Blood pressure measuring device and method, electronic equipment, storage medium and program product

Abstract

The invention relates to the technical field of sphygmomanometers, and discloses a blood pressure measuring device and method, electronic equipment, a storage medium and a program product, and the blood pressure measuring device comprises a cuff, a piezoelectric sensor, an air pipe, a pressure sensor and a host. The cuff comprises a first cuff part and a second cuff part, an air storage chamber is defined by the first cuff part and the second cuff part, the first cuff part and the second cuff part are integrally formed or connected in a sealed mode, an air nozzle is arranged on the cuff, and the air nozzle is communicated with the air storage chamber; the piezoelectric sensor is fixed in the air storage chamber and used for collecting Korotkoff sound signals, the air pipe is provided with a first end and a second end, the first end is connected to the air nozzle, and the air pipe is communicated with the air storage chamber; the pressure sensor is connected to the second end and used for collecting a pressure signal, and the pressure signal is used for representing the internal air pressure of the air storage chamber; the host is used for predicting the blood pressure measurement result according to the Korotkoff sound signal and the pressure signal. Compared with the prior art, the accuracy of the blood pressure measurement result can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of sphygmomanometers, and in particular to a blood pressure measuring device, method, electronic device, storage medium and program product. Background Art

[0002] Electronic sphygmomanometers based on the Korotkoff sound method have the advantages of easy operation, wide application range, and reliable measurement results. Therefore, electronic sphygmomanometers based on the Korotkoff sound method have been widely used. The cuff of a conventional electronic sphygmomanometer based on the Korotkoff sound method is provided with an independent air bag, and the piezoelectric sensor is placed under the air bag and is close to the skin when in use to collect signals. However, since the piezoelectric sensor is located under the air bag, the upper air bag will always be impacted by the blood flow. This impact is transmitted to the lower piezoelectric sensor through the air bag, causing the signal collected by the piezoelectric sensor to be affected by the blood flow impact, thereby increasing the difficulty of judging the Korotkoff sound signal and affecting the accuracy of the blood pressure measurement results. Summary of the invention

[0003] In view of this, the present invention provides a blood pressure measurement device, method, electronic device, storage medium and program product to solve the problem that blood pressure measurement results are easily affected by blood flow shock.

[0004] In a first aspect, the present invention provides a blood pressure measuring device, the device comprising:

[0005] The cuff comprises a first cuff portion and a second cuff portion, the first cuff portion and the second cuff portion enclose an air storage chamber, the first cuff portion and the second cuff portion are integrally formed or sealedly connected, and an air nozzle is provided on the cuff, and the air nozzle is communicated with the air storage chamber;

[0006] A piezoelectric sensor is fixed inside the gas storage chamber, and the piezoelectric sensor is used to collect Korotkoff sound signals;

[0007] An air pipe having a first end and a second end, wherein the first end is connected to the air nozzle, and the air pipe is in communication with the air storage chamber;

[0008] A pressure sensor connected to the second end, the pressure sensor is used to collect a pressure signal, and the pressure signal is used to represent the internal air pressure of the air storage chamber;

[0009] The host is used to obtain Korotkoff sound signals and pressure signals, and to predict blood pressure measurement results according to the Korotkoff sound signals and pressure signals. The piezoelectric sensor and the pressure sensor are respectively connected to the host for communication.

[0010] The present invention does not separately set up an airbag, but provides a design scheme of a cuff without an airbag, in which a piezoelectric sensor is set in the cuff, and specifically set in an air storage chamber surrounded by a first cuff portion and a second cuff portion, so as to reduce or even avoid the influence of the airbag pressure generated by blood flow impact on the collection of Korotkoff sound signals during the blood pressure measurement process, thereby improving the accuracy of Korotkoff sound collection. Compared with conventional technologies, the present invention can not only improve the accuracy of blood pressure measurement results, but also help to reduce materials and reduce the manufacturing cost of blood pressure measuring devices. The assembly process of the blood pressure measuring device is simpler, which is convenient for the production and manufacturing of the blood pressure measuring device.

[0011] In a second aspect, the present invention provides a blood pressure measurement method, the method comprising:

[0012] Acquire a Korotkoff sound signal collected by a piezoelectric sensor fixed inside an air storage chamber and a pressure signal collected by a pressure sensor, wherein the air storage chamber is surrounded by a first cuff portion and a second cuff portion in a cuff, wherein the first cuff portion and the second cuff portion are integrally formed or sealed, and the cuff is provided with an air nozzle connected to the air storage chamber, wherein the air nozzle is used to connect to a first end of an air pipe, wherein the air pipe is connected to the air storage chamber, and the pressure sensor is connected to a second end of the air pipe, and the pressure signal is used to characterize the internal air pressure of the air storage chamber;

[0013] The blood pressure measurement results are predicted based on the Korotkoff sound signals and pressure signals.

[0014] The piezoelectric sensor of the blood pressure measuring device used in the blood pressure measurement process of the present invention is arranged in an air storage chamber surrounded by the first cuff part and the second cuff part. This method can reduce or even avoid the influence of the air bag pressure generated by blood flow impact on the collection of Korotkoff sound signals during the blood pressure measurement process, improve the accuracy of Korotkoff sound collection, and improve the accuracy of blood pressure measurement results.

[0015] In a third aspect, the present invention provides an electronic device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the blood pressure measurement method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0016] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the blood pressure measurement method of the first aspect or any corresponding embodiment thereof.

[0017] In a fifth aspect, the present invention provides a computer program product, comprising computer instructions for causing a computer to execute the blood pressure measurement method of the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the related technologies, the drawings required for use in the specific embodiments or the related technical descriptions will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 Schematic diagram of the structure of a blood pressure measuring device according to an embodiment of the present invention.

[0020] Figure 2 is a schematic structural diagram of a gas nozzle according to an embodiment of the present invention.

[0021] Figure 3 is a curve diagram of a pressure signal according to an embodiment of the present invention.

[0022] Figure 4 is a schematic diagram of a curve of a Korotkoff sound signal according to an embodiment of the present invention.

[0023] Figure 5 is a schematic diagram of a curve of a pressure signal in a pressure drop stage according to an embodiment of the present invention.

[0024] Figure 6 is a schematic diagram of a curve of an oscillation wave signal according to an embodiment of the present invention.

[0025] Figure 7 is a schematic diagram of a curve of a Korotkoff sound signal in a pressure drop stage according to an embodiment of the present invention.

[0026] Figure 8 is a schematic diagram of a Korotkoff sound signal locked according to an oscillating wave signal according to an embodiment of the present invention.

[0027] Fig. 9 4 is a schematic diagram of a spectrum diagram of a Korotkoff sound signal according to an embodiment of the present invention.

[0028] Fig.10 is a schematic diagram of a convolutional bidirectional long short-term memory neural network according to an embodiment of the present invention.

[0029] Fig.11 is a schematic diagram of a signal frequency mean and a signal frequency variance according to an embodiment of the present invention.

[0030] Fig.12 4 is a flow chart of a blood pressure measurement method according to an embodiment of the present invention.

[0031] Fig.13 4 is a flow chart of another blood pressure measurement method according to an embodiment of the present invention.

[0032] Fig.14 It is a schematic diagram of the hardware structure of the electronic device according to an embodiment of the present invention.

[0033] Reference numerals:

[0034] 10. Cuff; 11. First cuff portion; 12. Second cuff portion; 101. Air storage chamber; 102. Air nozzle; 20. Piezoelectric sensor; 30. Air tube; 40. Pressure sensor; 50. Host. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0036] For non-invasive blood pressure measurement technology, the most recognized and accurate method is the Korotkoff sound method. The process of blood pressure measurement based on the Korotkoff sound method is generally as follows: place the stethoscope on the lower side of the air bag of the cuff and close it to the skin, pressurize the air bag, and when the air pressure in the air bag reaches a certain pressure value, the blood flow in the upper arm artery can be blocked, and then release the gas in the air bag to slowly reduce the pressure. When the first pulse sound appears in the stethoscope, the pressure value at this time is a high pressure value, and the pressure is continuously released. When the last pulse sound in the stethoscope disappears, the pressure value at this time is a low pressure value. With the rapid development of non-invasive blood pressure measurement technology, electronic sphygmomanometers based on the Korotkoff sound method have emerged. Electronic sphygmomanometers generally include a cuff and a host. The cuff is used to collect piezoelectric signals and pressure signals, and the host is used to calculate the blood pressure measurement results based on the collected piezoelectric signals and pressure signals; conventional cuffs need to be provided with air bags, and the piezoelectric sensor is placed under the air bag and close to the skin. However, since the piezoelectric sensor is located below the airbag, the upper airbag will always be subjected to the impact of blocked blood flow. This impact is transmitted to the lower piezoelectric sensor through the airbag, causing the signal collected by the piezoelectric sensor to be affected by the blood flow impact (i.e., the interference of the airbag signal caused by the blood flow impact), thereby increasing the difficulty of judging the Korotkoff sound signal and affecting the accuracy of the blood pressure measurement results.

[0037] like Figure 1 As shown, and can be combined Figure 2 An embodiment of the present invention provides a blood pressure measuring device, which includes but is not limited to a cuff 10, a piezoelectric sensor 20, a trachea 30, a pressure sensor 40 and a host 50.

[0038] The cuff 10 includes a first cuff portion 11 and a second cuff portion 12, the first cuff portion 11 and the second cuff portion 12 enclose an air storage chamber 101, the first cuff portion 11 and the second cuff portion 12 are integrally formed or sealed, and an air nozzle 102 is provided on the cuff 10, and the air nozzle 102 is connected to the air storage chamber 101. Of course, the cuff 10 may also include a third cuff portion (the blank area on the right side of the air storage chamber 101 in the figure, not marked), the third cuff portion may be a structure extending from the first cuff portion 11 or the second cuff portion 12, or a piece of non-woven fabric fixedly connected to the first cuff portion 11 and / or the second cuff portion 12, the third cuff portion may be used to roll the cuff to cover the user's arm when the cuff 10 is used, and the air storage chamber 101 may be attached to the user's skin (specifically, the artery of the arm) through a fixing component such as Velcro provided on the cuff 10; wherein, the air storage chamber 101 may be used to store air.

[0039] Specifically, in this embodiment, the edges of the first cuff portion 11 and the second cuff portion 12 are pressed together to seal the edge of the first cuff portion 11 with the edge of the second cuff portion 12, thereby forming an air storage chamber 101; alternatively, the air storage chamber 101 can also be directly formed inside the cuff 10 manufactured by integral molding.

[0040] The piezoelectric sensor 20 is fixed inside the air storage chamber 101, and the piezoelectric sensor 20 is used to collect Korotkoff sound signals; the piezoelectric sensor 20 is pressed into the cuff 10, and the piezoelectric sheet in the piezoelectric sensor is specifically fixed inside the air storage chamber 101. The Korotkoff sound signal is a piezoelectric signal collected by the piezoelectric sensor 20.

[0041] The number of the piezoelectric sensors 20 may be one or more. A piezoelectric sensor may include at least one piezoelectric sheet. All the piezoelectric sheets on the device of this embodiment collect the piezoelectric signal together to obtain the above-mentioned Korotkoff sound signal and transmit it to the host 50.

[0042] In some optional embodiments, the piezoelectric sensor 20 is communicatively connected to the host 50 via a connecting line built into the trachea 30. The piezoelectric sensor 20 may include a first piezoelectric sheet and a second piezoelectric sheet attached to a target inner wall of the air storage chamber 101. The target inner wall is the inner wall of the air storage chamber 101 close to the inner side of the cuff 10.

[0043] Specifically, the front side of the first piezoelectric sheet and the second piezoelectric sheet are attached to the side of the air storage chamber 101 facing the user's skin (when using the blood pressure measurement device), and the front side of the first piezoelectric sheet and the second piezoelectric sheet are bonded to the inner wall of the air storage chamber 101 (for example, by using 3M double-sided tape), and the piezoelectric sensor 20 can be further bonded to the inner wall of the air storage chamber 101 by acetate cloth tape, which plays a role of secondary reinforcement; the back side of the first piezoelectric sheet and the second piezoelectric sheet are attached to the sensor housing, and the connecting wire extends out through the sensor housing and is connected to the host 50 by being built into the trachea 30. This embodiment realizes the reuse of the trachea by placing the connecting wire in the trachea without leading out the connecting wire separately, so that the host 50 and the cuff 10 only need to be connected through the trachea, which simplifies the structure of the blood pressure measurement device, and the dual piezoelectric sheets attached to the inside of the airbag-free cuff and the gas fluctuations in the airbag jointly detect the Korotkoff sound signal, thereby improving the detection accuracy of the Korotkoff sound signal.

[0044] Optionally, the first piezoelectric sheet and the second piezoelectric sheet are distributed on both sides of the air nozzle 102 and close to the air nozzle 102, and are symmetrically arranged with respect to the cuff axis where the air nozzle 102 is located. In this embodiment, the distance between the first piezoelectric sheet and the second piezoelectric sheet can be 100 mm.

[0045] Specifically, the first piezoelectric sheet and the second piezoelectric sheet are disposed in the lower half of the cuff 10 . When blood pressure is measured, the first piezoelectric sheet and the second piezoelectric sheet are respectively placed on both sides of the artery.

[0046] The air pipe 30 has a first end and a second end, the first end is connected to the air nozzle 102, and the air pipe 30 is connected to the air storage chamber 101; in this embodiment, the second end of the air pipe 30 is connected to the main unit 50, and the main unit 50 is used to control the inflation into the air storage chamber 101 through the air pipe 30 or the deflation of the air storage chamber 101 through the air pipe 30.

[0047] Combination Figure 2 As shown, the first cuff portion 11 and the second cuff portion 12 form an air storage chamber 101 which is connected to the trachea 30 via an air nozzle 102. The air nozzle 102 can be arranged on an air nozzle joint, which is used to be fixedly connected to the first end of the trachea 30 and connect the trachea 30 with the air storage chamber 101.

[0048] The pressure sensor 40 is connected to the second end. The pressure sensor 40 is used to collect a pressure signal. The pressure signal is used to represent the internal air pressure of the air storage chamber 101 .

[0049] The host 50 is used to obtain Korotkoff sound signals and pressure signals, and the host 50 is used to predict blood pressure measurement results according to the Korotkoff sound signals and pressure signals. The piezoelectric sensor 20 and the pressure sensor 40 are respectively connected to the host 50 for communication.

[0050] The present embodiment provides a design scheme for a cuff without an airbag, in which a piezoelectric sensor is disposed in the cuff, and specifically disposed in an air storage chamber surrounded by a first cuff portion and a second cuff portion, so as to reduce or even avoid the influence of the airbag pressure generated by blood flow impact during the blood pressure measurement process on the collection of Korotkoff sound signals, thereby improving the accuracy of Korotkoff sound collection. Compared with conventional technologies, the present invention can not only improve the accuracy of blood pressure measurement results, but the present embodiment can also help reduce materials and reduce the manufacturing cost of the device. The assembly process of the device is simpler, which is convenient for the production and manufacturing of blood pressure measurement devices.

[0051] In some optional implementations, the host 50 includes a pre-processing module, an intermediate processing module, and a post-processing module that are connected in sequence.

[0052] The preprocessing module is used to perform time-frequency transformation on the first time domain feature corresponding to the arterial pulsation extracted from the Korotkoff sound signal to obtain the frequency domain feature, and to splice the frequency domain feature with the second time domain feature extracted from the Korotkoff sound signal to obtain the target data feature.

[0053] The preprocessing module of this embodiment is used to preprocess the collected pressure signal and Korotkoff sound signal to obtain target data features required for subsequent reasoning process.

[0054] The intermediate processing module is used to perform inference based on the target data characteristics to predict the blood pressure category classification result corresponding to the Korotkoff sound signal.

[0055] The intermediate processing module of this embodiment is used to classify the screened Korotkoff sound signals according to the target data characteristics, and infer the blood pressure category corresponding to the Korotkoff sound signals.

[0056] The post-processing module is used to determine the blood pressure measurement result according to the blood pressure category classification result and the pressure signal.

[0057] The post-processing module of this embodiment selects low-pressure signals and high-pressure signals from the pressure signal according to the blood pressure category to achieve the measurement of high blood pressure and low blood pressure.

[0058] The present invention can more accurately capture the change characteristics related to systolic and diastolic pressures by analyzing and processing the frequency domain of Korotkoff sound signals, and ensure the expression of comprehensive and useful information by splicing frequency domain characteristics with time domain characteristics, thereby further improving the accuracy of blood pressure measurement based on the principle of Korotkoff sound method.

[0059] In some optional embodiments, the preprocessing module includes a bandpass filtering unit, a signal extraction unit, a time-frequency transformation unit, a signal sampling unit and a feature splicing unit.

[0060] The bandpass filter unit is used to perform bandpass filtering on the pressure signal to obtain an oscillation wave signal for reflecting arterial pulsation.

[0061] The bandpass filter unit is specifically used to perform bandpass filtering on the pressure signal based on the pulse characteristics to obtain an oscillation wave signal, thereby realizing the function of signal denoising in the filtering process.

[0062] like Figure 3 As shown, the pressure signal collected by the pressure sensor 40 is shown. Figure 3 The horizontal axis represents time, and the unit can be seconds (s), and the vertical axis represents pressure, and the unit can be millimeters of mercury (mmHg). Figure 3 It can be seen that in the process of inflating and pressurizing the air storage chamber in the cuff, the blood flow in the user's upper arm artery will gradually be blocked, the amplitude of the pressure signal will increase, and then the amplitude of the pressure signal will gradually decrease in the process of slowly releasing the gas in the air storage chamber at a preset speed (for example, 2 to 4 mmHg).

[0063] Combination Figure 4 As shown, during the process of inflating the air storage chamber, the piezoelectric sensor detects a piezoelectric signal, and during the process of releasing the gas, the piezoelectric sensor detects a Korotkoff sound signal generated by the arterial pulsation.

[0064] like Figure 5 As shown, the horizontal axis represents time and the unit can be seconds (s), and the vertical axis represents pressure and the unit can be millimeters of mercury (mmHg); when measuring blood pressure, this embodiment focuses on the amplitude decrease process of the pressure signal, that is, the pressure curve diagram of the pressure drop section.

[0065] like Figure 6 As shown, the horizontal axis represents time, the unit can be seconds (s), the vertical axis represents the pressure change value, the unit can be millimeters of mercury (mmHg), and the points in the figure are used to indicate the positions of the maximum values.

[0066] The signal extraction unit is used to extract a first time domain feature corresponding to the arterial pulsation from the Korotkoff sound signal by taking the sampling time corresponding to the maximum value in the oscillation wave signal as a reference time point.

[0067] The signal extraction unit is specifically used to lock the position of the Korotkoff sound signal in the dual pressure sensor data (i.e., the Korotkoff sound signal) according to the maximum value of the shock wave. In this embodiment, each pulse beat is regarded as a moment, and a total of 512 data points (or characteristic points) before and after the position of the Korotkoff sound signal at each moment are extracted.

[0068] like Figure 7 As shown, the horizontal axis represents time, the unit can be seconds (s), and the vertical axis represents the amplitude of the Korotkoff sound signal, the unit can be volts (V). Figure 7On this basis, the exact position of the Korotkoff sound signal is locked by the time point corresponding to the maximum value in the oscillation wave signal.

[0069] like Figure 8 As shown, 25 first time domain features are shown, and each first time domain feature includes 512 data points.

[0070] The time-frequency transform unit is used to perform time-frequency transform processing on the first time domain feature to obtain a frequency domain feature.

[0071] The time-frequency transformation unit is specifically used to perform time-frequency transformation on the time series signal composed of the above 512 data points to obtain 256 data points as frequency domain features.

[0072] The signal sampling unit is used to perform sampling processing on the Korotkoff sound signal to extract the second time domain feature from the Korotkoff sound signal.

[0073] The feature concatenation unit is used to concatenate the frequency domain feature and the second time domain feature into a target data feature.

[0074] In this embodiment, the pressure signal is filtered to obtain an oscillation wave signal, and the position of the Korotkoff sound signal is identified by the maximum value of the oscillation wave signal. Then, time-frequency transformation can be performed on the basis of the locked Korotkoff sound signal, and the target data features for subsequent identification are obtained by splicing frequency domain signals with frequency domain signals. The processing method of the pressure signal and the Korotkoff sound signal in the above process can help to significantly improve the accuracy and reliability of the high and low pressure judgment results.

[0075] In some optional implementations, the signal extraction unit is used to determine a first preset number of first feature points before and after a reference time point extracted from the Korotkoff sound signal as the first time domain feature.

[0076] The first preset number is 512, for example.

[0077] The time-frequency transformation unit is used to determine a second preset number of second feature points obtained by performing time-frequency transformation on the first time-domain features as frequency-domain features.

[0078] The second preset number is smaller than the first preset number; the second preset number is 256, for example.

[0079] In this embodiment, the number of feature points is reduced during the time-frequency transformation process. This method can reduce the hardware requirements of the subsequent processing process while retaining key information, and can improve the processing efficiency of the subsequent process.

[0080] Specifically, the signal sampling unit is used to determine a third preset number of third feature points sampled from the Korotkoff sound signal as the second time domain feature; wherein the third preset number is less than the first preset number, and the third preset number is the same as or different from the second preset number; the third preset number is 256, for example.

[0081] The signal sampling unit can be specifically used to downsample the time series signal composed of the above 512 data points, so as to obtain 256 data points.

[0082] The feature splicing unit is used to splice a third preset number of third feature points and a second preset number of second feature points into a target data feature.

[0083] The feature splicing unit can be specifically used to splice the frequency domain signal represented by the 256 data points obtained by time domain transformation and the time domain signal represented by the 256 data points obtained by downsampling to obtain the target data features represented by the 512 data points obtained by splicing.

[0084] This embodiment retains the time domain key information and the frequency domain key information by splicing the third preset number of third feature points and the second preset number of second feature points, providing sufficient and reliable data support for the accurate identification of high and low voltage categories.

[0085] In some optional implementations, the intermediate processing module is a trained convolutional bidirectional long short-term memory neural network, and the trained convolutional bidirectional long short-term memory neural network includes a convolutional neural network and a bidirectional long short-term memory network connected in sequence;

[0086] The convolutional neural network is used to perform convolution operations on the target data features to extract spatial features from the target data features;

[0087] The bidirectional long short-term memory network is used to output a fourth preset number of blood pressure category classification results based on spatial features. The blood pressure category classification results include classification results between high and low pressure and classification results that are not between high and low pressure. The fourth preset number is the sum of the third preset number and the second preset number.

[0088] In this embodiment, the extracted target data features are input into a trained convolutional bidirectional long short-term memory neural network including a convolutional neural network and a bidirectional long short-term memory network connected in sequence, so as to learn the deep connection between the pressure signal and the high and low pressure signals of Korotkoff sounds through the convolutional bidirectional long short-term memory neural network. This method can greatly improve the accuracy of the blood pressure classification result while retaining the original characteristics of the signal, and transfer the extracted Korotkoff sound signal to the frequency domain for high and low pressure judgment, thereby further improving the accuracy of high and low pressure identification. Moreover, compared with the manual judgment method in the conventional scheme, the convolutional bidirectional long short-term memory neural network adopted in the present invention can also eliminate the problem of inaccurate blood pressure measurement results caused by human subjective factors.

[0089] In some optional embodiments, the bidirectional long short-term memory network includes a long short-term memory network layer and a fully connected layer connected in sequence.

[0090] The long short-term memory network layer is used to extract contextual features from spatial features.

[0091] The fully connected layer is used to output a fourth preset number of blood pressure category classification results based on the context features.

[0092] The present invention comprehensively considers the spatial features and context features in the target data features, thereby fully analyzing and learning the situation of the pressure signal and the Korotkoff sound high and low pressure signal and the relationship between the two, so as to obtain a blood pressure category classification result with higher accuracy.

[0093] In some optional embodiments, the post-processing module is used to determine the pressure signal corresponding to the first blood pressure category classification result between high and low pressure as a high pressure signal, and to determine the pressure signal corresponding to the last blood pressure category classification result between high and low pressure as a low pressure signal.

[0094] This embodiment can determine the high pressure signal and the low pressure signal according to the relationship between the blood pressure category classification result and the pressure signal, that is, realize accurate measurement of the high pressure value and the low pressure value.

[0095] In some optional embodiments, the host 50 includes an analysis module, an identification module, and a judgment module connected in sequence;

[0096] The analysis module is used to extract the target time domain feature corresponding to the arterial pulsation from the Korotkoff sound signal according to the pressure signal, and to determine the signal frequency mean and signal frequency variance corresponding to the target time domain feature;

[0097] The identification module is used to determine the signal change trend according to the signal frequency mean and the signal frequency variance, and is used to determine the high pressure change trend and the low pressure change trend according to the signal change trend;

[0098] The judgment module is used to determine the pressure signal corresponding to the high-pressure change trend as the high-pressure signal, and is used to determine the pressure signal corresponding to the low-pressure change trend as the low-pressure signal.

[0099] The present embodiment can also determine the signal change trend by using the signal frequency mean and signal frequency variance corresponding to the target time domain features extracted from the Korotkoff sound signal, thereby identifying the high pressure signal and the low pressure signal from the pressure signal according to the signal change trend, so as to achieve rapid and accurate determination of the high pressure value and the low pressure value. Moreover, compared with the manual judgment method in the conventional scheme, the analysis module, recognition module and judgment module adopted in the present invention can eliminate the problem of inaccurate blood pressure measurement results caused by human subjective factors.

[0100] According to an embodiment of the present invention, an embodiment of a blood pressure measurement method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0101] In this embodiment, a blood pressure measurement method is provided, which can be used in the above-mentioned host. Fig.12 is a flow chart of a blood pressure measurement method according to an embodiment of the present invention. Fig.12 As shown, the process includes the following steps:

[0102] Step S1201, obtaining a Korotkoff sound signal collected by a piezoelectric sensor fixed inside an air storage chamber and a pressure signal collected by a pressure sensor, wherein the air storage chamber is surrounded by a first cuff portion and a second cuff portion in a cuff, wherein the first cuff portion and the second cuff portion are integrally formed or sealedly connected, and an air nozzle connected to the air storage chamber is provided on the cuff, wherein the air nozzle is used to connect to a first end of an trachea, wherein the trachea is connected to the air storage chamber, and the pressure sensor is connected to a second end of the trachea, and the pressure signal is used to characterize the internal air pressure of the air storage chamber.

[0103] In some optional embodiments, the piezoelectric sensor is connected to the host for communication via a connecting line built into the trachea, and the piezoelectric sensor includes a first piezoelectric sheet and a second piezoelectric sheet attached to a target inner wall of the air storage chamber, where the target inner wall is the inner wall of the air storage chamber close to the inner side of the cuff.

[0104] Specifically, the Korotkoff sound signal of this embodiment is dual piezoelectric sensor data collected by the first piezoelectric sheet and the second piezoelectric sheet.

[0105] Step S1202: predicting the blood pressure measurement result according to the Korotkoff sound signal and the pressure signal.

[0106] Specifically, in the method of measuring blood pressure based on the Korotkoff sound method, a high pressure signal and a low pressure signal are selected from the pressure signal according to the Korotkoff sound signal, thereby obtaining a blood pressure measurement result, i.e., a high pressure value corresponding to the high pressure signal and a low pressure value corresponding to the low pressure signal.

[0107] In this embodiment, a blood pressure measurement method is provided, which can be used in the above host. Fig.13 is a flow chart of a blood pressure measurement method according to an embodiment of the present invention. Fig.13 As shown, the process includes the following steps:

[0108] Step S1301, obtain the Korotkoff sound signal collected by the piezoelectric sensor fixed inside the air storage chamber and the pressure signal collected by the pressure sensor, the air storage chamber is surrounded by the first cuff part and the second cuff part in the cuff, the first cuff part and the second cuff part are integrally formed or sealed, the cuff is provided with an air nozzle connected to the air storage chamber, the air nozzle is used to connect the first end of the trachea, the trachea is connected to the air storage chamber, the pressure sensor is connected to the second end of the trachea, and the pressure signal is used to characterize the internal air pressure of the air storage chamber. For details, please refer to Fig.12 Step S1201 of the illustrated embodiment will not be described in detail here.

[0109] Step S1302: predicting the blood pressure measurement result according to the Korotkoff sound signal and the pressure signal.

[0110] This embodiment provides a blood pressure measurement method based on a non-airbag cuff design, in which a piezoelectric sensor is disposed in the cuff, and specifically disposed in an air storage chamber surrounded by a first cuff portion and a second cuff portion, so as to reduce or even avoid the influence of the airbag pressure generated by blood flow impact on the acquisition of Korotkoff sound signals during the blood pressure measurement process, thereby improving the accuracy of Korotkoff sound acquisition. Compared with conventional technologies, the present invention can significantly improve the accuracy of blood pressure measurement results.

[0111] Specifically, the above step S1302 includes:

[0112] Step S13021: Perform time-frequency transformation on the first time domain feature corresponding to the arterial pulsation extracted from the Korotkoff sound signal to obtain a frequency domain feature.

[0113] In some optional implementations, step S13021 includes but is not limited to step a1 and step a2.

[0114] Step a1, performing bandpass filtering on the pressure signal to obtain an oscillation wave signal for reflecting arterial pulsation.

[0115] Specifically, the pressure signal is subjected to bandpass filtering based on the pulse characteristics to obtain an oscillation wave signal.

[0116] like Figure 3As shown, the pressure signal collected by the pressure sensor 40 is shown. Figure 3 The horizontal axis represents time, and the unit can be seconds (s), and the vertical axis represents pressure, and the unit can be millimeters of mercury (mmHg). Figure 3 It can be seen that in the process of inflating and pressurizing the air storage chamber in the cuff, the blood flow in the user's upper arm artery will gradually be blocked, the amplitude of the pressure signal will increase, and then the amplitude of the pressure signal will gradually decrease in the process of slowly releasing the gas in the air storage chamber at a preset speed (for example, 2 to 4 mmHg).

[0117] Combination Figure 4 As shown, during the process of inflating the air storage chamber, the piezoelectric sensor detects a piezoelectric signal, and during the process of releasing the gas, the piezoelectric sensor detects a Korotkoff sound signal generated by the arterial pulsation.

[0118] like Figure 5 As shown, the horizontal axis represents time and the unit can be seconds (s), and the vertical axis represents pressure and the unit can be millimeters of mercury (mmHg); when measuring blood pressure, this embodiment focuses on the amplitude decrease process of the pressure signal, that is, the pressure curve diagram of the pressure drop section.

[0119] like Figure 6 As shown, the horizontal axis represents time, the unit can be seconds (s), the vertical axis represents the pressure change value, the unit can be millimeters of mercury (mmHg), and the points in the figure are used to indicate the positions of the maximum values.

[0120] Step a2: taking the sampling time corresponding to the maximum value in the oscillation wave signal as the reference time point, extracting the first time domain feature corresponding to the arterial pulsation from the Korotkoff sound signal.

[0121] Specifically, step a2 includes: determining a first preset number of first feature points before and after a reference time point extracted from the Korotkoff sound signal as the first time domain feature.

[0122] The first preset number is 512, for example.

[0123] This embodiment locks the position of the Korotkoff sound signal in the dual pressure sensor data (i.e., the Korotkoff sound signal) based on the maximum value of the oscillation wave. This embodiment takes each pulse beat as a moment and takes out a total of 512 data points (or characteristic points) before and after the position of the Korotkoff sound signal at each moment.

[0124] like Figure 7 As shown, the horizontal axis represents time, the unit can be seconds (s), and the vertical axis represents the amplitude of the Korotkoff sound signal, the unit can be volts (V). Figure 7 On this basis, the exact position of the Korotkoff sound signal is locked by the time point corresponding to the maximum value in the oscillation wave signal.

[0125] like Figure 8 As shown, 25 first time domain features are shown, and each first time domain feature includes 512 data points.

[0126] Step a3: performing time-frequency transformation on the first time domain feature to obtain a frequency domain feature.

[0127] Specifically, step a3 includes: determining a second preset number of second feature points obtained by performing time-frequency transformation on the first time domain feature as frequency domain features; wherein the second preset number is smaller than the first preset number.

[0128] The second preset number is 256, for example.

[0129] In this embodiment, the time series signal composed of the above 512 data points is subjected to time-frequency transformation to obtain 256 data points as frequency domain features. In this embodiment, the number of feature points is reduced during the time-frequency transformation process. This method can reduce the hardware requirements of the subsequent processing process while retaining key information, and can improve the processing efficiency of the subsequent process.

[0130] Step S13022: concatenate the frequency domain feature with the second time domain feature extracted from the Korotkoff sound signal to obtain the target data feature.

[0131] In some optional implementations, step S13022 includes:

[0132] Step b1, sampling the Korotkoff sound signal to extract a second time domain feature from the Korotkoff sound signal.

[0133] Specifically, step b1 includes: determining a third preset number of third feature points sampled from the Korotkoff sound signal as the second time domain feature; wherein the third preset number is smaller than the first preset number, and the third preset number is the same as or different from the second preset number.

[0134] The third preset number is 256, for example.

[0135] This embodiment may also perform downsampling processing on the time series signal composed of the above 512 data points, so as to obtain 256 data points.

[0136] Step b2, concatenating the frequency domain features and the second time domain features into target data features.

[0137] Specifically, step b2 includes: splicing a third preset number of third feature points and a second preset number of second feature points into target data features.

[0138] In this embodiment, the frequency domain signal represented by the 256 data points obtained by time domain transformation and the time domain signal represented by the 256 data points obtained by downsampling are spliced ​​to obtain target data features represented by 512 data points obtained by splicing.

[0139] This embodiment obtains an oscillation wave signal by filtering the pressure signal, thereby identifying the position of the Korotkoff sound signal through the maximum value of the oscillation wave signal, and then performing time-frequency transformation on the locked Korotkoff sound signal, thereby obtaining the target data features for subsequent identification by splicing the frequency domain signal with the frequency domain signal. The processing method of the pressure signal and the Korotkoff sound signal in the above process can help to significantly improve the accuracy and reliability of the high and low pressure judgment results. This embodiment retains the key information in the time domain and the frequency domain by splicing the third preset number of third feature points and the second preset number of second feature points, providing sufficient and reliable data support for the accurate identification of high and low pressure categories.

[0140] Step S13023, inferring based on the target data features to predict the blood pressure category classification result corresponding to the Korotkoff sound signal.

[0141] Specifically, step S13023 includes:

[0142] Step c1, performing a convolution operation on the target data features through the convolutional neural network in the trained convolutional bidirectional long short-term memory neural network to extract spatial features from the target data features.

[0143] Specifically, a convolutional neural network may include a convolution layer, a pooling layer, and an activation function layer; the convolution layer is used to perform convolution operations on the input data using multiple convolution kernels to extract local features; the pooling layer is used to perform pooling operations on the output of the convolution layer to reduce the data dimension while retaining important features; the activation function layer is used to introduce nonlinear factors to enhance the expressive power of the model.

[0144] Step c2, the bidirectional long short-term memory network in the trained convolutional bidirectional long short-term memory neural network outputs a fourth preset number of blood pressure category classification results according to spatial features, the blood pressure category classification results include classification results between high and low pressure and classification results not between high and low pressure, and the fourth preset number is the sum of the third preset number and the second preset number.

[0145] like Fig.10 As shown, in this embodiment, the pulse number N=25 corresponding to the input target data feature W, H represents a 512-dimensional feature, that is, the fourth preset number mentioned above, the channel number C is 1, and is input into the bidirectional long short-term memory network to distinguish each pulse wave.

[0146] Specifically, the bidirectional long short-term memory network may include a forward long short-term memory network layer (forward LSTM layer) and a backward long short-term memory network layer (backward LSTM layer) and an output layer; the forward long short-term memory network layer is used to process the forward propagation of sequence data and capture past information; the backward long short-term memory network layer is used to process the backward propagation of sequence data and capture future information; the output layer is used to predict or classify according to the output of the forward LSTM layer and the backward LSTM layer. In this embodiment, whether the Korotkoff sound signal is between high and low pressure (including high pressure value points and low pressure value points) is output. When the bidirectional long short-term memory network is trained in advance, the training samples will be marked. If it is between high and low pressure values, the output is 1, and if it is between high and low pressure values, the output is zero.

[0147] Specifically, a convolutional bidirectional long short-term memory neural network (CNN-BiLSTM) is a deep learning model, and CNN-BiLSTM specifically combines the advantages of both convolutional neural network (CNN) and bidirectional long short-term memory network (BiLSTM).

[0148] In this embodiment, the extracted target data features are input into a trained convolutional bidirectional long short-term memory neural network including a convolutional neural network and a bidirectional long short-term memory network connected in sequence, so as to learn the deep connection between the pressure signal and the high and low pressure signals of Korotkoff sounds through the convolutional bidirectional long short-term memory neural network. This method can greatly improve the accuracy of the blood pressure classification result while retaining the original characteristics of the signal, and transfer the extracted Korotkoff sound signal to the frequency domain for high and low pressure judgment, thereby further improving the accuracy of high and low pressure identification. Moreover, compared with the manual judgment method in the conventional scheme, the convolutional bidirectional long short-term memory neural network adopted in the present invention can also eliminate the problem of inaccurate blood pressure measurement results caused by human subjective factors.

[0149] This embodiment combines a convolutional neural network with a bidirectional long short-term memory network, which can fully utilize the advantages of the convolutional neural network in better extracting spatial features in the data, and fully utilize the advantages of the bidirectional long short-term memory network in better extracting temporal features in the data, better process target data features in the form of sequence data, and determine a more accurate prediction of the blood pressure category classification result corresponding to the Korotkoff sound signal based on the temporal features and spatial features.

[0150] Specifically, step c2 includes:

[0151] Step d1, extracting contextual features from spatial features through the long short-term memory network layer in the bidirectional long short-term memory network.

[0152] Step d2, outputting a fourth preset number of blood pressure category classification results according to context features through a fully connected layer in a bidirectional long short-term memory network.

[0153] Step S13024, determining the blood pressure measurement result according to the blood pressure category classification result and the pressure signal.

[0154] Specifically, step S13024 includes: determining the pressure signal corresponding to the first blood pressure category classification result between high and low pressure as a high pressure signal, and determining the pressure signal corresponding to the last blood pressure category classification result between high and low pressure as a low pressure signal.

[0155] The present invention can more accurately capture the change characteristics related to systolic and diastolic pressures by analyzing and processing the frequency domain of Korotkoff sound signals, and ensure the expression of comprehensive and useful information by splicing frequency domain characteristics with time domain characteristics, thereby further improving the accuracy of blood pressure measurement based on the principle of Korotkoff sound method.

[0156] In some optional implementations, in the above step S1302, the blood pressure measurement result is predicted based on the Korotkoff sound signal and the pressure signal, including but not limited to steps e1 to e3.

[0157] Step e1, extracting a target time domain feature corresponding to arterial pulsation from the Korotkoff sound signal according to the pressure signal, and determining a signal frequency mean and a signal frequency variance corresponding to the target time domain feature.

[0158] Specifically, this embodiment performs filtering processing on the pressure signal, for example, using a first-order Butterworth bandpass filter to extract the pulsation signal, and finding the maximum value of the pulsation signal. The point where the maximum value is located is the pulse moment, that is, each pulse moment is obtained, for example Figure 6 As shown; Next, the pulse moment is used to locate the corresponding moment of the Korotkoff sound signal, that is, the corresponding moment of the piezoelectric signal. Since the sampling frequencies of the piezoelectric signal and the pressure signal may be different, in order to avoid the deviation caused by direct correspondence, this embodiment searches for the position of the maximum value between the first preset time period (e.g., 0.2 seconds) and the second preset time period (e.g., 0.1 seconds) of the corresponding piezoelectric signal as the Korotkoff sound signal position, for example Figure 7 As shown; according to the Korotkoff sound signal position, a total of 256 points are taken as the target time domain features corresponding to the arterial pulsation, for example Figure 8 The 26 target time domain features shown in the figure are used to process the extracted Korotkoff sound signal through a windowing function, thereby smoothing the edge of the signal to a certain extent so that the signal will not produce too large abrupt changes when it is intercepted; then, the Korotkoff sound signal processed by the windowing function is subjected to a short-time Fourier transform to obtain a corresponding spectrum diagram. In this embodiment, the energy distribution of the Korotkoff sound signal at different frequencies can be visualized through the spectrum diagram, for example Fig. 9 As shown, the data preprocessing is now completed.

[0159] In this embodiment, after the target time domain features are processed by windowing function and short-time Fourier transform, the frequency domain features are obtained. By processing the frequency domain features, the signal frequency mean and signal frequency variance corresponding to each target time domain feature can be directly calculated. Fig.11 As shown, the horizontal axis represents the time domain characteristics of each target, the vertical axis represents the mean and variance, the vertical bar in the figure represents the mean, the vertical line represents the variance, and the point where the vertical bar and the vertical line intersect represents the difference between the mean and the variance.

[0160] Step e2, determining the signal change trend according to the signal frequency mean and the signal frequency variance, and determining the high pressure change trend and the low pressure change trend according to the signal change trend.

[0161] Specifically, this embodiment determines the signal change trend according to the difference between the signal frequency mean and the signal frequency variance or the sum of the signal frequency mean and the signal frequency variance. Fig.11 In the figure, the curve formed by the lines connecting the points where the vertical bars and the vertical lines intersect is the changing trend of the difference between the mean and the variance.

[0162] In this embodiment, the signal change trend is the change trend of the difference between the mean and the variance, or the change trend of the sum of the mean and the variance. Among them, the high-pressure change trend refers to the change trend between the first (i.e., the first) signal frequency mean and the signal frequency variance of the signal change trend, and the low-pressure change trend refers to the change trend between the last (i.e., the last) signal frequency mean and the signal frequency variance of the signal change trend, and the signal frequency variance of the signal change trend.

[0163] Step e3, determining the pressure signal corresponding to the high-pressure change trend as the high-pressure signal, and determining the pressure signal corresponding to the low-pressure change trend as the low-pressure signal.

[0164] Combination Fig.11 As shown, according to the signal acquisition time sequence, it can be judged from the slope of the curve in the figure that the change from target time domain feature 13 to target time domain feature is the most obvious. Therefore, the pressure signal corresponding to target time domain feature 14 is a high-pressure signal. After target time domain feature 23, the changes in the mean and variance of target time domain feature 24 and target time domain feature 25 tend to be gentle, and the pressure signal corresponding to target time domain feature 23 is a low-pressure signal.

[0165] In this embodiment, a host 50 is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0166] This embodiment provides a host 50, combined with Figure 1 As shown, it includes a pre-processing module, an intermediate processing module and a post-processing module connected in sequence.

[0167] The preprocessing module is used to perform time-frequency transformation on the first time domain feature corresponding to the arterial pulsation extracted from the Korotkoff sound signal to obtain the frequency domain feature, and to splice the frequency domain feature with the second time domain feature extracted from the Korotkoff sound signal to obtain the target data feature.

[0168] The intermediate processing module is used to perform inference based on the target data characteristics to predict the blood pressure category classification result corresponding to the Korotkoff sound signal.

[0169] The post-processing module is used to determine the blood pressure measurement result according to the blood pressure category classification result and the pressure signal.

[0170] In some optional embodiments, the preprocessing module includes a bandpass filtering unit, a signal extraction unit, a time-frequency transformation unit, a signal sampling unit and a feature splicing unit.

[0171] The bandpass filter unit is used to perform bandpass filtering on the pressure signal to obtain an oscillation wave signal for reflecting arterial pulsation.

[0172] The signal extraction unit is used to extract a first time domain feature corresponding to the arterial pulsation from the Korotkoff sound signal by taking the sampling time corresponding to the maximum value in the oscillation wave signal as a reference time point.

[0173] The time-frequency transform unit is used to perform time-frequency transform processing on the first time domain feature to obtain a frequency domain feature.

[0174] The signal sampling unit is used to perform sampling processing on the Korotkoff sound signal to extract the second time domain feature from the Korotkoff sound signal.

[0175] The feature concatenation unit is used to concatenate the frequency domain feature and the second time domain feature into a target data feature.

[0176] In some optional embodiments, the signal extraction unit is used to determine a first preset number of first feature points before and after a reference time point extracted from the Korotkoff sound signal as a first time domain feature. The time-frequency conversion unit is used to determine a second preset number of second feature points obtained by performing time-frequency conversion processing on the first time domain feature as a frequency domain feature. The second preset number is less than the first preset number.

[0177] In some optional embodiments, the signal sampling unit is used to determine a third preset number of third feature points sampled from the Korotkoff sound signal as the second time domain feature; wherein the third preset number is less than the first preset number, and the third preset number is the same as or different from the second preset number;

[0178] The feature splicing unit is used to splice a third preset number of third feature points and a second preset number of second feature points into a target data feature.

[0179] In some optional implementations, the intermediate processing module is a trained convolutional bidirectional long short-term memory neural network, and the trained convolutional bidirectional long short-term memory neural network includes a convolutional neural network and a bidirectional long short-term memory network connected in sequence;

[0180] The convolutional neural network is used to perform convolution operations on the target data features to extract spatial features from the target data features.

[0181] The bidirectional long short-term memory network is used to output a fourth preset number of blood pressure category classification results based on spatial features. The blood pressure category classification results include classification results between high and low pressure and classification results that are not between high and low pressure. The fourth preset number is the sum of the third preset number and the second preset number.

[0182] In some optional embodiments, the bidirectional long short-term memory network includes a long short-term memory network layer and a fully connected layer connected in sequence; the long short-term memory network layer is used to extract contextual features from spatial features; and the fully connected layer is used to output a fourth preset number of blood pressure category classification results based on the contextual features.

[0183] In some optional embodiments, the post-processing module is used to determine the pressure signal corresponding to the first blood pressure category classification result between high and low pressure as a high pressure signal, and to determine the pressure signal corresponding to the last blood pressure category classification result between high and low pressure as a low pressure signal.

[0184] In some optional embodiments, the host 50 includes an analysis module, an identification module and a judgment module connected in sequence; the analysis module is used to extract the target time domain features corresponding to the arterial pulsation from the Korotkoff sound signal according to the pressure signal, and to determine the signal frequency mean and the signal frequency variance corresponding to the target time domain features; the identification module is used to determine the signal change trend according to the signal frequency mean and the signal frequency variance, and to determine the high pressure change trend and the low pressure change trend according to the signal change trend; the judgment module is used to determine the pressure signal corresponding to the high pressure change trend as a high pressure signal, and to determine the pressure signal corresponding to the low pressure change trend as a low pressure signal.

[0185] The further functional description of each of the above modules and / or units is the same as that of the above corresponding embodiments and will not be repeated here.

[0186] The host 50 in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0187] The embodiment of the present invention also provides an electronic device having the above Fig.14 The host 50 is shown.

[0188] See also Fig.14 , Fig.14 is a schematic diagram of the structure of an electronic device provided by an optional embodiment of the present invention, such as Fig.14 As shown, the electronic device includes: one or more processors 1410, a memory 1420, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process instructions executed in the electronic device, including instructions stored in or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple electronic devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Fig.14 A processor 1410 is taken as an example.

[0189] The processor 1410 may be a central processing unit, a network processor or a combination thereof. The processor 1410 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0190] The memory 1420 stores instructions executable by at least one processor 1410 so as to enable at least one processor 1410 to implement the method shown in the above embodiment.

[0191] The memory 1420 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the electronic device, etc. In addition, the memory 1420 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 1420 may optionally include a memory remotely arranged relative to the processor 1410, and these remote memories may be connected to the electronic device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0192] The memory 1420 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 1420 may also include a combination of the above types of memory.

[0193] The electronic device also includes a communication interface 1430, which is used for the electronic device to communicate with other devices or a communication network.

[0194] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0195] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of the computer program instruction in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc., and accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0196] In the description of the present invention, it is to be understood that the terms “center”, “longitudinal”, “lateral”, “length”, “width”, “thickness”, “up”, “down”, “front”, “back”, “left”, “right”, “vertical”, “horizontal”, “top”, “bottom”, “inside”, “outside”, “clockwise”, “counterclockwise”, “axial”, “radial”, “circumferential”, etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0197] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0198] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiment, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA, Programmable Gate Array), a field programmable gate array (FPGA, Field Programmable Gate Array), etc.

[0199] In the description of this specification, the description with reference to the terms "this embodiment", "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0200] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0201] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A blood pressure measuring device, characterized in that: The device comprises: A cuff (10) comprises a first cuff portion (11) and a second cuff portion (12), wherein the first cuff portion (11) and the second cuff portion (12) enclose an air storage chamber (101), the first cuff portion (11) and the second cuff portion (12) are integrally formed or sealedly connected, and an air nozzle (102) is provided on the cuff (10), and the air nozzle (102) is connected to the air storage chamber (101); A piezoelectric sensor (20) is fixed inside the gas storage chamber (101), and the piezoelectric sensor (20) is used to collect Korotkoff sound signals; An air pipe (30) having a first end and a second end, wherein the first end is connected to the air nozzle (102), and the air pipe (30) is in communication with the air storage chamber (101); A pressure sensor (40) connected to the second end, the pressure sensor (40) being used to collect a pressure signal, the pressure signal being used to represent the internal air pressure of the air storage chamber (101); The host (50) is used to obtain the Korotkoff sound signal and the pressure signal, and to predict the blood pressure measurement result according to the Korotkoff sound signal and the pressure signal. The piezoelectric sensor (20) and the pressure sensor (40) are respectively connected to the host (50) in communication.

2. The device according to claim 1, characterized in that The piezoelectric sensor (20) is communicatively connected to the host (50) via a connecting line built into the trachea (30), and the piezoelectric sensor (20) comprises a first piezoelectric sheet and a second piezoelectric sheet attached to a target inner wall of the air storage chamber (101), wherein the target inner wall is an inner wall of the air storage chamber (101) close to the inner side of the cuff (10).

3. The device according to claim 1 or 2, characterized in that: The host (50) comprises a pre-processing module, an intermediate processing module and a post-processing module which are connected in sequence; The preprocessing module is used to perform time-frequency transformation processing on the first time domain feature corresponding to the arterial pulsation extracted from the Korotkoff sound signal to obtain a frequency domain feature, and to splice the frequency domain feature with the second time domain feature extracted from the Korotkoff sound signal to obtain a target data feature; The intermediate processing module is used to perform reasoning based on the target data features to predict the blood pressure category classification result corresponding to the Korotkoff sound signal; The post-processing module is used to determine the blood pressure measurement result according to the blood pressure category classification result and the pressure signal.

4. The device according to claim 3, characterized in that The preprocessing module includes a bandpass filtering unit, a signal extraction unit, a time-frequency conversion unit, a signal sampling unit and a feature splicing unit; The bandpass filter unit is used to perform bandpass filtering on the pressure signal to obtain an oscillation wave signal reflecting arterial pulsation; The signal extraction unit is used to extract a first time domain feature corresponding to the arterial pulsation from the Korotkoff sound signal by taking the sampling time corresponding to the maximum value in the oscillation wave signal as a reference time point; The time-frequency conversion unit is used to perform time-frequency conversion processing on the first time domain feature to obtain a frequency domain feature; The signal sampling unit is used to perform sampling processing on the Korotkoff sound signal to extract the second time domain feature from the Korotkoff sound signal; The feature splicing unit is used to splice the frequency domain feature and the second time domain feature into the target data feature.

5. The device according to claim 4, characterized in that The signal extraction unit is used to determine a first preset number of first feature points before and after the reference time point extracted from the Korotkoff sound signal as the first time domain feature; The time-frequency transformation unit is used to determine a second preset number of second feature points obtained by performing time-frequency transformation processing on the first time domain feature as the frequency domain feature; The second preset number is smaller than the first preset number.

6. The device according to claim 5, characterized in that The signal sampling unit is used to determine a third preset number of third feature points sampled from the Korotkoff sound signal as the second time domain feature; wherein the third preset number is less than the first preset number, and the third preset number is the same as or different from the second preset number; The feature splicing unit is used to splice the third preset number of third feature points and the second preset number of second feature points into the target data feature.

7. The device according to claim 6, characterized in that The intermediate processing module is a trained convolutional bidirectional long short-term memory neural network, and the trained convolutional bidirectional long short-term memory neural network includes a convolutional neural network and a bidirectional long short-term memory network connected in sequence; The convolutional neural network is used to perform a convolution operation on the target data features to extract spatial features from the target data features; The bidirectional long short-term memory network is used to output a fourth preset number of blood pressure category classification results based on the spatial features, the blood pressure category classification results including classification results between high and low pressure and classification results not between high and low pressure, and the fourth preset number is the sum of the third preset number and the second preset number.

8. The device according to claim 7, characterized in that The bidirectional long short-term memory network includes a long short-term memory network layer and a fully connected layer connected in sequence; The long short-term memory network layer is used to extract context features from the spatial features; The fully connected layer is used to output the fourth preset number of blood pressure category classification results according to the context features.

9. The device according to claim 7, characterized in that The post-processing module is used to determine the pressure signal corresponding to the first blood pressure category classification result between high and low pressure as a high pressure signal, and to determine the pressure signal corresponding to the last blood pressure category classification result between high and low pressure as a low pressure signal.

10. The device according to claim 1, characterized in that The host (50) comprises an analysis module, an identification module and a judgment module which are connected in sequence; The analysis module is used to extract a target time domain feature corresponding to the arterial pulsation from the Korotkoff sound signal according to the pressure signal, and to determine a signal frequency mean and a signal frequency variance corresponding to the target time domain feature; The identification module is used to determine the signal change trend according to the signal frequency mean and the signal frequency variance, and to determine the high pressure change trend and the low pressure change trend according to the signal change trend; The judgment module is used to determine the pressure signal corresponding to the high-pressure change trend as a high-pressure signal, and to determine the pressure signal corresponding to the low-pressure change trend as a low-pressure signal.

11. A blood pressure measurement method, characterized in that: The method comprises: Acquiring a Korotkoff sound signal collected by a piezoelectric sensor (20) fixed inside an air storage chamber (101) and a pressure signal collected by a pressure sensor (40), wherein the air storage chamber (101) is surrounded by a first cuff portion (11) and a second cuff portion (12) in a cuff (10), wherein the first cuff portion (11) and the second cuff portion (12) are integrally formed or sealedly connected, and the cuff (10) is provided with an air nozzle (102) connected to the air storage chamber (101), wherein the air nozzle (102) is used to connect to a first end of an air pipe (30), wherein the air pipe (30) is connected to the air storage chamber (101), and the pressure sensor (40) is connected to a second end of the air pipe (30), and the pressure signal is used to characterize the internal air pressure of the air storage chamber (101); The blood pressure measurement result is predicted according to the Korotkoff sound signal and the pressure signal.

12. The blood pressure measurement method according to claim 11, characterized in that: The predicting of the blood pressure measurement result according to the Korotkoff sound signal and the pressure signal comprises: Performing time-frequency transformation processing on the first time domain feature corresponding to the arterial pulsation extracted from the Korotkoff sound signal to obtain a frequency domain feature; splicing the frequency domain feature with a second time domain feature extracted from the Korotkoff sound signal to obtain a target data feature; Performing reasoning based on the target data features to predict the blood pressure category classification result corresponding to the Korotkoff sound signal; The blood pressure measurement result is determined according to the blood pressure category classification result and the pressure signal.

13. The blood pressure measurement method according to claim 12, characterized in that: The step of performing a time-frequency transformation process on the first time domain feature corresponding to the arterial pulsation extracted from the Korotkoff sound signal to obtain a frequency domain feature includes: Performing bandpass filtering on the pressure signal to obtain an oscillating wave signal reflecting arterial pulsation; Taking the sampling time corresponding to the maximum value in the oscillation wave signal as a reference time point, extracting a first time domain feature corresponding to the arterial pulsation from the Korotkoff sound signal; Perform time-frequency transformation on the first time domain feature to obtain a frequency domain feature.

14. The blood pressure measurement method according to claim 13, characterized in that: The step of splicing the frequency domain feature with the second time domain feature extracted from the Korotkoff sound signal to obtain a target data feature includes: performing sampling processing on the Korotkoff sound signal to extract the second time domain feature from the Korotkoff sound signal; The frequency domain feature and the second time domain feature are concatenated into the target data feature.

15. The blood pressure measurement method according to claim 14, characterized in that: The step of taking the sampling moment corresponding to the maximum value in the oscillation wave signal as a reference time point and extracting the first time domain feature corresponding to the arterial pulsation from the Korotkoff sound signal comprises: determining a first preset number of first feature points before and after the reference time point extracted from the Korotkoff sound signal as the first time domain feature; The performing time-frequency transformation processing on the first time domain feature to obtain the frequency domain feature includes: determining a second preset number of second feature points obtained by performing time-frequency transformation processing on the first time domain feature as the frequency domain feature; wherein the second preset number is smaller than the first preset number.

16. The blood pressure measurement method according to claim 15, characterized in that: The sampling process of the Korotkoff sound signal to extract the second time domain feature from the Korotkoff sound signal includes: determining a third preset number of third feature points sampled from the Korotkoff sound signal as the second time domain feature; wherein the third preset number is less than the first preset number, and the third preset number is the same as or different from the second preset number; The splicing the frequency domain feature and the second time domain feature into the target data feature includes: splicing the third preset number of third feature points and the second preset number of second feature points into the target data feature.

17. The blood pressure measurement method according to claim 16, characterized in that: The reasoning according to the target data feature to predict the blood pressure category classification result corresponding to the Korotkoff sound signal includes: Performing a convolution operation on the target data features through a convolutional neural network in a trained convolutional bidirectional long short-term memory neural network to extract spatial features from the target data features; The bidirectional long short-term memory network in the convolutional bidirectional long short-term memory neural network completed through the training outputs a fourth preset number of blood pressure category classification results according to the spatial features, and the blood pressure category classification results include classification results between high and low pressure and classification results not between high and low pressure, and the fourth preset number is the sum of the third preset number and the second preset number.

18. The blood pressure measurement method according to claim 17, characterized in that: The bidirectional long short-term memory network in the convolutional bidirectional long short-term memory neural network completed through the training outputs a fourth preset number of blood pressure category classification results according to the spatial features, including: Extracting context features from the spatial features through a long short-term memory network layer in the bidirectional long short-term memory network; The fourth preset number of blood pressure category classification results are output according to the context features through the fully connected layer in the bidirectional long short-term memory network.

19. The blood pressure measurement method according to claim 17, characterized in that: The step of determining the blood pressure measurement result according to the blood pressure category classification result and the pressure signal includes: The pressure signal corresponding to the first blood pressure category classification result between high and low pressure is determined as a high pressure signal, and the pressure signal corresponding to the last blood pressure category classification result between high and low pressure is determined as a low pressure signal.

20. The blood pressure measurement method according to claim 11, characterized in that: The predicting of the blood pressure measurement result according to the Korotkoff sound signal and the pressure signal comprises: Extracting a target time domain feature corresponding to arterial pulsation from the Korotkoff sound signal according to the pressure signal, and determining a signal frequency mean and a signal frequency variance corresponding to the target time domain feature; Determine a signal change trend according to the signal frequency mean and the signal frequency variance, and determine a high pressure change trend and a low pressure change trend according to the signal change trend; The pressure signal corresponding to the high-pressure variation trend is determined as a high-pressure signal, and the pressure signal corresponding to the low-pressure variation trend is determined as a low-pressure signal.

21. An electronic device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the blood pressure measurement method according to any one of claims 11 to 20 by executing the computer instructions.

22. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the blood pressure measurement method according to any one of claims 11 to 20.

23. A computer program product, characterized in that The method comprises computer instructions for causing a computer to execute the blood pressure measurement method according to any one of claims 11 to 20.

Images