CN110974172A

CN110974172A - Real-time physiological parameter measuring system

Info

Publication number: CN110974172A
Application number: CN201911298712.5A
Authority: CN
Inventors: 许志; 杨向林; 张煜; 欧洋; 李延军; 宋晋忠; 杨辉; 叶树明
Original assignee: Information And Communication Office Of General Staff Of Space System Department Of Strategic Support Force Of People's Liberation Army; Beijing Institute of Technology BIT; Suzhou Industrial Technology Research Institute of ZJU; 63919 Troops of PLA
Current assignee: Information And Communication Office Of General Staff Of Space System Department Of Strategic Support Force Of People's Liberation Army; Beijing Institute of Technology BIT; Suzhou Industrial Technology Research Institute of ZJU; 63919 Troops of PLA
Priority date: 2019-12-13
Filing date: 2019-12-13
Publication date: 2020-04-10

Abstract

The embodiment of the invention provides a real-time physiological parameter measuring system, which comprises: dividing a plurality of interrupt time slices representing execution time by a timer of the measurement system; the electrocardio sensor acquires electrocardiosignals according to the interruption time slices; the pulse wave sensor acquires pulse wave signals according to the interruption time slices; the data processing module obtains the physiological parameters of the heartbeat in each beat according to the electrocardio signals and the pulse wave signals; the data processing module specifically executes the following steps: performing feature extraction on electrocardiosignals and pulse wave signals of a plurality of historical testees to obtain first feature data; establishing a multi-physiological parameter measurement model according to the first characteristic data and multi-physiological parameter values of the beat-to-beat of a plurality of historical testees; and extracting the characteristics of the electrocardiosignals and the pulse wave signals of the person to be measured to obtain second characteristic data, and inputting the second characteristic data into the multi-physiological-parameter measurement model to obtain the multi-physiological-parameter values of the beat and the heart of the person to be measured.

Description

Real-time physiological parameter measuring system

Technical Field

The invention relates to the technical field of medical measurement, in particular to a real-time physiological parameter measuring system.

Background

The fast pace of life has caused many people to be in sub-health, which can cause one or more chronic diseases in the human body for a long period of time, and more people now pay more attention to their own physical condition. Therefore, how to realize the measurement of the physiological parameters of the human body becomes an urgent problem to be solved.

Disclosure of Invention

In view of this, the embodiment of the present invention provides a real-time physiological parameter measurement system, so as to implement measurement of human physiological parameters.

In a first aspect of the present invention, a real-time physiological parameter measuring system is provided, which includes: the device comprises an electrocardio sensor, a pulse wave sensor and a data processing module; dividing, by a timer of the measurement system, a plurality of interrupt slots characterizing an execution time; the electrocardio sensor is used for acquiring electrocardiosignals according to the interruption time slices and sending the electrocardiosignals to the data processing module; the pulse wave sensor is used for acquiring pulse wave signals according to the interruption time slices and sending the pulse wave signals to the data processing module; the data processing module is used for obtaining physiological parameters of the beat-to-beat according to the electrocardiosignals and the pulse wave signals; the data processing module specifically executes the following steps: performing feature extraction on electrocardiosignals and pulse wave signals of a plurality of historical testees to obtain first feature data; establishing a multi-physiological parameter measurement model according to the first characteristic data and multi-physiological parameter values of the beat-to-beat of the plurality of historical testees; and extracting the characteristics of the electrocardiosignals and the pulse wave signals of the person to be measured to obtain second characteristic data, and inputting the second characteristic data into the multi-physiological-parameter measurement model to obtain the multi-physiological-parameter values of the beat and the heartbeat of the person to be measured.

With reference to the first aspect, in a first implementation manner of the first aspect, the data processing module includes: the heart rate calculation module is used for carrying out R wave peak value detection on the electrocardiosignals, determining each peak value point corresponding to the electrocardiosignals and calculating the heart rate HR (k) of each beat of the human body by the following formula: hr (k) 60/rr (k), where rr (k) denotes the cardiac cycle of the k-th cardiac cycle.

With reference to the first implementation manner of the first aspect, in a second implementation manner of the first aspect, the data processing module further includes: the pulse wave conduction time calculation module is used for calculating human pulse wave conduction time PTT (k) according to the electrocardiosignals and the pulse wave signals: ptt (k) ═ t₁(k)-t₂(k) Wherein, t₁(k) Time, t, representing the peak point correspondence of the pulse wave of the (k) th heart cycle₂(k) To representThe time corresponding to the peak point of the electrocardiographic R wave of the (k) th heart cycle.

With reference to the second implementation manner of the first aspect, in a third implementation manner of the first aspect, the data processing module further includes: the pulse wave velocity calculation module is used for calculating the human pulse wave velocity PWV (k) according to the pulse wave signal and the following formula: PWV (k) L/T_p(k) Wherein L represents the length of the blood vessel between two points of pulse, T_p(k) Represents the time difference of two paths of pulse wave peaks of the (k) th heart beat period.

With reference to the third implementation manner of the first aspect, in a fourth implementation manner of the first aspect, the data processing module further includes: a total artery compliance value calculating module, for calculating a total artery compliance value C (k) according to the human pulse wave velocity: c (k) ═ a/(ρ × pwv (k)²) Where a represents the artery area and ρ represents the blood density.

With reference to the fourth implementation manner of the first aspect, in a fifth implementation manner of the first aspect, the data processing module further includes: a blood pressure calculating module for calculating human body systolic pressure P according to the human body heart rate and the human body pulse wave conduction time_s(k) Diastolic pressure P_d(k) And the average pressure P_m(k)：P_s(k)＝a₁+b₁/PTT(k)²，P_d(k)＝a₂*HR(k)+b₂/PTT(k)²+c₂，P_m(k)＝(P_s(k)+2P_d(k) B) is/3, wherein a₁、b₁、a₂、b₂And c₂Both represent calibration coefficients.

With reference to the fifth implementation manner of the first aspect, in a sixth implementation manner of the first aspect, the data processing module further includes: a peripheral vascular resistance calculation module for calculating human peripheral vascular resistance R (k) from the systolic pressure, diastolic pressure and total arterial compliance values: r (k) ═ T (T-T)_s)/[C(k)*ln(P_d(k)/P_s(k))]Where T denotes the current time, T_s(k) Indicating the diastolic start time in the k-th heart cycle.

In combination with the sixth embodiment of the first aspectIn an embodiment, in a seventh implementation manner of the first aspect, the data processing module further includes: a stroke volume calculation module and a cardiac output volume calculation module; the stroke volume calculating module is used for calculating the human stroke volume SV (k) according to the peripheral vascular resistance, the average pressure and the cardiac cycle: sv (k) ═ P_m(k) Rr (k)/r (k); the cardiac output calculation module is used for calculating human cardiac output CO (k) according to the human heart rate and the human stroke volume: co (k) ═ hr (k) × sv (k).

With reference to the seventh implementation manner of the first aspect, in an eighth implementation manner of the first aspect, the data processing module further includes: a heart index and stroke index calculation module for calculating a human heart index CI (k) and a human stroke index SI (k) from the human cardiac output, the human stroke volume and the human surface area: ci (k) ═ co (k)/BSA, SI (k) ═ sv (k)/BSA, where BSA denotes the surface area of the human body, BSA ═ a ═ h + b ═ w + c, h denotes the height of the human body, w denotes the weight of the human body, a, b, and c are calibration coefficients, and the calibration may be measured using a regression method or a numerical optimization method, where a, b, and c include: a is 0.0061, b is 0.0128, and c is-0.1529.

With reference to the eighth implementation manner of the first aspect, in the ninth implementation manner of the first aspect, the multi-physiological parameter measurement model is a multi-lead pulse wave signal and multi-parameter depth belief network-based physiological parameter measurement model or a multi-lead pulse wave signal and multi-parameter cyclic neural network-based physiological parameter measurement model.

Compared with the prior art, the technical scheme of the invention at least has the following advantages:

the embodiment of the invention provides a real-time physiological parameter measuring system, which acquires electrocardiosignals through an electrocardio sensor and sends the electrocardiosignals to a data processing module, acquires pulse wave signals through a pulse wave sensor and sends the pulse wave signals to the data processing module, and the data processing module acquires various physiological parameters of each beat of a human body according to the electrocardiosignals and the pulse wave signals, thereby realizing the real-time measurement of the physiological parameters of the human body, being beneficial to knowing the physical condition of the human body and evaluating the cardiac function of the human body.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.

FIG. 1 is a functional block diagram of one particular example of a real-time physiological parameter measurement system according to an embodiment of the present invention;

FIG. 2 is a functional block diagram of another specific example of a real-time physiological parameter measurement system according to an embodiment of the present invention;

FIG. 3 is a functional block diagram of another specific example of a real-time physiological parameter measurement system according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a dynamic Bayesian network of a real-time physiological parameter measurement system according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of a temporal recurrent neural network of a real-time physiological parameter measurement system according to an embodiment of the present invention;

FIG. 6 is a schematic block diagram of a specific example of a real-time physiological parameter measurement system in an embodiment of the present invention;

FIG. 7 is a schematic block diagram of a specific example of a non-invasive hemodynamic parameter measurement apparatus without a sphygmomanometer in an embodiment of the present invention;

FIG. 8 is a schematic block diagram of a specific example of a non-invasive hemodynamic parameter measurement apparatus with a sphygmomanometer in an embodiment of the present invention;

FIG. 9 is a schematic block diagram of a specific example of a non-invasive hemodynamic parameter world-wide link measurement apparatus without a sphygmomanometer in an embodiment of the present invention;

FIG. 10 is a schematic block diagram of a specific example of a non-invasive hemodynamic parameter skywise link measurement apparatus with a sphygmomanometer in an embodiment of the present invention;

FIG. 11 is an information flow diagram of a system for measuring real-time physiological parameters in an embodiment of the present invention;

FIG. 12 is a block diagram of a system for measuring real-time physiological parameters in accordance with an embodiment of the present invention;

FIG. 13 is a block diagram of the general structure of a real-time physiological parameter measurement system according to an embodiment of the present invention;

FIG. 14 is a schematic diagram of the relationship between the real-time physiological parameter measurement system and the power interface of the medical information management host according to an embodiment of the present invention;

FIG. 15 is a schematic diagram of a data interface relationship between a real-time physiological parameter measurement system and a medical information management host according to an embodiment of the present invention;

FIG. 16 is a schematic diagram of the external interface of the hemodynamic parameter acquisition software in an embodiment of the invention;

FIG. 17 is an external interface control flow diagram of hemodynamic parameter acquisition software;

FIG. 18 is a functional block diagram of another specific example of a real-time physiological parameter measurement system in an embodiment of the present invention;

FIG. 19 is a logic flow diagram of one specific example of electrocardiosignal and pulse wave signal acquisition;

FIG. 20 is a logic flow diagram of CSU2-1 reading ADS 8344;

FIG. 21 is a logic flow diagram for CSU2-2 to extinguish PW1 red LED lamp;

FIG. 22 is a logic flow diagram for CSU2-2 to extinguish PW2 red LED lamp;

FIG. 23 is a logic flow diagram for the CSU5-5 to illuminate the PW1 red light;

FIG. 24 is a flowchart of the logic for the CSU5-5 to illuminate the PW2 red light;

FIG. 25 is a logic flow diagram for the CSU5-7 to illuminate the PW1 infrared LED;

FIG. 26 is a logic flow diagram for the CSU5-7 to illuminate the PW2 infrared LED;

FIG. 27 is a logic flow diagram of a CSU5-8 writing data to a USB BUF;

FIG. 28 is a data flow and control flow diagram of hemodynamic parameter acquisition software;

FIGS. 29A-29C are schematic diagrams of extracted features of cardiac electrical signals in an embodiment of the present invention;

FIG. 30 is a schematic diagram of the characteristics of a typical pulse wave signal;

FIG. 31 is a diagram illustrating a Markov blanket based causal feature selection process.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

An embodiment of the present invention provides a real-time physiological parameter measurement system, as shown in fig. 1, the real-time physiological parameter measurement system includes: the electrocardio sensor 1, the pulse wave sensor 2 and the data processing module 3; the electrocardio sensor 1 acquires electrocardiosignals and sends the electrocardiosignals to the data processing module 3; the pulse wave sensor 2 acquires a pulse wave signal and sends the pulse wave signal to the data processing module 3; the data processing module 3 obtains the physiological parameters of the heartbeat in each beat according to the electrocardio signals and the pulse wave signals.

According to the real-time physiological parameter measuring system provided by the embodiment of the invention, the electrocardio signal is obtained through the electrocardio sensor 1 and is sent to the data processing module 3, the pulse wave signal is obtained through the pulse wave sensor 2 and is sent to the data processing module 3, and the data processing module 3 obtains various physiological parameters of each beat heart of a human body according to the electrocardio signal and the pulse wave signal, so that the real-time measurement of the physiological parameters of the human body is realized, and the real-time physiological parameter measuring system is beneficial for people to know the physical condition of the people and evaluate the cardiac function of the people.

In some embodiments of the invention, one path of electrocardio signals, one path of finger end pulse waves and one path of ear lobe pulse wave signals can be adopted for synchronous acquisition, the acquired electrocardio signals and pulse wave signals are uploaded to the data processing module 3 in real time, the two paths of pulse waves are acquired simultaneously, and if one path of signals is poor, the other path of signals can be used for calculating various physiological parameters instead, so that the robustness of the measuring system is enhanced. In addition, the electrocardio signals and the pulse wave signals are continuously collected, and each path of pulse wave signals collects infrared signals and red light signals.

In other embodiments of the present invention, multiple pulse wave signals may be acquired by multiple pulse wave sensors, for example, pulse wave signals of forehead, carotid artery, ear, wrist, finger and fingertip may be acquired.

The real-time physiological parameter measuring system provided by the embodiment of the invention can measure more than ten kinds of physiological parameters of human bodies, including heart rate, pulse wave conduction time, pulse wave conduction speed, total artery compliance value, blood pressure, peripheral vascular resistance, stroke output, cardiac index, stroke index, left heart work and contraction time ratio and the like, comprehensively estimates the cardiac function state of the human body through the physiological parameters, and realizes the non-invasive hemodynamic monitoring of the human body.

An embodiment of the present invention provides a real-time physiological parameter measurement system, as shown in fig. 2, the real-time physiological parameter measurement system includes: a signal acquisition module 21, a feature extraction module 22 and a physiological parameter detection module 23; the signal acquisition module 21 acquires a first electrocardiosignal and a first pulse wave signal, and transmits the first electrocardiosignal and the first pulse wave signal to the feature extraction module 22; the feature extraction module 22 obtains first feature information according to the first cardiac signal and the first pulse wave signal, and transmits the first feature information to the physiological parameter detection module 23; the physiological parameter detection module 23 generates a detection result through a pre-established physiological parameter detection model according to the first characteristic information.

According to the real-time physiological parameter measuring system provided by the embodiment of the invention, the signal acquisition module 21 is used for acquiring the first electrocardiosignal and the first pulse wave signal, the characteristic extraction module 22 is used for extracting the first characteristic information of the first electrocardiosignal and the first pulse wave signal, and the physiological parameter detection module 23 is used for generating a detection result through a pre-established physiological parameter detection model according to the first characteristic information, so that the accuracy of physiological parameter detection is improved.

As shown in fig. 2, the signal acquisition module 21 includes: an electrocardiosignal acquisition module 211 and a pulse wave signal acquisition module 212; the electrocardiosignal acquisition module 211 acquires a first electrocardiosignal and transmits the first electrocardiosignal to the feature extraction module 22; the pulse wave collecting module 12 collects the first pulse wave signal and transmits the first pulse wave signal to the feature extracting module 22.

In an optional embodiment, the electrocardiograph signal acquisition module 211 may be a patch-type electrocardiograph electrode, specifically, may be an electrocardiograph electrode of the type RedDot-2223, and the electrocardiograph electrode is a silver electrode and has good conductivity. The pulse wave signal collecting module 212 can collect two paths of pulse wave signals, one path adopts a transmission type photoelectric pulse wave sensor, the other path adopts a reflection type photoelectric pulse wave sensor, and when the signal of the one path is poor, the other path can be used for collecting the pulse wave signals, so that the robustness of the real-time physiological parameter measuring system is enhanced. In addition, the transmission-type photoelectric pulse wave sensor and the reflection-type photoelectric pulse wave sensor respectively comprise an infrared LED and a red LED, and the transmission-type photoelectric pulse wave sensor and the reflection-type photoelectric pulse wave sensor respectively measure pulse wave signals of a finger tip and an earlobe of a human body, so that the finger tip infrared pulse wave signal, the finger tip red pulse wave signal, the earlobe infrared pulse wave signal and the earlobe red pulse wave signal are respectively extracted. Optionally, in other embodiments of the present invention, the pulse wave signal collecting module 212 may further include a plurality of photoelectric pulse wave sensors respectively used for measuring pulse wave signals of ears, wrists, heads, carotid arteries, and the like, and each of the photoelectric pulse wave sensors collects pulse wave signals of infrared and red light.

As shown in fig. 2, the system for measuring real-time physiological parameters according to the embodiment of the present invention further includes: a signal processing module 4; the signal processing module 4 performs filtering, amplification and analog-to-digital conversion on the first electrocardiosignal and the first pulse wave signal to generate digital signals of the first electrocardiosignal and the first pulse wave signal, and transmits the digital signals of the first electrocardiosignal and the first pulse wave signal to the feature extraction module 22.

As shown in fig. 2, the signal processing module 4 includes: the first-stage differential amplification module 41, the low-pass filtering module 42, the high-pass filtering module 43, the second-stage amplification module 44, and the analog-to-digital conversion module 45 are respectively configured to perform first-stage differential amplification, low-pass filtering, high-pass filtering, second-stage amplification, and analog-to-digital conversion on the first cardiac signal and the first pulse wave signal.

In an optional embodiment, the first feature information includes: the first heart rate information and the first pulse wave propagation time information, as shown in fig. 2, the feature extraction module 22 includes: a first peak point determining module 21, a second peak point determining module 22, a heart rate information generating module 23 and a pulse wave propagation time calculating module 24; the first peak point determining module 21 performs R-wave peak detection on the first electrocardiographic signal to determine a first peak point corresponding to the electrocardiographic signal; the second peak point determining module 22 performs dominant wave peak detection on the first pulse wave signal, and determines a second peak point corresponding to the pulse wave signal; the heart rate information generating module 23 obtains first heart rate information according to a time interval between time points corresponding to adjacent first peak points, specifically, calculates first heart rate information HR according to HR ═ 60/RR, where RR represents a time interval between time points corresponding to adjacent first peak points; the pulse wave propagation time calculation module 24 calculates to obtain first pulse wave propagation time information according to a time interval between a first time point corresponding to the first peak point and a second time point corresponding to a second peak point closest to the first peak point, specifically, a difference value between the first time point corresponding to the first peak point and the second time point corresponding to the second peak point closest to the first peak point, that is, the first pulse wave propagation time.

Specifically, the first peak point determining module 21 includes: the device comprises an R wave peak value acquisition module, a first judgment module and a first peak value point judgment module; the R wave peak value acquisition module acquires all R wave peak value points of the first electrocardiosignal; the first judgment module judges whether the difference value between each R wave peak point and the adjacent R wave peak point is greater than a first preset value, wherein the first preset value is preferably 0.7 times of the average amplitude of the first electrocardiosignal; the first peak point judging module judges that the R wave peak point of which the difference value between the adjacent R wave peak points is greater than a first preset value is the first peak point.

Specifically, the second peak point determining module 22 includes: the device comprises a main wave peak point acquisition module, a second judgment module and a second peak point judgment module; a main wave peak point acquisition module acquires all main wave peak points of the first pulse wave signal; the second judging module judges whether the difference value between each main wave peak point and the adjacent main wave peak point is greater than a second preset value, wherein the second preset value is preferably 0.7 times of the average amplitude of the first pulse wave signal; the second peak point judging module judges the main peak point of which the difference value between the adjacent main peak points is greater than a second preset value as a second peak point.

Optionally, the feature information extraction module 2 may further include: the electrocardiosignal denoising module is used for denoising power frequency interference (50Hz or 60Hz) of external equipment, myoelectric interference (10-300 Hz) in a human body and baseline drift interference (0.05-2 Hz) caused by respiration and motion, and specifically can perform empirical mode decomposition on a electrocardiosignal containing noise based on Hilbert-Huang transformation, perform Hilbert spectrum analysis on a decomposed Intrinsic Mode Function (IMF), wherein the Hilbert spectrum represents complete time-frequency distribution of the signal, the Hilbert-Huang transformation is a new self-adaptive time-frequency analysis method, self-adaptive time-frequency decomposition is performed according to local time-varying characteristics of the signal, considered factors are eliminated, the defects that a traditional method represents non-stationary and nonlinear signals by meaningless harmonic components are overcome, and extremely high time-frequency resolution is obtained, has good time-frequency aggregation.

Optionally, the feature information extraction module 2 may further include: and the pulse wave signal preprocessing module is used for eliminating baseline drift caused by respiratory motion and body displacement. The human pulse rate is 4-5 times of the respiratory rate, the respiratory rate component is usually below 0.8Hz, and the body displacement can also be characterized by a low frequency component, so that the baseline drift is basically a low frequency component. The pulse wave signal preprocessing module can adopt an adaptive filter based on Meyer wavelet, the high-frequency component of the original signal after wavelet decomposition is selected by reference input, the high-frequency component is related to the signal and is not related to noise, and the baseline drift of the pulse wave can be effectively eliminated. As shown in fig. 2, the physiological parameter detecting module 23 includes: a characteristic information receiving module 31 and a physiological parameter calculating module 32; the characteristic information receiving module 31 receives the first characteristic information and transmits the first characteristic information to the physiological parameter calculating module 32; the physiological parameter calculation module 32 inputs the first characteristic information into a pre-established physiological parameter detection model, and calculates to obtain a detection result.

In an alternative embodiment, the physiological parameter detection model is established by the following steps: acquiring multiple groups of second electrocardiosignals, second pulse wave signals and second physiological parameter detection information in a preset time sequence, wherein each group of second electrocardiosignal, second pulse wave signal and second physiological parameter detection information respectively corresponds to different time information; according to the second electrocardiosignal and the second pulse wave signal, second characteristic information is extracted, and the second characteristic information comprises: second heart rate information and second pulse wave transit time information; and training the dynamic Bayesian network model according to the second characteristic information and the second physiological parameter detection information to obtain a physiological parameter detection model.

In the embodiment of the present invention, a dynamic bayesian network model (DBN) is an extension of a bayesian network in the time domain, and is a random model for processing time series data formed by combining an original network structure with time information based on a static bayesian network, as shown in fig. 4. Due to the introduction of the time sequence relationship, the method not only can carry out probability modeling on the relationship among variables in the same time slice, but also can reflect the relationship of variable time sequences among different time slices. The method has the advantages of processing nonlinear relations, uncertain relations and dynamic relations. Definition of Z_t＝[Z_t1,Z_t2...Z_tN]Is a network of N features at time t, including SBP (systolic pressure), DBP (diastolic pressure), MBP (mean blood pressure), etc. Z_tAnd one Bayesian network at the time t is formed, and the Bayesian networks at different times form a dynamic Bayesian network, and the joint probability formula is shown as the following formula.

In an alternative embodiment, the physiological parameter detection model is established by the following steps: acquiring multiple groups of second electrocardiosignals, second pulse wave signals and second physiological parameter detection information in a preset time sequence, wherein each group of second electrocardiosignal, second pulse wave signal and second physiological parameter detection information respectively corresponds to different time information; according to the second electrocardiosignal and the second pulse wave signal, second characteristic information is extracted, and the second characteristic information comprises: second heart rate information and second pulse wave transit time information; and training a time Recurrent Neural Network (RNN) according to the second characteristic information and the second physiological parameter detection information to obtain a physiological parameter detection model.

In practical application, deep learning is successfully applied to the fields of voice recognition, video detection, image recognition and the like. The feedforward network such as the multilayer perceptron and the convolutional neural network assumes that the input is an independent unit without context relation, but the blood pressure of each stroke and the characteristics thereof have obvious time series characteristics, and the output blood pressure is related to the previous blood pressure. Therefore, an RNN model with a certain "memory capacity" is employed.

As shown in fig. 5, the RNN includes an input layer, an output layer, and a hidden layer, and its inter-neuron connections form a directed graph. The RNN memorizes the previous information and applies it to the calculation of the current output, i.e. the nodes between the hidden layers are no longer connected but connected, and the input of the hidden layer includes not only the output of the input layer but also the output of the hidden layer at the previous moment.

In an optional embodiment of the present invention, the acquiring of the second cardiac electric signal, the second pulse wave signal, and the second physiological parameter detection information may be acquiring cardiac electric signals, pulse wave signals, and blood pressure values that have been measured by a plurality of subjects. It should be noted that, the process of extracting the second characteristic information according to the second cardiac signal and the second pulse wave signal may refer to the above description about extracting the first characteristic information, and is not described herein again.

In an optional embodiment, the physiological parameter detection model is obtained by taking the second characteristic information as input, that is, taking the second heart rate information and the second pulse wave transit time as input, taking the second blood pressure information as output, and training by using a deep neural network. It should be noted that, the training of the physiological parameter detection model by using the deep neural network is only a preferred implementation manner provided by the embodiment of the present invention, and other training methods may be selected in practical applications, which is not limited to the present invention.

As shown in fig. 3, the system for measuring real-time physiological parameters according to the embodiment of the present invention further includes: and the display module 5 is used for displaying the detection result.

In another optional embodiment of the present invention, the first characteristic information may include heart rate variability, electrocardiographic signal waveform variability, pulse wave signal waveform dominant wave height, dominant wave rise time, dicrotic wave height, dicrotic wave relative height, central descending isthmus relative height, and cardiac output per stroke, and may further include an analytic characteristic, a representation characteristic, a transform domain characteristic, a fusion characteristic, and the like. The analysis characteristics refer to the characteristics of geometrical parameters such as the amplitude, interval, area and angle of waveforms of the electrocardiosignals and the pulse wave signals; the appearance characteristics refer to that data dimensions of electrocardiosignals and pulse wave signals are reduced by methods such as Principal Component Analysis (PCA), Latent Dirichlet Allocation (LDA), hidden Dirichlet Allocation (KLT-Loeve Transform), Polynomial Distance Measurement (PDM), a functional data analysis method, phase space reconstruction and the like, characteristics suitable for blood pressure estimation are extracted, the information of the whole electrocardiosignals and pulse wave signals is utilized, redundant and unimportant information is removed through transformation, and the data volume is reduced and a large amount of information is utilized; the transform domain characteristics refer to the characteristics of electrocardiosignals and pulse wave signals on various transform domains such as wavelet transform, Fourier transform, Hilbert transform, cosine transform, Hilbert-Huang transform and the like, new effective characteristics can be found by extracting the transform domain characteristics, and the transform domain waveform characteristics are stable and reflect the essential characteristics of the signals; the fusion characteristics are effective characteristics obtained by fusing the analytic characteristics, the appearance characteristics, the transform domain characteristics and the like of the electrocardiosignals and the pulse wave signals, so that more information is utilized to extract more abundant characteristics. Specifically, how to extract the listed first feature information and how to fuse the analysis feature, the appearance feature and the transform domain feature may refer to an implementation method in the prior art, which is not described in detail in the embodiments of the present invention.

Based on the extracted first feature information, the physiological parameter detection module 23 of the embodiment of the present invention receives the first feature information through the feature information receiving module 31, and transmits the first feature information to the physiological parameter calculation module 32; the physiological parameter calculation module 32 inputs the first characteristic information into a pre-established physiological parameter detection model, and calculates to obtain a detection result.

An embodiment of the present invention provides a system for measuring real-time physiological parameters, as shown in fig. 6, the system for measuring real-time physiological parameters includes: a monitoring host 504, an information management system 505, a cuff type sphygmomanometer 503, an electrocardio sensor 501 and at least one pulse wave sensor 502; the electrocardio sensor 501 acquires an electrocardiosignal and transmits the electrocardiosignal to the monitoring host 504; the at least one pulse wave sensor 502 acquires a pulse wave signal and transmits the pulse wave signal to the monitoring host 504; the cuff type sphygmomanometer 503 acquires a first blood pressure signal and transmits the first blood pressure signal to the monitoring host 504; the monitoring host 504 transmits the electrocardiosignal, the pulse wave signal and the first blood pressure signal to the information management system 505; the information management system 505 may be a computer, and obtains a plurality of non-invasive hemodynamic parameters according to the electrocardiographic signal and the pulse wave signal, wherein the plurality of non-invasive hemodynamic parameters include: and the information management system 505 calibrates the second blood pressure signal according to the first blood pressure signal to obtain a calibrated second blood pressure signal.

According to the real-time physiological parameter measuring system provided by the embodiment of the invention, the electrocardiosignals acquired by the electrocardio sensor 501, the pulse wave signals acquired by the pulse wave sensor 502 and the first blood pressure signals acquired by the cuff type sphygmomanometer 503 are transmitted to the information management system 505 through the monitoring host 504, the information management system 505 obtains the second blood pressure signals according to the electrocardiosignals and the pulse wave signals, and calibrates the second blood pressure signals according to the first blood pressure signals to obtain the calibrated second blood pressure signals, so that the accuracy of measuring the dynamic blood pressure is improved.

Fig. 7 to 10 show four embodiments of the real-time physiological parameter measuring system, wherein fig. 7 shows the real-time physiological parameter measuring system without the sphygmomanometer, fig. 8 shows the real-time physiological parameter measuring system with the sphygmomanometer, fig. 9 shows the world link measuring device without the sphygmomanometer, and fig. 10 shows the world link measuring device with the real-time physiological parameter of the sphygmomanometer, which can be used for transmitting the measured real-time physiological parameter of the astronaut to the ground monitoring system 506 through the world link when the astronaut is in orbit flight.

Optionally, in some embodiments of the present invention, the cuff-type sphygmomanometer 503 may transmit the first blood pressure signal to the monitoring host 504 through a wired or wireless communication manner. Specifically, the cuff type sphygmomanometer 503 may be a bluetooth cuff type sphygmomanometer, and transmits the first blood pressure signal to the monitoring host 504 via bluetooth.

Optionally, in other embodiments of the present invention, the cuff-type sphygmomanometer 503 may also be connected to the monitoring host 504 through a serial port and transmit the first blood pressure signal to the monitoring host 504.

Fig. 11 is an information flow diagram of a real-time physiological parameter measurement system, which is an instrument for non-invasively detecting each hemodynamic parameter such as stroke volume, cardiac output volume, peripheral vascular resistance of an astronaut in regular medical examination and transmitting the hemodynamic parameter to a medical information comprehensive management host in real time. As shown in fig. 12, the real-time physiological parameter measuring system is composed of a non-invasive hemodynamic monitor host, a sensor, an on-orbit blood pressure calibration interface, and a dedicated cable including a data line and a power line.

The real-time physiological parameter measuring system only works under the participation of people, and does not work when no people exist. During testing, the astronaut connects an external dynamic blood pressure instrument to the non-invasive hemodynamic monitor host through a serial port, then connects the non-invasive hemodynamic monitor host to the medical information comprehensive management host through a USB interface, and the instrument completes self-checking after being electrified. Selecting or creating an astronaut identifier in-orbit hemodynamic parameter detection display software on a medical information comprehensive management host platform by an astronaut, wearing an electrocardioelectrode and a blood oxygen sensor, and wearing a sleeve according to the requirement of an external dynamic blood pressure instrument; and finally, opening the switch to start testing.

If the medical information comprehensive management host displays that the electrocardio sensor is not connected normally, whether the electrocardio electrode is connected well needs to be checked; if the pulse wave sensor is not normally connected, whether the pulse wave sensor is connected or not needs to be confirmed; if the on-orbit blood pressure is displayed to be abnormal in connection, whether the serial port line is normally connected or not needs to be confirmed, whether the external dynamic blood pressure instrument starts blood pressure detection or not needs to be started, and if all the on-orbit blood pressure lines are normal, an option of acquiring the on-orbit blood pressure in software can be selected or a measurement system host waiting for real-time physiological parameters can automatically acquire the on-orbit blood pressure. During detection, the detected tester keeps a rest state.

The measured physiological curve and the on-orbit blood pressure data are transmitted to a medical information comprehensive management host computer in real time through a USB, the medical information comprehensive management host computer stores the data and timely downloads the data to the ground, ground analysis software corrects the blood pressure per beat in real time according to the blood pressure value obtained by the on-orbit blood pressure calibration interface, and calculates and displays corresponding hemodynamic parameter indexes.

After the test is finished, the tester closes the switch of the external dynamic blood pressure instrument and the switch of the real-time physiological parameter measuring system, takes off the worn sensor, and finishes the detection.

The detection targets of the real-time physiological parameter measuring system are three physiological indexes of continuous blood pressure, stroke cardiac output and blood oxygen saturation. In order to meet the requirements of the performance indexes, a multi-channel 16-bit high-speed analog/digital conversion and acquisition system is designed to realize data acquisition of electrocardio and blood oxygen signals, an electrocardio and blood oxygen saturation processing circuit is developed to realize real-time processing of the signals and feature extraction of the signals, and a CAN bus interface is designed to realize the receiving and sending of the parameters and the real-time signals of a medical information management host.

Fig. 13 is a block diagram of the general structure of the real-time physiological parameter measuring system, and as can be seen from fig. 13, the real-time physiological parameter measuring system mainly comprises the following parts:

a. sensor with a sensor element

A real-time physiological parameter measurement system interfaces with the astronaut through a sensor. The sensor comprises an electrocardio sensor and a pulse wave sensor, the electrocardio electrode is used for detecting an electrocardio waveform, and the pulse wave sensor measures the variation of the light absorption quantity along with the change of the pulsation of the artery and integrates the information of the electrocardio waveform to calculate the wave speed of the pulse wave.

b. Preprocessing circuit

Before analyzing the signals collected by the sensor, the signals must be preprocessed, the electrocardiosignals and the blood oxygen signals must respectively pass through an electrocardiosignal conditioning circuit and a pulse wave signal conditioning circuit to amplify the signals and eliminate the influence of noise and interference, and then the data collection of the electrocardiosignals and the blood oxygen signals is realized through a multi-channel gating circuit and an A/D conversion circuit.

c. Data processing

The core of the real-time physiological parameter measuring system is the design of analyzing digital signals obtained by an A/D conversion circuit, and a microprocessor is required to complete the calculation and measurement of continuous blood pressure, stroke volume and continuous blood oxygen, the transmission of obtained parameters and the management of devices. In order to realize multi-parameter measurement, modular processing is carried out, and the blood pressure per stroke, the stroke volume and the blood oxygen saturation are calculated.

Continuous blood pressure and stroke volume module: the non-invasive continuous blood pressure measurement adopts a pulse wave measurement method, takes the wave crest of an R wave in an electrocardiosignal as a starting point, measures the time required by transmitting the R wave triggering pressure pulse wave to the tail end of a finger, establishes a regression equation of systolic pressure and diastolic pressure, and converts the transmission time into the systolic pressure and the diastolic pressure so as to obtain the continuous blood pressure.

Extracting a pulse wave form coefficient characteristic quantity K value according to the change of the area of the pulse wave chart, wherein the K value formula is as follows:

wherein

Mean arterial pressure, which is equal to the mean of the pulse pressure p (t) in one cardiac cycle. P_s、P_dSystolic and diastolic, respectively. Then the stroke output can be obtained according to a hemodynamics model:

cardiac output:

where Kc is a constant and T is the pulse wave period.

A blood oxygen saturation degree module: the pulse wave sensor detects the absorption coefficient of blood oxygen when two kinds of constant light with specific wavelengths irradiate the finger, and according to the formula of the blood oxygen saturation degree:

a continuously measured blood oxygen saturation can be obtained. Wherein HbO₂、HO₂Respectively, bound oxygen capacity and total bindable oxygen capacity.

d. Astronaut key selection

When the astronaut uses the real-time physiological parameter measurement system, the corresponding keys "a", "B" and "C" must be pressed to distinguish different astronauts such as "a", "B" and "C", so the device needs to be configured with a selection key to distinguish which astronaut uses the device.

e. Switch control

After the astronaut presses the selection key, the astronaut needs to press the switch control key, the power supply is switched on, and the real-time physiological parameter measuring system can start to detect; when the detection time is up, the astronaut presses the switch control button to turn off the power supply, and the noninvasive cardiac function monitor stops detecting.

f. Data transmission

The real-time physiological parameter measuring system is in data communication with the medical information management host through a CAN bus, and a CAN bus interface circuit is required between the two systems. The control software of the non-invasive cardiac function monitor CAN safely and reliably complete the operation key control of the non-invasive cardiac function monitor, the calculation of physiological indexes such as continuous blood pressure, stroke volume, blood oxygen saturation and the like, and the CAN bus interface protocol with the medical information management host. The physiological indexes such as continuous blood pressure, stroke volume, blood oxygen saturation and the like measured by the noninvasive cardiac function monitor are received, stored and displayed through a CAN bus interface of the simulated medical information management host.

In addition, the real-time physiological parameter measuring system provided by the embodiment of the invention is also designed with safety, and comprises the following aspects:

a. the host shell of the real-time physiological parameter measuring system is designed to be smooth and chamfered, so that the astronaut cannot be mechanically damaged during operation;

b. the sensor has smooth edge without burrs, and soft lead, and can not stimulate and damage the skin of astronaut.

In addition, the real-time physiological parameter measurement system provided by the embodiment of the invention is also designed with reliability, and comprises the following aspects:

a. selection and control of components

The selection principle of the component model in the development stage is as follows: the adopted components are aerospace-grade products or military products as similar as possible, and the components which are mature in the first period and are subjected to flight tests are adopted as much as possible.

b. Electromagnetic compatibility design

When the printed board is designed, a power supply and a ground wire are thickened and are independently wired, a data ground and an analog ground are separated, the input end of the power supply is connected with a decoupling capacitor, various signals are respectively wired, and the anti-interference capability is enhanced.

c. Thermal design

The components with low voltage and low power consumption are selected during product design, so that heating is reduced; the layout of the components is reasonable, and the heat dissipation is facilitated.

d. Anti-static design

And the grounding design of the shell of the product is required during product design.

e. Design of resistance science

When designing a product, vibration reduction measures aiming at mechanical factors such as vibration, impact and the like are properly taken.

f. Software reliability design

During software design, redundancy processing is required to be adopted on important data in data processing, and a processing mode of taking 2 out of 3 is adopted on important marks, so that the reliability of the important data and the important marks in software is ensured;

the software requirements can be processed aiming at the abnormal conditions of all functional modules in the software, so that the reliability of a software system is guaranteed when the functional modules are in the abnormal conditions;

software requires to design a software trap, and the software trap comprises functions of enabling an interrupt to be unlocked and enabling a program to return to a program head position, so that a system can be normally recovered, and the reliability of a software system is guaranteed.

Fig. 14 is a schematic diagram of the relationship between the real-time physiological parameter measurement system and the power interface of the medical information management host, which is responsible for supplying ± 6V power to the real-time physiological parameter measurement system. Fig. 15 is a schematic diagram of a data interface relationship between a real-time physiological parameter measurement system and a medical information management host, in which the real-time physiological parameter measurement system communicates with the medical information management host through a CAN bus to transmit various measured parameters to the medical information management host, and the medical information management host downloads the parameters to the ground. Fig. 16 is a schematic diagram of an external interface of hemodynamic parameter acquisition software, fig. 17 is a control flow diagram of the external interface of the hemodynamic parameter acquisition software, the hemodynamic parameter acquisition software is embedded software in a non-invasive hemodynamic monitor, the software uses a real-time physiological parameter measurement system as a hardware platform, and has a main function of information interaction with host software, detecting a cardiac lead signal and two cardiac lead signals of a tester, and transmitting a cardiac lead data and pulse data and a device state parameter set to the host software, each external interface of the hemodynamic parameter acquisition software corresponds to a name and a unique identification number of an external device, and a function description of each external interface is shown in table 1.

TABLE 1

As shown in fig. 18, in an alternative embodiment, the number of the pulse wave sensors 502 is two, including: a first pulse wave sensor 5021 and a second pulse wave sensor 5022; the first pulse wave sensor 5021 is a photoplethysmography sensor, is connected to the monitoring host 504 through a first lead wire, and is used for acquiring a first pulse wave signal at a finger tip; the second pulse wave sensor 5022 is a reflective pulse wave sensor, and is connected to the monitoring host 504 through a second lead wire for collecting a second pulse wave signal of the forehead or earlobe. First pulse wave sensor 5021 and second pulse wave sensor 5022 all include infrared LED and ruddiness LED, and infrared LED can send 940 nm's infrared light, and ruddiness LED can send 660 nm's ruddiness, in addition, can also include the LED that can send other wavelengths. In specific implementation, the first pulse wave sensor 5021 is clamped at the front end of the finger, the second pulse wave sensor 5022 is attached to the forehead of the subject, adhered by an adhesive tape, and further fixed by a bandage to prevent the second pulse wave sensor 5021 from falling off. According to the embodiment of the invention, two paths of pulse wave sensors 502 are adopted to acquire pulse wave signals, and when one path of signals is poor, the other path of signals can be used for acquiring the pulse wave signals, so that the robustness of the real-time physiological parameter measuring system is enhanced. Moreover, the pulse wave signals are simultaneously acquired by the two pulse wave sensors 502, so that the pulse wave conduction velocity PWV can be obtained, specifically, the pulse wave conduction velocity PWV can be obtained by

Calculated, where L represents the length of the blood vessel between two points of pulse, T_pThe method represents the time difference of wave crests of two paths of pulse waves, and the pulse wave conduction speed is an important index for representing the arterial complianceThe pulse wave signal of the sensor 502 obtains the pulse wave conduction velocity PWV, which has important reference value for the hemodynamics research.

Optionally, in other embodiments of the present invention, the real-time physiological parameter measuring system may further include more than three pulse wave sensors 502, and when there is a pulse wave sensor 502 with a poor signal, the pulse wave sensors 502 with other good signals may be used to obtain pulse wave signals, so as to enhance the robustness of the real-time physiological parameter measuring system. Moreover, pulse wave signals of the ear, the wrist, the head, the carotid artery and other parts can be respectively measured by adopting a plurality of pulse wave sensors, and each photoelectric pulse wave sensor acquires two paths of infrared pulse wave signals and red pulse wave signals.

Fig. 19 is a logic flow diagram of a specific example of the above-mentioned electrocardiosignal and pulse wave signal acquisition, and the acquisition process is mainly completed in a timer interrupt service routine. The timer is set to 3000 times/s, the timer interrupt program is divided into a plurality of time slices according to the LEDSTAT value, 12 time slices are taken as an example for illustration, CSU2-1 to CSU2-7 all represent software modules, PW1 and PW2 all represent pulse wave sensors:

a. time slices 0(LED _ ECG0), 3(LED _ ECG1), 6(LED _ ECG2), and 9(LED _ ECG3) control ADS8344 to acquire the electrocardiosignal through CSU 2-1.

b. Time slice 1(LED _ OFF0) and time slice 7(LED _ OFF1) extinguish PW1 red light through CSU2-2 and PW2 red light through CSU 2-3.

c. Time slice 2(LED _ RED _ ON) lights PW1 RED light through CSU2-4 and collects PW1 RED light direct current through CSU 2-1; PW2 red light is illuminated through CSU2-5, and PW2 red direct current is collected through CSU 2-1.

d. Time slice 4(LED _ RED _ HOLD) collects the DC of PW1 RED light through CSU2-1 and writes it to buffer AD _ Buf [6], and collects the DC of PW2 RED light through CSU2-1 and writes it to AD _ Buf [7 ].

e. Time slice 5(LED _ RED _ OK) collects PW1 RED light alternating current through CSU2_1 and writes the RED light alternating current into a buffer AD _ Buf [8], and collects PW2 RED light alternating current through CSU2_1 and writes the RED light alternating current into AD _ Buf [9 ].

f. Time slice 8(LED _ IR _ ON) collects PW1 lamp full black direct current through CSU2_1 and lights PW1 infrared LED through CSU2_ 6; the PW2 lamp full black direct current is collected through the CSU2_1, and the PW2 infrared LED is lightened through the CSU2_ 7.

g. The time slice 10(LED _ IR _ HOLD) collects PW1 infrared direct current through CSU2-1 and writes the direct current into a buffer AD _ Buf [10 ]; PW2 infrared direct current is collected by CSU2-1 and written into AD _ Buf [11 ].

h. The time slice 11(LED _ IR _ OK) respectively collects PW1 infrared communication through CSU2-1 and writes the infrared communication into a buffer AD _ Buf [1 ]; PW2 infrared communication is collected through CSU2-1 and written into AD _ Buf [2 ].

g. When a group of data is sampled, the physiological data is written into the USB buffer.

The timer interrupt frequency is 3000 times/s-250 times/s-12 times/s; the timer is internally divided into 12 interrupt time slices, and each interrupt can only enter one time slice to process the slice program. The electrocardio is collected for 4 times in 12 interruptions, namely the sampling rate of the electrocardio is 1000Hz, and the sampling rate of 2 lead pulse waves is 250Hz, thereby meeting the technical requirements. Through the setting of the interrupt time slices, the following functions are realized:

1) CSU2-1 ADS8344 data acquisition Read _ AD functions to Read physiological signal data from the ADS8344, including 1 lead cardiac and 2 lead pulse waves.

The main information of the detailed design of the electrocardio and pulse wave signal acquisition CSU2-1 is as follows:

a. the sampling frequency of the electrocardiosignal is 1KHz, the data length is 16 bits, the high bit is in front, and the low bit is in back.

b. The sampling frequency of the pulse wave digital signal is 250Hz, the data length is 16bit, and the high order is in front of the low order.

c. When pulse wave signals are collected, the value of the sensor is once when the red light LED is required to be on and the infrared LED is required to be off (the collected value is called red pulse wave for short), the value of the sensor is once when the infrared LED is off and the sensor is used for on (the collected value is called infrared pulse wave for short), and the value of the sensor is twice when the red light LED and the infrared LED are off.

d. The red pulse wave acquisition takes two timer interruptions. And the red light LED is firstly lightened during the previous interruption, the infrared LED is extinguished, and the infrared LED is collected in the next interruption period, so that the stability of the light intensity of the red light LED is ensured, and the interference is reduced.

e. The infrared pulse wave acquisition takes two timer interruptions. The infrared LED is required to be turned on firstly in the previous interruption, the red LED is turned off, and the infrared LED is collected in the next interruption period, so that the stability of the light intensity of the infrared LED is ensured, and the interference is reduced.

f. The length of a data buffer area of the electrocardiosignals and the pulse wave signals is 24K, the buffer area adopts an FIFO mode, and when the buffer area is full, the data written firstly is discarded, and the latest data is filled;

g. error processing: when the buffer is full, the oldest written data is discarded and the latest data is filled.

h. The logic flow chart of the electrocardio-pulse wave signal acquisition CSU3-1 is shown in FIG. 20.

2) The function of the CSU2-2 for turning off the PW1 red LED lamp PW1BlackAllLED is shown in fig. 21, the PW1 red LED and the infrared LED pins are cleared, and the PW1 red LED and the infrared LED are turned off.

3) The CSU2-3 extinguishes the PW2 red LED lamp PW2BlackAllLED, and the PW2 red LED and the infrared LED pins are cleared, and the PW2 red LED and the infrared LED are extinguished, as shown in FIG. 22.

4) The function of the CSU2-4 to light the PW1 red PW1LightRedLED is shown in FIG. 23, the PW1 infrared LED is turned off first, and then the PW1 red LED is turned on.

5) The function of the CSU2-5 to light the PW2 red PW2LightRedLED is shown in FIG. 24, the PW2 infrared LED is turned off first, and then the PW2 red LED is turned on.

6) The function of the CSU2-6 to light the PW1 infrared LED PW1LightIRLED is shown in FIG. 25, namely, the PW1 red LED is turned off first, and then the PW1 infrared LED is turned on.

7) The function of the CSU2-7 for lighting the PW2 infrared LED PW1LightIRLED is shown in FIG. 26, namely, the PW2 red LED is turned off firstly, and then the PW2 infrared LED is turned on.

8) The function of the CSU2-8 writing AD _ Buf data into the USB _ Buf InsertDataToUSBBuf is shown in FIG. 27, and the detailed design of the function is as follows:

a. inputting the elements: none.

b. Outputting an element: electrocardio digital signals and pulse wave digital signals.

c. The sampling frequency of the electrocardio digital signal is 1000Hz, the data length is 16bit, the high bit is in front, and the low bit is in back.

d. The sampling frequency of the pulse wave digital signal is 250Hz, the data length is 16bit, and the high order is in front of the low order.

e. Error processing: when the buffer is full, the oldest written data is discarded and the latest data is filled.

The AD _ buf has a length of 12, and the writing of the physiological data packet into the USB _ buf is 24K.

g. The length of a single physiological data is 24 bytes, and in an information body, a physiological data packet consists of 41 physiological data, and the total length is 984 bytes.

h. The electrocardiosignal and the pulse wave signal are respectively sampled in the interruption of a 3000Hz timer, and the acquired data are respectively stored in a data buffer area.

i. The electrocardio-electrode state signals and the pulse wave sensor state signals are collected every 4s, and the results are stored in electrocardio-electrode falling marks and pulse wave sensor error marks.

j. Feeding dogs is performed every 4 ms.

The information management system 505 obtains a second blood pressure signal from the electrocardiographic signal and the pulse wave signal, first obtains the pulse wave transit time PTT from the electrocardiographic signal and the pulse wave signal, and then fits and establishes a functional relationship between the pulse wave transit time and the blood pressure according to a linear correlation relationship between the pulse wave transit time and the blood pressure, specifically, P may be a + bPTT, where P represents the blood pressure, PTT represents the pulse wave transit time, a and b are fitting coefficients, and a and b may be preset according to experience. The information management system 505 calibrates the second blood pressure signal according to the first blood pressure signal to obtain a calibrated second blood pressure signal, and calibrates the blood pressure signal obtained according to the functional relationship between the pulse wave transit time and the blood pressure, specifically, calibrates the fitting coefficients a and b in the functional relationship to obtain a calibrated blood pressure signal, according to the blood pressure signal obtained by the cuff type sphygmomanometer 503.

In an optional embodiment of the present invention, the electrocardiograph sensor 501 is a metal snap-button patch electrocardiograph electrode, and is connected to the monitoring host 504 through a third lead wire, one end of the third lead wire is detachably connected to the electrocardiograph sensor 501 through a metal snap, and the other end of the third lead wire is detachably connected to the monitoring host 504 through a USB interface. The electrocardiograph sensor 501 may be an electrocardiograph electrode of the type RedDot-2223, which is a silver electrode, has good conductivity, is connected to the electrocardiograph vest, and is attached to the chest of the tester during the specific implementation.

In an optional embodiment, the real-time physiological parameter measurement system provided by the embodiment of the invention can measure non-invasive hemodynamic multi-parameters such as heart rate, blood oxygen saturation and arterial compliance besides blood pressure. In particular, the measurement of heart rate may be according to HR-60/RR intervals, where RR intervals represent heart beat intervals; the measurement of the blood oxygen saturation can be pulse waves measured by a pulse wave sensor, and the blood oxygen saturation parameter is obtained according to the variation and the absolute quantity of the absorbance of light under specific wavelength; total arterial compliance may be based on

Wherein the pulse wave velocity PWV can be obtained by the above-mentioned

And calculating to obtain an artery compliance parameter C, wherein A represents the artery area, and rho represents the blood density, and the artery compliance parameter C can be calculated according to the pulse wave velocity PWV, the artery area A and the blood density rho.

As shown in fig. 18, in an alternative embodiment, the monitoring host 504 includes: signal conditioning chip, sampling chip 44 and processor 45, sampling chip 44 may be the low power consumption chip that adopts ADS1298, and processor 45 may adopt STM32 series. The signal conditioning chip filters and amplifies the pulse wave signals and the electrocardiosignals, and transmits the filtered and amplified pulse wave signals and electrocardiosignals to the sampling chip 44, the sampling chip 44 converts the filtered and amplified pulse wave signals and electrocardiosignals into digital signals, and transmits the digital signals to the processor 45, and the processor 45 transmits the digital signals to the information management system 505 through the USB interface.

As shown in fig. 14, in an alternative embodiment, the signal conditioning chip includes: an electrocardiosignal conditioning chip 41, a first pulse wave signal conditioning chip 42 and a second pulse wave signal conditioning chip 43; the electrocardiosignal conditioning chip 41 filters and amplifies the electrocardiosignals and sends the filtered and amplified electrocardiosignals to the sampling chip 44; the first pulse wave signal conditioning chip 42 filters and amplifies the first pulse wave signal, and sends the filtered and amplified first pulse wave signal to the sampling chip 44; the second pulse wave signal conditioning chip 43 filters and amplifies the second pulse wave signal, and sends the filtered and amplified second pulse wave signal to the sampling chip 44.

In an optional embodiment, the sampling chip 44 may further collect state parameters of the electrocardiograph electrode, the first lead wire, the second lead wire, and the third lead wire, and transmit the state parameters to the processor 45; the processor 45 determines whether the electrocardiograph electrode, the first lead wire, the second lead wire and the third lead wire fall off or not according to the state parameters. It should be noted that, it is the prior art to determine whether the electrocardiograph electrodes and the lead wires fall off through the sampling chip 44, and details thereof are not described herein.

As shown in fig. 18, in an optional embodiment, the monitoring host 504 further includes: a first indicator lamp 46, a second indicator lamp 47, a third indicator lamp 48 and a fourth indicator lamp 49, which are respectively used for indicating the states of the electrocardioelectrode, the first lead wire, the second lead wire and the third lead wire; when the electrocardio-electrode, the first lead wire, the second lead wire or the third lead wire are not fallen off, the processor 45 controls the indicator lights corresponding to the electrocardio-electrode, the first lead wire, the second lead wire or the third lead wire to be on; when the electrocardio electrode, the first lead wire, the second lead wire or the third lead wire fall off, the processor 45 generates an alarm signal, and the indicator lamp corresponding to the fallen electrocardio electrode, the first lead wire, the second lead wire or the third lead wire is controlled to be turned off through the alarm signal. Through the arrangement of the indicator lamp, whether the electrocardio-electrode and the lead wire are abnormal or not can be judged, namely whether the electrocardio-electrode and the lead wire fall off or not and the falling position occur or not can be judged, a foundation is laid for eliminating the abnormality, a basis is provided, and the efficiency of abnormality treatment is improved.

Specifically, whether the electrocardio-electrode and the three lead wires are connected or not can be judged through an I/O port of a processor, two detection points are reserved in each cable interface in the design of a cable interface of the signal lead wire, one point is GND, and the other pin is pulled up by 3.3V and then is connected to a GPIO pin of the CPU. For impedance detection, a self-checking mode and a physiological data acquisition mode, the states of the sensor are actual connection states or not, and the connection states of the sensor are judged under the states. When no lead wire is inserted, the detection pin is pulled up to high level by 3.3V; after the lead line is connected, the pin level ground becomes a low level. Reading GPIO pin information, and if the GPIO pin is at a high level, updating the state of the sensor to be unconnected; if a low level is detected, the status is updated as connected.

As shown in fig. 18, in an optional embodiment, the monitoring host 504 provided in the embodiment of the present invention further includes: and a memory 410 connected to the processor 45 for storing at least one of the electrocardiographic signal, the pulse wave signal, the first blood pressure signal and the second blood pressure signal.

As shown in fig. 18, in an optional embodiment, the monitoring host 504 further includes: and the power supply 411 is connected with the processor 45 and supplies power to the monitoring host through the processor 45.

Fig. 28 is a data flow and control flow diagram of hemodynamic parameter acquisition software, the system for measuring real-time physiological parameters according to the embodiment of the present invention can measure ten kinds of non-invasive hemodynamic multiple parameters of a human body, including heart rate, pulse wave conduction time, pulse wave conduction velocity, total artery compliance value, blood pressure, peripheral vascular resistance, stroke output, cardiac index, stroke index, left cardiac work and contraction time ratio, and the like.

Specifically, the processor includes: the heart rate calculation module is used for carrying out R wave peak value detection on the electrocardiosignals, determining each peak value point corresponding to the electrocardiosignals and calculating the human heart rate HR through the following formula:

HR＝60/RR，

wherein RR denotes a cardiac cycle, i.e. an interval between time points corresponding to adjacent peak points of the electrocardiographic signal.

Specifically, the processor further includes: the pulse rate calculation module is used for carrying out peak detection on the pulse signals, determining each peak point corresponding to the pulse wave signals, and calculating the human heart rate PR according to the following formula:

PR＝60/PP，

wherein PP represents the pulse wave period, i.e. the interval between the time points corresponding to the adjacent peak points of the pulse wave signal.

The processor further comprises: the pulse wave conduction time calculation module is used for calculating the human pulse wave conduction time PTT according to the electrocardiosignals and the pulse wave signals:

PTT＝t₁-t₂，

wherein, t₁Representing the time, t, at which the peak point of the pulse wave corresponds₂And the time corresponding to the peak point of the electrocardio R wave is shown.

The processor further comprises: the pulse wave conduction velocity calculation module is used for calculating the human pulse wave conduction velocity PWV according to the pulse wave signals and the following formula:

PWV＝L/T_p，

wherein L represents the length of the blood vessel between two points of pulse, T_pAnd the time difference of the wave crests of the two pulse waves is shown.

The processor further comprises: the total artery compliance value calculating module is used for calculating a human body total artery compliance value C according to the human body pulse wave conduction velocity:

C＝A/(ρ*PWV²)，

where a represents the artery area and ρ represents the blood density.

The processor further comprises: a blood pressure calculating module for calculating human body systolic pressure P according to human body heart rate and human body pulse wave conduction time_sDiastolic pressure P_dAnd the average pressure P_m：

P_s＝a₁+b₁/PTT²，

P_d＝a₂*HR+b₂/PTT²+c₂，

P_m＝(P_s+2P_d)/3，

Wherein, a₁、b₁、a₂、b₂And c₂All represent calibration coefficients, which can be obtained by respectively calibrating synchronous blood pressure measurement. It should be noted that the above coefficient obtained by the synchronous blood pressure measurement calibration may be obtained by using a calibration method in the prior art, and the present invention is not limited thereto.

The processor further comprises: the peripheral vascular resistance calculation module is used for calculating the peripheral vascular resistance R of the human body according to the systolic pressure, the diastolic pressure and the total arterial compliance value:

R＝(t-T_s)/[C*ln(P_d/P_s)]，

where T denotes the current time, T_sIndicating the diastolic starting time.

The processor further comprises: the stroke volume calculation module is used for calculating the human stroke volume SV according to the peripheral vascular resistance, the average pressure and the cardiac cycle:

SV＝(P_m*RR)/R。

the processor further comprises: the cardiac output calculation module is used for calculating the human cardiac output CO according to the human heart rate and the human stroke volume:

CO＝HR*SV。

the processor further comprises: the heart index and stroke index calculation module is used for calculating a human heart index CI and a human stroke index SI according to the human heart output, the human stroke output and the human surface area:

CI＝CO/BSA，

SI＝SV/BSA，

wherein BSA represents a human body surface area, BSA is 0.0061h +0.0128w-0.1529, h represents a human body height, and w represents a human body weight.

The processor 45 may further include: the left heart work doing calculation module is used for calculating the left heart work doing LCW of the human body according to the ventricular pressure and the cardiac output:

LCW＝(MAP-PAOP)*CO*0.0144，

wherein MAP-PAOP represents ventricular pressure and MAP represents mean pressure (i.e., P as described above)_m) PAOP represents the mean left ventricular pressure (typically 6mmHg) and CO represents cardiac output.

The embodiment of the invention can also calculate contraction time ratio (STR, Systolic TimeRatio) which is the ratio between the myocardial electrical excitation period and the mechanical contraction period, and is calculated by the ratio of the pre-ejection period and the left-heart ejection period in the contraction time period:

STR＝PEP/LVET，

wherein PEP represents a pre-ejection period in the contraction time period, LVET represents left-heart ejection time, and the pre-ejection period and the left-heart ejection time are obtained by detecting pulse wave characteristic points.

According to the real-time physiological parameter measuring system provided by the embodiment of the invention, the collected electrocardiosignals and pulse wave signals are uploaded to the processor in real time, and the heart rate, the pulse wave conduction time, the pulse wave conduction speed, the total artery compliance value, the blood pressure, the peripheral blood vessel pressure, the stroke output quantity, the heart index, the stroke index, the left heart work doing and contraction time ratio of the human body are calculated by the processor, so that various physiological parameters of the human body are obtained, the heart function state of the human body is fully reflected, the human body can know the own physical condition in real time, the non-invasive hemodynamic monitoring of the human body is realized, and an important basis is provided for medical decision making of medical personnel.

The real-time physiological parameter measuring system provided by the embodiment of the invention acquires ECG and pulse wave signals through the sensor, detects the characteristic points in real time from the acquired ECG and pulse wave signals, and extracts the characteristics of each heart cycle. Using the features to calculate multiple physiological parameters for each cardiac cycle, the physiological parameters including: blood pressure, heart rate, stroke volume, cardiac output, peripheral vascular resistance, ejection time, left heart ejection time, cardiac index, stroke index, left heart work done, contraction time ratio, etc. Finally, the physiological parameters are displayed through a display screen, so that a user can conveniently obtain the hemodynamic parameters of each heartbeat in real time. In addition, the non-invasive hemodynamic monitor has a storage function (for example, an SD card is used for realizing the storage function), and stored data comprises acquired ECG and pulse wave physiological signals, extracted features per heart cycle and hemodynamic parameters. The noninvasive hemodynamic monitor is attached with analysis software, the analysis software has a driving function, data can be conveniently transmitted to a computer, and the noninvasive hemodynamic monitor has the functions of displaying, counting, analyzing and the like on the computer.

In the embodiment of the present invention, the cardiac function of the human body may be evaluated comprehensively by using the above multiple physiological indexes, or may be evaluated by using a part of the physiological indexes, which may be selected according to actual needs.

In addition, the real-time physiological parameter measuring system provided by the embodiment of the invention can also obtain a continuous blood pressure value through the collected electrocardiosignals and pulse wave signals, and obtain a continuous blood pressure measuring value by extracting the characteristics of the collected electrocardiosignals and pulse wave signals and inputting the extracted characteristics into a pre-established physiological parameter calculation model.

Specifically, the method may include acquiring electrocardiosignals, pulse wave signals and blood pressure values of a plurality of testees, establishing a physiological parameter calculation model through feature extraction, performing feature extraction on the acquired electrocardiosignals and pulse wave signals of the testees, and inputting the extracted features into the physiological parameter calculation model to obtain continuous blood pressure measurement values.

The features extracted in the process of establishing the physiological parameter calculation model and the features extracted in the process of measuring the blood pressure of the person to be measured can comprise heart rate variability, electrocardiosignal waveform variability, pulse wave signal waveform dominant wave height, dominant wave rise time, counterpulsation wave height, counterpulsation wave relative height, central isthmus descending relative height and cardiac output per stroke, and can also comprise analysis features, appearance features, transformation domain features, fusion features and the like.

The analysis characteristics refer to the characteristics of geometrical parameters such as the amplitude, interval, area and angle of waveforms of the electrocardiosignals and the pulse wave signals; in the embodiment of the present invention, the extracted electrocardiographic signal features are shown in fig. 29A to fig. 29C and table 2, where fig. 29A is a schematic diagram of an amplitude characteristic curve; FIG. 29B is a schematic view of interval characteristics; FIG. 29C is a schematic view of a triangular characteristic curve.

TABLE 2

As shown in fig. 30, the main features of a typical pulse wave signal are: (1) ascending branches (A-B): when the heart contracts, the left ventricle ejects blood to the aorta, so that the aortic blood pressure is rapidly increased, and the aortic blood flow is increased. (2) Descending (B-C): in the later stage of left ventricular ejection, due to the slow speed of ejection, when the inflow blood volume of the aortic root is lower than the outflow blood volume to the periphery, the pressure is reduced, and the elastic retraction of the aortic tube is formed. The curve of segment A-B-C constitutes the dominant wave, whose amplitude and morphology are related to the ejection function of the heart and the aortic pressure changes. (3) The descending isthmus (C), which occurs at the instant the aortic valve closes, has an amplitude that is affected by peripheral resistance and aortic valve function: the descending isthmus raises when peripheral resistance increases and lowers otherwise. (4) Dicrotic wave (D): is a wavelet after the central isthmus is fallen. At the beginning of diastole, the aortic valve suddenly closes and peripheral blood regurgitation causes vasodilation.

In the embodiment of the present invention, the extracted pulse wave signal features mainly include: RI, SI, K values, AmBE, DfAmBE, g, LeBA, TmCpt, and RtH.

Wherein, RI is a reflection coefficient, the larger RI is, the stronger reflected wave is generally, the better the elasticity of the blood vessel is generally, and the expression of RI is:

RI＝b/a，

wherein a is the amplitude of the dominant wave and b is the amplitude of the dicrotic wave.

SI is a hardness coefficient, the larger SI is, the smaller DT is generally, the higher PWV is generally, and thus the vessel wall hardness is generally higher, and the expression of SI is:

SI＝h/DT，

wherein h is the height of the subject.

When the peripheral resistance is low or the vessel wall elasticity is good, the K value is generally small; conversely, when the peripheral resistance and the degree of vessel wall hardening increase, the K value generally increases; generally, the smaller K, the smaller the resistance experienced by the PPG, and the expression of the value of K is:

wherein P is_sIs SBP, P_dIs DBP, P_mIs MBP.

Both SI and RI are related to the precise location of point D. When the peripheral resistance of the blood vessel is too large, the dicrotic wave is often not obvious or even indistinguishable, and the accurate positioning of the D point is very difficult at the moment. In addition, the K value is a macroscopic value, which cannot track the waveform shape change in detail, and PPG waveforms of different shapes may correspond to the same K value.

In the embodiment of the invention, six new indexes for measuring PPG waveform change are defined by the inventor except the three characteristic indexes, and the change rule of the indexes is found by referring to the change rule of the K value. Let the abscissa of the point A on the PPG be A_xAnd the other characteristic point coordinate forms are also the same. Referring to fig. 30, a feature point E is defined_xAnd F_x。

E_x＝B_x+100ms，

F_x＝C_x+160ms，

Wherein, B_xRepresents the abscissa of point B, C, on the PPG_xRepresents the abscissa of the C point on the PPG.

Accordingly, AmBE is the average amplitude of the curve of the BE segment of PPG with the amplitude of the point E as the reference,

wherein M is length (Bx: Ex). The larger AmBE is, the steeper the BE segment curve of the PPG is, and the better the elasticity of the blood vessel wall or the smaller the resistance to the PPG is.

DfAmBE: the dfAmBE is the difference mean of the BE segment curves of the PPG.

Wherein K is length (Bx: Ex-1). The larger the DfAmBE is, the steeper the BE segment curve of the PPG is, and the better the elasticity of the blood vessel wall or the smaller the resistance applied to the PPG is.

g: g is the amplitude difference between the BA' line segment and the PPG curve at the point C. The larger g, the lower the relative position of C, the better the elasticity of the vessel wall or the lower the resistance to PPG.

g＝f_BA'(C_x)-ppg(C_x)，

Wherein f is_BA'(C_x) Indicating the magnitude of the BA' line segment at point C.

LeBA: LeBA is the linear fit error of the BA' curve of PPG.

Wherein N ═ length (Bx: Ax'). The larger the LeBA, the less linear the BA' curve, the better the vessel wall elasticity or the lower the resistance experienced by the PPG.

Tmcpt: TmCpt is the cumulative time above point C in the curve of the CF segment of the PPG. The bigger the TmCpt is, the more obvious the dicrotic wave of the PPG is, and the better the elasticity of the vessel wall is or the smaller the resistance borne by the PPG is.

Wherein Fs represents the sampling frequency of the PPG signal; t represents the number of points in the PPG that are greater than PPG (cx).

RtH: RtH is the ratio of the relative amplitude of the dicrotic wave trough point to the relative amplitude of the main wave peak point. RtH, the less elastic the vessel wall or the less resistive the PPG is.

RtH＝h₂/h₁，

Wherein h is₁Representing the relative amplitude of the main wave peak point; h is₂Representing the relative amplitude of the dicrotic trough points.

The appearance characteristics refer to that data dimensions of electrocardiosignals and pulse wave signals are reduced by methods such as Principal Component Analysis (PCA), Latent Dirichlet Allocation (LDA), hidden Dirichlet Allocation (KLT-Loeve Transform), Polynomial Distance Measurement (PDM), a functional data analysis method, phase space reconstruction and the like, characteristics suitable for blood pressure estimation are extracted, the whole electrocardiosignal and pulse wave signal information is utilized, redundant and unimportant information is removed through transformation, and the data volume is reduced and a large amount of information is utilized.

The transform domain characteristics refer to the characteristics of electrocardiosignals and pulse wave signals in various transform domains such as wavelet transform, Fourier transform, Hilbert transform, cosine transform, Hilbert-Huang transform and the like, new effective characteristics can be found by extracting the transform domain characteristics, and the transform domain waveform characteristics are stable and reflect the essential characteristics of the signals.

The fusion characteristics are effective characteristics obtained by fusing the analytic characteristics, the appearance characteristics, the transform domain characteristics and the like of the electrocardiosignals and the pulse wave signals, so that more information is utilized to extract more abundant characteristics. Specifically, how to extract the listed first feature information and how to fuse the analysis feature, the appearance feature and the transform domain feature may refer to an implementation method in the prior art, which is not described in detail in the embodiments of the present invention.

In the embodiment of the present invention, a plurality of features in feature extraction are used to perform feature selection, and for example, a Markov Blanket (MB) based method is used to perform causal feature selection, so as to select a physiological parameter equation estimation feature and a physiological parameter equation selection feature.

The markov blanket concept stems from a bayesian network in which the markov blanket of a variable T is a minimum subset of variables, given the markov blanket mb (T) of the variable T, other variable conditions on the bayesian network are independent of the variable T. While finding a markov blanket of target nodes has proven to be equivalent to an optimal solution for feature selection.

Markov blanket uniqueness: if bayes network B ═ G; p > where the directed acyclic graph G and the probability distribution P satisfy a loyalty condition with each other, are unique for any T, mb (T), including its parent, child and spouse nodes (parent of child). As shown in fig. 31, the optimal feature combination of the prediction target T at time T in the time series is markov blanket MB (T) ═ B (T-1), T (T-1), a (T), B (T), e (T), s (T)).

The physiological parameter calculation model may be a continuous blood pressure measurement model based on a multi-lead pulse wave signal and a multi-parameter DBN (deep belief Network), or may be a continuous blood pressure measurement model based on a multi-lead pulse wave signal and a multi-parameter RNN (Recurrent Neural Network).

The DBN is the expansion of a Bayesian network in the time field, and is based on a static Bayesian network, and combines the original network structure with time information to form a random model for processing time series data. The time sequence relation is introduced, so that the probability modeling can be performed on the relation among variables in the same time slice, and the relation of variable time sequences among different time slices can be reflected. The method has the advantages of processing nonlinear relations, uncertain relations and dynamic relations. Definition of Z_t＝[Z_t1，Z_t2,…,Z_tn]Is a network of N features at time t, including diastolic, systolic, and mean pressures. Z_tAnd one Bayesian network at the time t is formed, and the Bayesian networks at different times form a dynamic Bayesian network.

Time series modeling of continuous blood pressure estimates was performed using RNN. The feedforward network such as the multilayer perceptron and the convolutional neural network assumes that the input is an independent unit without context relation, but the blood pressure of each stroke and the characteristics thereof have obvious time series characteristics, and the output blood pressure is related to the previous blood pressure. The RNN includes input, output and hidden layers, with connections between neurons forming a directed graph. The RNN memorizes the previous information and applies it to the calculation of the current output, i.e. the nodes between the hidden layers are no longer connected but connected, and the input of the hidden layer includes not only the output of the input layer but also the output of the hidden layer at the previous moment. Besides the advantage of long memory in time series modeling, the RNN method directly takes signals as input, does not need feature extraction, and features of the middle layer are selected by the network, so that the complexity of processes such as feature point detection, feature extraction and the like is simplified, and errors are reduced.

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic Disk, an optical Disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a Flash Memory (Flash Memory), a Hard Disk (Hard Disk Drive, abbreviated as HDD) or a Solid State Drive (SSD), etc.; the storage medium may also comprise a combination of memories of the kind described above.

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 invention, and such modifications and variations fall within the scope defined by the appended claims.

Claims

1. A system for real-time measurement of physiological parameters, comprising: the device comprises an electrocardio sensor, a pulse wave sensor and a data processing module;

dividing, by a timer of the measurement system, a plurality of interrupt slots characterizing an execution time;

the electrocardio sensor is used for acquiring electrocardiosignals according to the interruption time slices and sending the electrocardiosignals to the data processing module;

the pulse wave sensor is used for acquiring pulse wave signals according to the interruption time slices and sending the pulse wave signals to the data processing module;

the data processing module is used for obtaining physiological parameters of the beat-to-beat according to the electrocardiosignals and the pulse wave signals;

the data processing module specifically executes the following steps:

performing feature extraction on electrocardiosignals and pulse wave signals of a plurality of historical testees to obtain first feature data;

establishing a multi-physiological parameter measurement model according to the first characteristic data and multi-physiological parameter values of the beat-to-beat of the plurality of historical testees;

and extracting the characteristics of the electrocardiosignals and the pulse wave signals of the person to be measured to obtain second characteristic data, and inputting the second characteristic data into the multi-physiological-parameter measurement model to obtain the multi-physiological-parameter values of the beat and the heartbeat of the person to be measured.

2. The measurement system of claim 1, wherein the data processing module comprises: the heart rate calculation module is used for carrying out R wave peak value detection on the electrocardiosignals, determining each peak value point corresponding to the electrocardiosignals and calculating the heart rate HR (k) of each beat of the human body by the following formula:

HR(k)＝60/RR(k)，

where rr (k) denotes the cardiac cycle of the k-th heart cycle.

3. The measurement system of claim 2, wherein the data processing module further comprises: the pulse wave conduction time calculation module is used for calculating human pulse wave conduction time PTT (k) according to the electrocardiosignals and the pulse wave signals:

PTT(k)＝t₁(k)-t₂(k)，

wherein, t₁(k) Time, t, corresponding to a peak point of a pulse wave representing the k-th heart cycle₂(k) And the time corresponding to the peak point of the electrocardio R wave of the kth heart cycle is shown.

4. The measurement system of claim 3, wherein the data processing module further comprises: the pulse wave velocity calculation module is used for calculating the human pulse wave velocity PWV (k) according to the pulse wave signal and the following formula:

PWV(k)＝L/T_p(k)，

wherein L represents the length of the blood vessel between two points of pulse, T_p(k) Representing the time difference of two paths of pulse wave peaks in the kth heart cycle。

5. The measurement system of claim 4, wherein the data processing module further comprises: a total artery compliance value calculating module, for calculating a total artery compliance value C (k) according to the human pulse wave velocity:

C(k)＝A/(ρ*PWV(k)²)，

where a represents the artery area and ρ represents the blood density.

6. The measurement system of claim 5, wherein the data processing module further comprises: a blood pressure calculating module for calculating human body systolic pressure P according to the human body heart rate and the human body pulse wave conduction time_s(k) Diastolic pressure P_d(k) And the average pressure P_m(k)：

P_s(k)＝a₁+b₁/PTT(k)²，

P_d(k)＝a₂*HR(k)+b₂/PTT(k)²+c₂，

P_m(k)＝(P_s(k)+2P_d(k))/3，

Wherein, a₁、b₁、a₂、b₂And c₂Both represent calibration coefficients.

7. The measurement system of claim 6, wherein the data processing module further comprises: a peripheral vascular resistance calculation module for calculating human peripheral vascular resistance R (k) from the systolic pressure, diastolic pressure and total arterial compliance values:

R(k)＝(t-T_s)/[C(k)*ln(P_d(k)/P_s(k))]，

where T denotes the current time, T_s(k) Indicating the diastolic start time in the k-th heart cycle.

8. The measurement system of claim 7, wherein the data processing module further comprises: a stroke volume calculation module and a cardiac output volume calculation module;

the stroke volume calculating module is used for calculating the human stroke volume SV (k) according to the peripheral vascular resistance, the average pressure and the cardiac cycle:

SV(k)＝(P_m(k)*RR(k))/R(k)；

the cardiac output calculation module is used for calculating human cardiac output CO (k) according to the human heart rate and the human stroke volume:

CO(k)＝HR(k)*SV(k)。

9. the measurement system of claim 8, wherein the data processing module further comprises: a heart index and stroke index calculation module for calculating a human heart index CI (k) and a human stroke index SI (k) from the human cardiac output, the human stroke volume and the human surface area:

CI(k)＝CO(k)/BSA，

SI(k)＝SV(k)/BSA，

wherein, BSA represents the surface area of human body, BSA ═ a × h + b × w + c, h represents the height of human body, w represents the weight of human body, a, b, c are calibration coefficients, and the calibration can be measured by adopting a regression method or a numerical optimization method, wherein a group of characteristic values of a, b, c include: a is 0.0061, b is 0.0128, and c is-0.1529.

10. The measurement system according to claim 1, wherein the multi-physiological parameter measurement model is a multi-lead pulse wave signal, multi-parameter deep belief network based physiological parameter measurement model or a multi-lead pulse wave signal, multi-parameter cyclic neural network based physiological parameter measurement model.