neural network model according to the judgment result, the pressure difference can be calculated specifically through the corresponding model according to the individual situation of the target user, so as to obtain a more accurate blood pressure estimation result.

[0099] Step S3, based on the original average pressure and the pressure difference, calculate the diastolic blood pressure and systolic blood pressure as the processing result, and store and display the processing result. This step S3 specifically includes the following steps S600 to step S900.

[0100] Step S600, compare the obtained pressure difference with the second threshold to determine whether the obtained pressure difference exceeds the second threshold. If so, execute step S610; otherwise, proceed to step S700. Specifically, in this embodiment, the second threshold is preset according to statistical data. For example, the second threshold can be set to 10 mmHg to 125 mmHg. Thus, when the pressure difference is less than 10 mmHg or greater than 125 mmHg, it is determined that the second threshold is exceeded; while when the pressure difference is in the range of 10 mmHg to 125 mmHg, it is determined that the second threshold is not exceeded. By comparing the obtained pressure difference with the second threshold and then deciding whether to give an error reminder or perform further processing according to the judgment result, problems in the processing process can be responded to in a timely manner to avoid invalid calculations.

[0101] Step S610, output an error, output an error message to the target user, and return to step S100.

[0102] Step S700, calculate the estimated average pressure through the obtained pressure difference and the original average pressure, and calibrate the estimated average pressure to obtain the calibrated estimated average pressure. This calibration is to convert the unlabeled data in the original data set into a structured and machine-readable data set.

[0103] Step S800, use the obtained calibrated estimated average pressure as the diastolic blood pressure estimate value, perform correction and calculation to obtain the diastolic blood pressure. Refer to Figure 4 , which shows the specific steps and process of correcting the diastolic blood pressure in this embodiment. As Figure 4 shown in the flowchart, this step specifically includes the following steps S810 to step S830.

[0104] Step S810: According to the judgment result of the visitor mode, adjust the preset temporary diastolic pressure amplitude coefficient to obtain the diastolic pressure amplitude coefficient, and use the calibrated estimated mean pressure as the diastolic pressure estimate value, which is reduced by this diastolic pressure amplitude coefficient to obtain the reduced diastolic pressure estimate value. This reduction is not a simple multiplication calculation. The temporary diastolic pressure amplitude coefficient is preset according to statistical data. In this embodiment, when it is determined that the target user is in the visitor mode, this temporary diastolic pressure amplitude coefficient is directly used; otherwise, for the personal user mode, the temporary diastolic pressure amplitude coefficient can be modified according to the information preset or provided by the target user (for example, the personal data or historical data of the target user).

[0105] Step S820: Perform offset correction on the reduced diastolic pressure estimate value to obtain the time of the diastolic pressure. Specifically, in this embodiment, this offset correction process is performed through a data analysis method. Optionally, this offset correction may include subtracting the time difference corresponding to the pressure difference from the time corresponding to the reduced estimated mean pressure to obtain the time of the diastolic pressure.

[0106] Step S830: Calculate the diastolic pressure based on the time of the diastolic pressure. This calculation includes finding the static pressure value corresponding to the time of the diastolic pressure on the obtained static pressure straight line to obtain the diastolic pressure.

[0107] Step S900: Calculate the systolic pressure based on the obtained diastolic pressure and the calibrated estimated mean pressure. Specifically, in this embodiment, combine the calibrated estimated mean pressure and the diastolic pressure, calculate their difference, amplify this difference by the systolic pressure accuracy coefficient, and add the value of the calibrated estimated mean pressure to obtain the systolic pressure. The systolic pressure accuracy coefficient can be preset according to data statistics. This amplification is not a simple multiplication amplification.

[0108] The working process of the blood pressure estimation method provided in this embodiment is as follows: First, the piezoelectric micropump of the blood pressure estimation device 100 inflates the airbag. During the inflation process, the pressure sensor and the tactile sensor measure and obtain the original air pressure data of the target user, and this original air pressure data includes air pressure data and contact data; perform data preprocessing on the obtained original air pressure data, including filtering and fitting, to obtain the original mean pressure; compare and judge the original mean pressure with the first threshold, and select to use the first blood pressure artificial neural network model or the second blood pressure artificial neural network model; then input the original mean pressure into the selected blood pressure artificial neural network model for processing to obtain the pressure difference, and then compare and judge this pressure difference with the second threshold to determine whether the pressure difference exceeds the range of the second threshold. If it exceeds the range, a wrong reminder is fed back and returned. Otherwise, the calibrated estimated mean pressure is obtained through the processing of the pressure difference and the original mean pressure; then the diastolic pressure is calculated through the calibrated estimated mean pressure; finally, the systolic pressure is calculated using the diastolic pressure and the calibrated estimated mean pressure. Output and display the obtained diastolic pressure, systolic pressure, and the result of the pulse rate.

[0109] Example 2

[0110] Figure 5 It is a logic block diagram of a blood pressure estimation device provided according to an embodiment of the present invention, which shows the specific components and processing procedures of the blood pressure estimation device 100 in this embodiment. Refer to Figure 5 , which shows a blood pressure estimation device 100 for performing the blood pressure estimation method of the above-mentioned Embodiment 1 provided according to another embodiment of the present invention, including: a barometric pressure detection module 110 for detecting and collecting the original barometric pressure data of the target user; a data preprocessing module 120 connected to the barometric pressure detection module 110, receiving the original barometric pressure data of the target user and performing data preprocessing to obtain the original mean pressure; a blood pressure artificial neural network module 130 connected to the data preprocessing module 120, receiving the original mean pressure obtained after preprocessing to estimate the pressure difference and obtain the pressure difference; a central processing unit module 160 for calculating and processing the results obtained by each module and controlling each module of the blood pressure estimation device 100; a data storage module 140 for storing the personal data of the target user and the data and results processed by each module; a data transmission module 150 for data transmission between internal modules of the device, and transmitting the processed data and results to external devices and for display according to the instructions of the central processing unit module 160.

[0111] Optionally, the barometric pressure detection module 110 may include a piezoelectric micropump, an airbag, and a pressure sensor. By using a piezoelectric micropump and a micro airbag, it is beneficial to obtain a more miniaturized blood pressure estimation device, which is convenient for integration or assembly into a miniature wearable device or medical device. Among them, the piezoelectric micropump in this embodiment may be a MEMS piezoelectric micropump. When in use, the target user first sets the barometric pressure detection module 110 of the blood pressure estimation device around the wrist. When starting the measurement, the piezoelectric micropump is used to inflate the airbag, and during the inflation process, the pressure sensor measures the original barometric pressure data of the target user.

[0112] As Figure 5As shown, in this embodiment, the blood pressure estimation device 100 further includes a blood pressure artificial neural network module 130. The blood pressure artificial neural network module 130 includes a first blood pressure artificial neural network model and a second blood pressure artificial neural network model. The first blood pressure artificial neural network model includes an ordinary blood pressure estimation module, and the second blood pressure artificial neural network model includes a high blood pressure estimation module and an ordinary blood pressure estimation module. The blood pressure artificial neural network module 130 can be set to use the first blood pressure artificial neural network model to calculate and estimate the blood pressure of target users with an original mean pressure less than a first threshold, and the second blood pressure artificial neural network model to calculate and estimate the blood pressure of target users with an original mean pressure greater than the first threshold. Optionally, the blood pressure artificial neural network used in this embodiment can be an ANN network. Among them, the ordinary blood pressure estimation module can be obtained by training with the statistical data of the general population; while the high blood pressure estimation module can be obtained by training with the statistical data of the high blood pressure population. By specifically using different models for data processing and blood pressure estimation for target users in different blood pressure ranges, more accurate estimation values can be obtained.

[0113] As Figure 5 As shown, the blood pressure estimation device 100 of this embodiment further includes a data preprocessing module 120. The data preprocessing module 120 can call the central processing unit module 160 for data processing. The blood pressure estimation device 100 further includes a data storage module 140. In this embodiment, FLASH flash memory is used as the data storage module 140 to store the data obtained by each module. The blood pressure estimation device 100 further includes a data transmission module 150. The data transmission module 150 can include wired data transmission and / or wireless data transmission. In this embodiment, the various modules inside the blood pressure estimation device 100 can perform wired data transmission, and the data transmission between the blood pressure estimation device 100 and external devices can use Bluetooth and wireless networks.

[0114] The working process of a blood pressure estimation device 100 in this embodiment is as follows: After the target user wears the blood pressure estimation device 100 at the root of the styloid process of the ulna on the wrist, the user first selects to use the personal mode or the visitor mode; then clicks to start measuring blood pressure. At this time, the piezoelectric micropump of the air pressure detection module 110 in the blood pressure estimation device 100 starts to work and inflates the airbag. During this process, the pressure sensor measures and collects the original air pressure data of the target user, and inputs the collected original air pressure data into the data preprocessing module 120; the data preprocessing module 120 calls the central processing unit module 160 to perform data preprocessing to obtain the preprocessed data (i.e., the original mean pressure); then the preprocessed data is compared and judged with the first threshold through the central processing unit module 160, and according to the judgment result, the first blood pressure artificial neural network model or the second blood pressure artificial neural network model in the blood pressure artificial neural network module 130 is selected; according to the selection result, the preprocessed data is input into the blood pressure artificial neural network module 130 to output the estimated pressure difference; the pressure difference is compared and judged with the second threshold, and according to the judgment result, an error reminder is executed and returned, or the blood pressure estimation is continued; the central processing unit module 160 calculates the diastolic blood pressure and systolic blood pressure by using the mean pressure and the pressure difference, and outputs the estimated results of the diastolic blood pressure and systolic blood pressure to the target user.

[0115] The following Table 1 shows the measurement results obtained from a blood pressure value estimation test 1 for an exemplary blood pressure estimation device according to Embodiment 2 of the present invention. For the sake of brevity, only a part of the measurement results is cited here.

[0116] In this test 1, 255 people were randomly selected for auscultation comparison tests. Auscultation was performed by specialized nurses, and two people auscultated simultaneously and took the average value as the control group for the test. The measurement group was for control measurement, and the subjects were required to follow the measurement standard postures: sit still for more than 3 minutes; measure according to the posture required by F1-PRO; keep still, not talk, and not move during the measurement process; no one who can be measured was excluded, and samples were randomly collected; through the statistical analysis of the deviations between the measurement group and the control group, the average deviation of the systolic blood pressure was 0.85 mmHg, the standard deviation of the systolic blood pressure was 6.87 mmHg, the average deviation of the diastolic blood pressure was 2.41 mmHg, and the standard deviation of the diastolic blood pressure was 6.21 mmHg. Comparing with the standard document "YY 0670-2008", the results show that the clinical test meets the acceptance standards for medical-grade sphygmomanometers.

[0117] Table 1 Test data table for Test 1

[0118]

[0119] The following Table 2 is a comparison table of test data for Test 2. In this Test 2, data calculated by the blood pressure estimation device according to Embodiment 2 of the present invention is used, and the measured data of the Omron 7121 model blood pressure measurement device is used as a reference to compare the test data of blood pressure estimation. The pressure difference between the measured values of the blood pressure estimation device of Embodiment 2 and the Omron 7121 is small, and the standard deviation of the blood pressure estimation device of Embodiment 2 is very small, indicating that both the accuracy and consistency of the blood pressure estimation device of Embodiment 2 are relatively high.

[0120] Table 2 Comparison Table of Test Data for Test 2

[0121]

[0122] Embodiment 3

[0123] According to Embodiment 3 of the present invention, a computer-readable storage medium is also provided. A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the method provided in Embodiment 1 of the present invention can be implemented.

[0124] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

[0125] All of the above optional technical solutions can be combined arbitrarily to form optional embodiments of the present application, which will not be elaborated here one by one.

[0126] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0127] The above is only the preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A blood pressure estimation method, characterized in that: The following steps are involved: Step S1, collecting the original air pressure data of the measured object, and performing data preprocessing on the original air pressure data to obtain the original average pressure; Step S2, based on the original average pressure, the pressure difference is obtained by processing through the blood pressure artificial neural network module; Step S3, based on the original mean pressure and the pressure difference, the diastolic pressure and the systolic pressure are calculated as the processing result, and the processing result is stored and displayed; Wherein, the step S1 specifically includes the following steps: Step S100, collecting original air pressure data, the collected original air pressure data includes air pressure data and contact data; Step S200, preprocessing the collected original air pressure data to obtain the original average pressure as preprocessed data; Step S300, judging the visitor mode, judging whether the target user is in the visitor mode according to the target user's selection, if yes, judging it as the visitor mode and executing step S400, otherwise judging it as the personal mode and executing step S310; Step S310, reading the personal data and historical data of the target user; Step S400, inputting the original average pressure into the central processing unit module; Wherein, step S3 specifically includes: Step S600, comparing the obtained pressure difference with the second threshold value to determine whether the obtained pressure difference exceeds the second threshold value, if so, executing step S610, otherwise executing step S700; Step S610, outputting an error, outputting an error prompt message to the target user, and returning to step S100; Step S700, calculating an estimated average pressure by using the obtained pressure difference and the original average pressure, and calibrating the estimated average pressure to obtain a calibrated estimated average pressure; Step S800, taking the obtained calibrated estimated mean pressure as the estimated value of diastolic pressure, correcting and calculating to obtain the diastolic pressure; Step S900, calculating the systolic pressure based on the obtained diastolic pressure and the calibrated estimated mean pressure, storing the obtained diastolic pressure and systolic pressure as processing results, and outputting and displaying them; Wherein, step S800 specifically includes: Step S810, adjusting the preset temporary diastolic pressure amplitude coefficient according to the result of the visitor mode determination to obtain the diastolic pressure amplitude coefficient, and using the calibrated estimated mean pressure as the diastolic pressure estimated value, and reducing it by the diastolic pressure amplitude coefficient to obtain the reduced diastolic pressure estimated value; Step S820, performing offset correction on the reduced diastolic pressure estimate to obtain the diastolic pressure time; Step S830, calculating the diastolic pressure time to obtain the diastolic pressure.

2. The blood pressure estimation method according to claim 1, characterized in that in, The step S2 specifically includes the following steps: Step S500, comparing the inputted original average pressure with the first threshold, and determining whether the original average pressure exceeds the first threshold, if so, executing step S520, otherwise executing step S510; Step S510, inputting the original average pressure into the first blood pressure artificial neural network model, outputting the pressure difference after calculation, and executing step S530; Step S520, inputting the original average pressure into the second blood pressure artificial neural network model, and outputting the pressure difference after calculation; Step S530, input the pressure difference into the central processing module and store it in the data storage module.

3. The blood pressure estimation method according to claim 1, characterized in that: The step S200 further includes the following steps: Step S210, performing a first filtering process on the collected raw air pressure data to obtain an air pressure sequence and a contact sequence; Step S220, performing a second filtering process on the air pressure sequence to obtain dynamic pressure data and static pressure data; Step S230, performing Gaussian fitting on the dynamic pressure data to obtain a Gaussian curve; Step S240, performing static pressure straight line fitting on the static pressure data to obtain a static pressure straight line; Step S250, obtaining the original average pressure through Gaussian curve calculation.

4. The blood pressure estimation method according to claim 3, characterized in that: The first filtering process of step S210 specifically includes: performing Gaussian filtering on the air pressure data in the original air pressure data to obtain an air pressure sequence; and performing mean filtering on the contact data in the original air pressure data to obtain a contact sequence.

5. The blood pressure estimation method according to claim 3, characterized in that: The second filtering process of step S220 includes: performing Butterworth filtering on the air pressure sequence to obtain dynamic pressure data; and performing N-order IIR filtering on the air pressure sequence to obtain static pressure data.

6. The blood pressure estimation method according to claim 2, characterized in that: The first blood pressure artificial neural network model includes a common blood pressure estimation module, and the second blood pressure artificial neural network model includes a high blood pressure estimation module and a common blood pressure estimation module.

7. A device for executing the blood pressure estimation method according to any one of claims 1 to 6, characterized in that: include: Air pressure detection module, used to detect and collect the original air pressure data of the target user, including piezoelectric micro pump, air bag and pressure sensor; The data preprocessing module is connected to the air pressure detection module, receives the original air pressure data of the target user and performs data preprocessing to obtain the original average pressure; The blood pressure artificial neural network module is connected to the data preprocessing module, receives the original average pressure obtained after preprocessing, performs pressure difference estimation, and obtains the pressure difference; The central processing unit module calculates and processes the results obtained by each module and controls each module; Data storage module, which stores the personal data of the target user and the data and results processed by each module; The data transmission module transmits the processed data and results to the external device for display according to the instructions of the central processing unit module.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the blood pressure estimation method according to any one of claims 1 to 6 is implemented.

Images