A non-cuff dynamic blood pressure measurement system

Abstract

The present invention discloses a non-cuff type dynamic blood pressure measurement system. The system includes an acquisition module and a processing module, wherein the processing module includes a heartbeat signal extraction unit, an instantaneous phase extraction unit and a blood pressure prediction unit, wherein the acquisition module acquires the initial heartbeat signal of the measured person; the heartbeat signal extraction unit filters and de-noises the initial heartbeat signal, and extracts the heartbeat signal based on energy information; the instantaneous phase extraction unit extracts the corresponding instantaneous phase feature through Hilbert transform calculation for each valid single heartbeat signal extracted, and selects the instantaneous phase feature set using principal component analysis; the blood pressure prediction unit inputs the instantaneous phase feature set into a blood pressure prediction model, and outputs the predicted blood pressure information, and the blood pressure prediction model is obtained by pre-training a least squares support vector regression model. The present invention realizes dynamic and non-sensitive blood pressure monitoring at the same time, and the blood pressure measurement has high accuracy, which can meet the needs of medical clinical blood pressure measurement.

Description

Technical Field

[0001] The present invention relates to the technical field of biological signal detection, and more specifically, to a non-cuff type dynamic blood pressure measurement system. Background Art

[0002] Blood pressure is an important physiological parameter of the human body. Frequent blood pressure measurement helps in the diagnosis and treatment of many serious diseases (such as heart disease, exhaustion, renal failure, hypertension and hemodialysis). In recent years, the number of people with hypertension has continued to rise, and hypertension caused by various causes has posed a serious threat to people's health. According to the gold standard of international medicine, people with hypertension usually need to measure their blood pressure every 15 minutes, and wearing cumbersome and uncomfortable mobile devices will affect their normal activities. Therefore, exploring non-invasive continuous blood pressure measurement methods that can continuously and dynamically reflect the blood pressure status of the monitored person will help improve the quality of life of the measured person, effectively and flexibly monitor blood pressure changes, and achieve better prevention and diagnosis and treatment effects.

[0003] At present, the intra-arterial blood pressure of the arterial catheter commonly used in clinical blood pressure measurement is regarded as the "gold standard" for continuous blood pressure monitoring due to its timeliness and accuracy; the cuff-type electronic arm sphygmomanometer is the "gold standard" for non-invasive blood pressure monitoring. However, invasive blood pressure measurement will cause discomfort to the person being measured and there are potential complications; and intermittent blood pressure monitoring will lose continuous hemodynamic information. However, many non-invasive continuous blood pressure measurement methods in the prior art, such as pulse wave conduction time, pulse wave conduction velocity, applanation intraocular pressure technology, etc., have the disadvantages of being susceptible to interference, low accuracy, requiring active cooperation in monitoring, and easily causing discomfort to the person being measured. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a non-cuff dynamic blood pressure measurement technology based on the ballistocardiac signal, which can capture the ballistocardiac (BCG) signal of the measured person through an inertial sensor without the measured person feeling anything, thereby achieving the effect of dynamic blood pressure detection and disease prevention and diagnosis.

[0005] According to the first aspect of the present invention, a non-cuff type dynamic blood pressure measurement system is provided. The system includes an acquisition module and a processing module, wherein the processing module includes a heartbeat signal extraction unit, an instantaneous phase extraction unit and a blood pressure prediction unit, wherein the acquisition module is used to acquire the initial heartbeat signal of the subject; the heartbeat signal extraction unit is used to filter and de-noise the initial heartbeat signal, and extract the heartbeat signal based on energy information; the instantaneous phase extraction unit is used to extract the corresponding instantaneous phase feature through Hilbert transform calculation for each valid single heartbeat signal extracted, and use principal component analysis to filter out the instantaneous phase feature set; the blood pressure prediction unit is used to input the obtained instantaneous phase feature set into a blood pressure prediction model, and output the predicted blood pressure information, and the blood pressure prediction model is obtained by pre-training the least squares support vector regression model.

[0006] According to a second aspect of the present invention, there is provided a computer-readable storage medium having a computer program stored thereon, wherein when the program is executed by a processor, the following steps are implemented:

[0007] Collecting the initial cardiac signal of the subject;

[0008] Perform filtering and noise reduction on the initial heartbeat signal, and extract the heartbeat signal based on energy information;

[0009] For each valid single heartbeat signal extracted, the corresponding instantaneous phase feature is extracted by Hilbert transform calculation, and the instantaneous phase feature set is screened out using principal component analysis;

[0010] The obtained instantaneous phase feature set is input into a blood pressure prediction model, and the predicted blood pressure information is output. The blood pressure prediction model is obtained by pre-training a least squares support vector regression model.

[0011] Compared with the prior art, the present invention uses an inertial sensor to capture the cardiac impact signal of the person being measured, extracts the instantaneous phase characteristics of each heartbeat signal after relevant data calculation, and predicts blood pressure through a regression model. This method of blood pressure measurement has high accuracy, meets medical standards, and can meet the needs of daily clinical blood pressure measurement. In addition, the present invention can transmit the real-time monitored blood pressure conditions to the cloud and PC wirelessly to achieve remote medical monitoring, significantly saving the time, space and labor costs of medical monitoring.

[0012] Further features and advantages of the present invention will become apparent from the following detailed description of exemplary embodiments of the present invention with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.

[0014] Figure 1 is a flow chart of a non-cuff type dynamic blood pressure measurement method according to an embodiment of the present invention;

[0015] Figure 2 It is a typical cardiac impact signal and corresponding arterial blood flow schematic diagram in the prior art;

[0016] Figure 3 is a schematic diagram of a non-cuff type dynamic blood pressure measurement system based on a heart ballistometry signal according to an embodiment of the present invention;

[0017] Figure 4 The figure is a schematic diagram of the implementation process of a non-cuff type dynamic blood pressure measurement method based on a cardiac ballistic signal according to an embodiment of the present invention. DETAILED DESCRIPTION

[0018] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present invention unless otherwise specifically stated.

[0019] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.

[0020] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered as part of the specification.

[0021] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0022] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0023] The present invention provides a non-cuff type dynamic blood pressure measurement system based on the heartbeat signal, which can monitor blood pressure more quickly, more comfortably, in real time, non-invasively, non-sensitively and accurately. For ease of understanding, the process of using the system to measure blood pressure is first described, combined with Figure 1 As shown, the specific steps include the following steps.

[0024] Step S1, using a measuring device to collect the cardiac signal of the measured person, filtering and denoising the signal and extracting the heartbeat signal based on energy.

[0025] The measuring device may be embedded with an accelerometer or a piezoelectric film sensor to collect ballistocardiogram (BCG) signals, and collecting signals in this way improves the comfort of the measured person. The present invention is described below using an accelerometer as an example.

[0026] In one embodiment, step S1 includes the following sub-steps:

[0027] S11, install the measuring device on a chair or bed, with the y-axis direction of the accelerometer being from the footsteps to the head. When the person being measured is in a sitting or lying state, the measuring device will continuously collect the BCG signal of the person being measured for 30 seconds.

[0028] S12, using a second-order Butterworth bandpass filter (such as 0.5 Hz-10 Hz) to filter the collected y-axis signal of the accelerometer, that is, the BCG signal, to remove interference from non-heartbeat signals such as breathing and human body movement.

[0029] S13, extracting a valid heartbeat signal based on an energy threshold heartbeat signal extraction algorithm.

[0030] The detection algorithm based on energy threshold first calculates the energy of the initial BCG signal, then observes the energy range of the normal BCG signal, and uses this range as the standard to detect the BCG signal interval and the non-BCG signal interval, where the non-BCG signal comes from environmental factors such as impact and touch. In this way, the interference caused by the external environment can be eliminated, and the effective heartbeat signal can be extracted, thereby improving the accuracy of subsequent blood pressure measurement.

[0031] For example, the energy calculation method is to multiply the integral result of the BCG signal by the BCG signal. The threshold method is used to detect the energy extreme point, and the signal within 0.3s before and 0.5s after the extreme point is the extracted single heartbeat BCG signal. Multiple valid heartbeat BCG signals can be detected for the collected 30s signal.

[0032] For BCG signals, a typical single heartbeat BCG signal and the corresponding arterial blood flow are as follows: Figure 2 As shown, Figure 2 Figure (a) is a typical BCG signal. Figure 2Figure (b) in the figure shows the direction of the force generated by the aortic arch and left ventricular ejection. Due to the relatively large fluctuation amplitude and clearness, the I-wave and J-wave are the main points of interest in the present invention. The I-wave is caused by the blood in the left ventricle being ejected into the ascending aorta toward the head (head direction). According to Newton's third law of motion, the body will produce a reaction force toward the feet (foot direction), that is, the body produces foot-directed movement. Similar to the I-wave, the J-wave is caused by the blood in the descending aorta flowing toward the feet. The body will produce a head-directed reaction force, that is, the body produces head-directed movement;

[0033] Step S2, using Hilbert transform to extract the instantaneous phase of the ballistocardiac signal, and further performing principal component extraction on the instantaneous phase to obtain a final instantaneous phase feature set.

[0034] In one embodiment, step S2 includes the following sub-steps:

[0035] S21, for each valid single heartbeat BCG signal, extract the instantaneous phase through Hilbert transform.

[0036] After the signal is Hilbert transformed, the amplitude of each frequency component in the frequency domain remains unchanged, but the phase will shift by 90°, that is, it lags behind π / 2 for positive frequencies and leads π / 2 for negative frequencies. The Hilbert transform (Hilbert transform, denoted as H here) of a real-valued function is to convolve the signal s(t) with 1 / (πt) to obtain s′(t). Therefore, the Hilbert transform result s′(t) can be interpreted as the output of a linear time invariant system whose input is s(t), and the impulse response of this system is 1 / (πt). Specifically, the Hilbert transform formula is:

[0037]

[0038] in,

[0039] S22, based on the PCA algorithm, the extracted instantaneous phase is further subjected to principal component extraction, that is, the initial instantaneous phase feature is projected into a certain space so that the reconstruction error after projection is minimized, and the first k features with the largest energy are selected, that is, the first k features with the largest eigenvalues ​​are selected.

[0040] Principal component extraction refers to selecting the k feature combinations that can best reconstruct the original signal from a given feature set. Its goal is to retain the relevant features for learning and remove redundant and irrelevant features, thereby reducing the number of features, improving model accuracy, and reducing running time.

[0041] Step S3, using the extracted instantaneous phase feature set, using a regression model to measure blood pressure, and recording the systolic and diastolic blood pressures of each measurement.

[0042] In one embodiment, the regression model adopts the least square support vector regression model (LSSVR), which is a least square support vector regression algorithm based on the principle of support vector machine (Support Vector Machine). In LSSVR, all input samples are mapped from low-dimensional space to high-dimensional space, so as to find a hyperplane as the regression plane, and the samples are distributed on the hyperplane or on both sides of the hyperplane, and the goal is to minimize the error of regression prediction.

[0043] It should be understood that the least squares support vector regression model needs to be pre-trained based on the training data set to obtain model parameters (such as weights, biases, etc.) that meet the set optimization goals, and then the trained model can be used as a blood pressure prediction model for actual blood pressure measurement. When actually used for testing, the model will output the predicted systolic and diastolic blood pressures at the same time. During the measurement process, the person being measured needs to be in a sitting or lying state.

[0044] The training data set includes multiple groups of data samples, each group of data samples includes a screened benchmark feature set and corresponding blood pressure values ​​measured using the "gold standard".

[0045] In one embodiment, when using LSSVR for model training, the optimization target of the model is set to:

[0046]

[0047] Among them, w represents the weight coefficient of the regression model, b represents the bias parameter, and y k represents the true value of the kth sample, β is the penalty parameter of the model, which is set for better robustness, and e k represents the hinge loss corresponding to the k-th sample, The function maps samples from a low-dimensional space to a linearly separable high-dimensional space.

[0048] In order to solve the above optimization objectives, the Lagrange multiplier method can be used to construct the following linear equations about α and b:

[0049]

[0050] where Y = [y1,...,y N ] represents the true value of the sample, α=[α1,...,α N ] represents the Lagrange multiplier (support spectrum), let Then Ω=ZZ T represents the kernel matrix.

[0051] By solving the above linear equations (3), the regression model for blood pressure prediction can be obtained, and its regression prediction can be expressed as the following formula:

[0052]

[0053] in, Represents the predicted value corresponding to sample x.

[0054] Accordingly, the present invention provides a non-cuff type dynamic blood pressure measurement system or device based on the heart ballistic signal, which is used to implement one or more aspects of the above method. Figure 3 As shown, the system generally includes an acquisition module, a processing module (or a blood pressure prediction module), a control module and a display module.

[0055] The acquisition module is used to be placed on a medium in contact with the measured person to obtain the heart ballistic signal. The medium may include a bed, a chair, etc. The acquisition module may use an accelerometer or a piezoelectric film sensor to acquire effective signals.

[0056] The processing module includes: an initial value acquisition unit, which is used to obtain an initial cardiac impact signal, wherein the initial cardiac impact signal is measured by the acquisition module; a heartbeat signal extraction unit, which is used to cut the initial cardiac impact signal into a single heartbeat signal combination and remove the interference of non-heartbeat signals, wherein the non-heartbeat signal includes vibrations caused by non-heartbeats, such as impact on a chair, etc.; a noise reduction filter unit, which is used to perform noise reduction and bandpass filtering on the extracted heartbeat signal; an instantaneous phase extraction unit, which is used to extract the instantaneous phase characteristics corresponding to each heartbeat signal; and a blood pressure prediction unit, which is used to predict the current blood pressure value of the person being measured based on the extracted instantaneous phase characteristics.

[0057] The display module is connected to the processing module and is used to display the blood pressure prediction results in real time. For example, the display module includes a device display, a mobile phone display, or a computer display, which can be used with the application end of the present invention to transmit the prediction results to different terminals through wireless communication technology.

[0058] The control module is used to control the frequency and time of blood pressure measurement and the coordination of other modules. For example, the acquisition module can control the start and end time of signal acquisition through acquisition instructions, or control the relevant parameters of signal acquisition. The control module is implemented by software, processor or FPGA.

[0059] The working principle of the present invention is that the heart pumping blood during the heartbeat cycle can cause the body to produce corresponding movements. The motion signal is picked up by a high-sensitivity sensor and described as a waveform, which is called a ballistocardiogram (BCG). Because the BCG signal records the body movement synchronized with the heartbeat caused by the heart pumping blood, the BCG signal indirectly reflects the heart's power and movement state. Analysis of the BCG signal can obtain relevant vital signs or physiological parameters, such as blood pressure. When the human body is in a sitting or lying state, the body movement caused by the heartbeat can be obtained by a high-sensitivity inertial sensor installed on a chair or a bed. Different BCG signals will be generated under different blood pressures. By extracting the instantaneous phase feature set from the BCG signal, the dynamic characteristics of the blood in this period can be reflected, and the blood pressure information is obtained through the blood pressure regression prediction model, that is, the instantaneous phase feature set is mapped to the corresponding diastolic pressure and systolic pressure.

[0060] See also Figure 4 As shown, the working process of the non-cuff type dynamic blood pressure measurement system of the present invention includes the following steps:

[0061] S401, the acquisition module receives an acquisition instruction from the control module and acquires an initial BCG signal;

[0062] S402, using bandpass filtering to reduce noise of the original signal to improve the signal-to-noise ratio;

[0063] S403, detecting and extracting the heartbeat signal based on the energy threshold detection algorithm, and removing interference from non-heartbeat signals;

[0064] S404, determining whether a valid heartbeat signal is extracted;

[0065] S405, if no valid heartbeat signal is extracted, remind the measured person to adjust his posture, sit or lie still, and then return to step S401 to collect again;

[0066] S406, if the heartbeat signal is extracted, extract the instantaneous phase of the BCG signal using Hilbert transform;

[0067] S407, using PCA principal component analysis to obtain the final instantaneous phase feature set;

[0068] S408, predicting blood pressure using the pre-trained LSSVR regression model, with the input being the filtered instantaneous phase feature set and the output being the diastolic pressure and the systolic pressure;

[0069] S409, determining whether the blood pressure is high or low;

[0070] S410, if the blood pressure is normal, display and record the current blood pressure value, and proceed to the next round of measurement;

[0071] S411, if the blood pressure is abnormal, an alarm is sent to the person being measured, and then the current blood pressure value is displayed and recorded, and the next round of measurement is continued.

[0072] In summary, the present invention has at least the following advantages:

[0073] 1) The inertial sensor is used to capture the cardiac signal of the person being measured, and the blood pressure is predicted by combining it with machine learning to achieve fast and accurate dynamic blood pressure monitoring, that is, it can simultaneously realize dynamic and non-sensitive blood pressure monitoring functions.

[0074] 2) The non-cuff blood pressure measurement method is adopted. Compared with the various cuff blood pressure measurement methods currently widely used, the present invention has the advantages of being faster, simpler and more comfortable.

[0075] 3) It is suitable for the person being measured to be sitting or lying still, which is more conducive to continuous long-term blood pressure monitoring at night, in the office, and during leisure time. At the same time, it dynamically records the blood pressure information at each moment, which is helpful for the dynamic treatment evaluation of diseases such as hypertension.

[0076] 4) Based on the machine learning method, the instantaneous phase characteristics are extracted using the cardiac signal. Compared with the current number of parameters, which is often tens of millions or even hundreds of millions, the regression model of the present invention has fewer parameters, which significantly improves the efficiency of algorithm execution and reduces the time required for blood pressure measurement.

[0077] 5) The blood pressure measurement accuracy of the present invention has reached the medical standard and can meet the needs of daily clinical blood pressure measurement.

[0078] 6) The present invention can transmit the real-time monitored blood pressure of the user to the cloud and PC wirelessly, realizing remote medical monitoring, which significantly saves the time, space and manpower costs of medical monitoring.

[0079] The present invention may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present invention.

[0080] Computer readable storage medium can be a tangible device that can hold and store instructions used by an instruction execution device. Computer readable storage medium can be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific examples (non-exhaustive list) of computer readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, for example, a punch card or a convex structure in a groove on which instructions are stored, and any suitable combination thereof. The computer readable storage medium used here is not interpreted as a transient signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagated by a waveguide or other transmission medium (for example, a light pulse by an optical fiber cable), or an electrical signal transmitted by a wire.

[0081] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0082] The computer program instructions for performing the operation of the present invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, Python, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed entirely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be personalized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present invention.

[0083] Various aspects of the present invention are described herein with reference to the flow charts and / or block diagrams of the methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each box of the flow chart and / or block diagram and the combination of each box in the flow chart and / or block diagram can be implemented by computer-readable program instructions.

[0084] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processor of the computer or other programmable data processing device, a device that implements the functions / actions specified in one or more boxes in the flowchart and / or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing device, and / or other equipment to work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0085] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0086] The flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to multiple embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of an instruction, and the module, a program segment or a part of an instruction contains one or more executable instructions for realizing the specified logical function. In some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of the boxes in the block diagram and / or the flowchart can be implemented by a dedicated hardware-based system that performs the specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions. It is well known to those skilled in the art that it is equivalent to implement it by hardware, implement it by software, and implement it by combining software and hardware.

[0087] Embodiments of the present invention have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the marketplace, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein. The scope of the present invention is defined by the appended claims.

Claims

1. A non-cuff type dynamic blood pressure measurement system, comprising an acquisition module and a processing module, wherein the processing module comprises a heartbeat signal extraction unit, an instantaneous phase extraction unit and a blood pressure prediction unit, wherein: The acquisition module is used to acquire the initial cardiac impact signal of the subject; the heartbeat signal extraction unit is used to filter and reduce noise on the initial cardiac impact signal, and extract the heartbeat signal based on energy information; the instantaneous phase extraction unit is used to extract the corresponding instantaneous phase feature through Hilbert transform calculation for each valid single heartbeat signal extracted, and use principal component analysis to screen out the instantaneous phase feature set; the blood pressure prediction unit is used to input the obtained instantaneous phase feature set into the blood pressure prediction model, and output the predicted blood pressure information, and the blood pressure prediction model is obtained by pre-training the least squares support vector regression model; The acquisition module is provided with a measuring device, which is embedded with a cardiac impact sensor and can be installed on a chair or a bed. When the measured person is in a sitting or lying state, the measuring device is configured to continuously collect the cardiac impact signal of the measured person for a period of time to collect the initial cardiac impact signal; Wherein, the heartbeat signal extraction unit performs: The acquired initial heartbeat signal is filtered using a second-order Butterworth bandpass filter to eliminate non-heartbeat signals; The energy extreme point is detected by using the threshold method, and the signal within the set time range is intercepted with the energy extreme point as a reference as the extracted single heartbeat cardiac impulse signal, wherein the energy calculation method is to multiply the integral result of the cardiac impulse signal by the cardiac impulse signal; The instantaneous phase extraction unit obtains the instantaneous phase feature set according to the following steps: For each valid single heartbeat impulse signal, the instantaneous phase is extracted by Hilbert transform calculation; The principal component extraction is performed on the extracted instantaneous phase to project the initial instantaneous phase features into space so that the reconstruction error after projection is minimized, thereby selecting the first k features with the largest eigenvalues ​​as the screened instantaneous phase feature set.

2. The system according to claim 1, characterized in that The training of the least squares support vector regression model includes: Constructing a training data set, which includes multiple groups of data samples, each group of samples reflects the correspondence between the instantaneous phase feature set and the blood pressure value measured using the "gold standard"; The set objective function is used as the optimization target, the least squares support vector regression model is trained using the training data set, and the trained model is used as the blood pressure prediction model.

3. The system according to claim 2, characterized in that The objective function is set as: Among them, w represents the weight coefficient, b represents the bias parameter, and y k represents the true value of the kth sample, β is the penalty parameter, e k represents the hinge loss corresponding to the k-th sample, The function maps samples from a low-dimensional space to a linearly separable high-dimensional space.

4. The system according to claim 1, characterized in that The heart impact sensor includes an accelerometer or a piezoelectric film sensor.

5. The system according to claim 2, characterized in that It also includes a display module and a control module, wherein the display module is communicatively connected with the processing module for displaying the blood pressure prediction results; the control module is communicatively connected with the display module, the acquisition module and the processing module for controlling the frequency and time of blood pressure measurement and controlling the coordination between the relevant modules.

6. The system according to claim 1, characterized in that The time range is set to be 0.3s before and 0.5s after the energy extreme value point.

7. A computer-readable storage medium having a computer program stored thereon, wherein: When the program is executed by the processor, the following steps are implemented: Collecting the initial cardiac signal of the subject; The initial cardiac impact signal is filtered and denoised, and the heartbeat signal is extracted based on the energy information, including: using a second-order Butterworth bandpass filter to filter the collected initial cardiac impact signal to eliminate non-heartbeat signals; using a threshold method to detect the energy extreme point, using the energy extreme point as a reference, intercepting the signal within a set time range as the extracted single heartbeat cardiac impact signal, wherein the energy calculation method is to multiply the integral result of the cardiac impact signal by the cardiac impact signal; For each valid single heartbeat signal extracted, the corresponding instantaneous phase feature is extracted by Hilbert transform calculation, and the instantaneous phase feature set is screened out using principal component analysis; The obtained instantaneous phase feature set is input into a blood pressure prediction model, and the predicted blood pressure information is output, wherein the blood pressure prediction model is obtained by a pre-trained least squares support vector regression model; The initial cardiac impact signal is collected by a measuring device, which is embedded with a cardiac impact sensor and can be installed on a chair or a bed. When the measured person is in a sitting or lying state, the measuring device is configured to continuously collect the cardiac impact signal of the measured person for a period of time to collect the initial cardiac impact signal; The instantaneous phase feature set is screened according to the following steps: For each valid single heartbeat impulse signal, the instantaneous phase is extracted by Hilbert transform calculation; The principal component extraction is performed on the extracted instantaneous phase to project the initial instantaneous phase features into space so that the reconstruction error after projection is minimized, thereby selecting the first k features with the largest eigenvalues ​​as the screened instantaneous phase feature set.

Images