WO2019096175A1 - 一种生命体征信号分析处理方法和生命体征监测设备 - Google Patents

一种生命体征信号分析处理方法和生命体征监测设备 Download PDF

Info

Publication number
WO2019096175A1
WO2019096175A1 PCT/CN2018/115490 CN2018115490W WO2019096175A1 WO 2019096175 A1 WO2019096175 A1 WO 2019096175A1 CN 2018115490 W CN2018115490 W CN 2018115490W WO 2019096175 A1 WO2019096175 A1 WO 2019096175A1
Authority
WO
WIPO (PCT)
Prior art keywords
domain signal
vital sign
respiratory
rate
frequency
Prior art date
Application number
PCT/CN2018/115490
Other languages
English (en)
French (fr)
Inventor
叶飞
胡峻浩
杨超
Original Assignee
深圳市大耳马科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳市大耳马科技有限公司 filed Critical 深圳市大耳马科技有限公司
Publication of WO2019096175A1 publication Critical patent/WO2019096175A1/zh

Links

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
    • A61B5/024—Detecting, measuring or recording pulse rate or heart rate
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/08—Detecting, measuring or recording devices for evaluating the respiratory organs
    • A61B5/0816—Measuring devices for examining respiratory frequency
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B2562/00—Details of sensors; Constructional details of sensor housings or probes; Accessories for sensors
    • A61B2562/02—Details of sensors specially adapted for in-vivo measurements
    • A61B2562/0247—Pressure sensors

Definitions

  • the invention belongs to the field of signal processing, and in particular relates to a vital sign signal analysis processing method and a vital sign monitoring device.
  • Electrocardiogram is a more effective monitoring method, which can accurately capture the electrical activity signal of each beat of the human heart, and then calculate the beat-by-beat heart rate and the waveform characteristics of each beat. Based on this, the heart rate variability can be analyzed.
  • Heart Rate Variability HRV
  • characteristic information of the amplitude or width of the wave group and the respiratory rate obtained by the amplitude feature extraction of the R wave.
  • HRV Heart Rate Variability
  • characteristic information of the amplitude or width of the wave group characteristic information of the amplitude or width of the wave group
  • the respiratory rate obtained by the amplitude feature extraction of the R wave.
  • ECG monitoring requires the application of electrode pads on the subject, which imposes certain constraints on the daily activities of the subject. There may be some doubts or rejections or even a psychological burden on the body's psychological feeling of attaching the sensor, especially when using multi-lead. It is more intense in the case where a plurality of electrode sheets are to be pasted.
  • Photoplethysmograph can also monitor vital signs of respiratory rate and heart rate (in this case, pulse rate).
  • PPG signal acquisition is easily interfered by many factors, such as the wearing position of the sensor, the wearer's skin, and the artifacts caused by the movement of the wearing part, which may cause signal interference, resulting in inaccurate detection of pulse rate per pulse;
  • the calculated respiration rate requires a long-term stable high-quality signal, which needs to be extracted from the high-quality PPG signal to calculate the low-frequency breathing profile, which is very susceptible to subtle low-frequency interference and affects the calculation; and PPG still needs to wear the sensor (finger clip, ear) Clip, nose clip, etc.), although the repelling sensation is slightly weaker than the ECG pasted electrode, it is still impossible to completely separate the subject from the psychological constraints of wearing the sensor.
  • the Ballistocardiography (BCG) technique can non-invasively measure the effect of blood ejected by the human body on each movement of the heart, and obtain the corresponding BCG waveform signal.
  • BCG signal acquisition technology can non-invasively measure the body's fine vibration signals caused by heartbeat and respiration, so it can monitor the patient's respiratory rate and heart rate in a non-contact manner.
  • BCG reflects cardiac mechanical activity, which can also reflect beat-by-heart rate, per-wave waveform characteristics, and the subject does not need any sensor that touches the body;
  • PPG technology BCG can be more detailed Reflecting the waveform characteristics of each Bo signal, the waveform carries a much richer information than PPG. Under the premise of clarifying the physiological meaning of each wave group, it can have more favorable diagnostic conditions and advantages than PPG, and the subject does not need Any sensor that touches the body.
  • BCG signals acquired by each method will have different waveform time domain morphological details.
  • the BCG obtained by the same acquisition device due to different physiological structures of different subjects The signal will also have different waveform time domain morphological features; and, the same individual different vertical planes (head-to-foot vertical surface, belly-back front and back, left and right sides, etc.), the resulting BCG signal will also have different waveforms.
  • the object of the present invention is to provide a vital sign signal analysis processing method, a computer readable storage medium and a vital sign monitoring device, which aim to solve the prior art BCG technology, when the BCG signal has certain interference and the quality of the collected signal is not high. Under the measurement of heart rate and respiratory rate calculation results have large errors.
  • the present invention provides a method for analyzing vital sign signals, the method comprising:
  • the vital sign time domain signal comprising a BCG time domain signal and/or a respiratory time domain signal
  • the first vital sign parameter includes a first heart rate and/or a first respiration rate; and performing a time-frequency transform of the vital sign time domain signal of the preset duration to obtain a vital sign a frequency domain signal, and calculating a second vital sign parameter based on the vital sign frequency domain signal;
  • the vital sign frequency domain signal includes a BCG frequency domain signal and/or a respiratory frequency domain signal, and the second vital sign parameter includes a second heart rate And/or a second respiratory rate;
  • a final vital sign parameter is calculated based on the first vital sign parameter and the second vital sign parameter, the final vital sign parameter including a final heart rate and/or a final respiration rate.
  • the present invention provides a computer readable storage medium storing a computer program, the computer program being executed by a processor to implement the steps of the vital sign signal analysis processing method as described above.
  • the present invention provides a vital sign monitoring device, comprising:
  • One or more processors are One or more processors;
  • One or more computer programs wherein the one or more computer programs are stored in the memory and configured to be executed by the one or more processors, the processor implementing the computer program The steps of the vital sign signal analysis processing method as described above.
  • the vital sign time domain signal of the preset duration is time-frequency transformed to obtain the vital sign frequency domain signal, and is calculated based on the vital sign frequency domain signal.
  • the second vital sign parameter; the final vital sign parameter is calculated based on the first vital sign parameter and the second vital sign parameter.
  • the two methods are processed in parallel and refer to the auxiliary calculations to enhance the anti-interference ability while satisfying the real-time performance, which can greatly improve the accuracy and reliability of the calculation results.
  • the heart rate and the respiration rate can be measured under the condition that the BCG signal has certain interference and the quality of the collected signal is not high, and the performance is stable and the result is accurate.
  • Embodiment 1 is a flow chart of a method for analyzing vital sign signals according to Embodiment 1 of the present invention.
  • Figure 2 is a schematic diagram of the original signal waveform acquired by the BCG sensor.
  • FIG. 3 is a schematic diagram of a BCG time domain signal waveform.
  • FIG. 4 is a schematic diagram of another BCG time domain signal waveform.
  • FIG. 5 is a schematic diagram of another BCG time domain signal waveform.
  • Figure 6 is a graphical representation of the time-domain search results of the cardiac shock signal when the subject remains stationary and there is almost no disturbance in the physiological frequency bandwidth range when the original signal is acquired.
  • FIG. 7 is a schematic diagram of a frequency domain signal waveform obtained by selecting a time domain signal of FIG. 6 after time-frequency transform.
  • Figure 8 is a schematic diagram showing the results of the time-domain search calculation of the heart-impact signal when there is occasional subtle body shake in the subject, which may not destroy the waveform characteristics of the time domain signal waveform, but when the detail consistency between the waveforms changes. .
  • FIG. 9 is a schematic diagram of a frequency domain signal waveform obtained by selecting a time domain signal of FIG. 8 after time-frequency transform.
  • FIG. 10 is a schematic diagram showing the frequency domain signal waveform obtained by time-frequency transform of the time domain signal waveform of FIG. 5.
  • FIG. 10 is a schematic diagram showing the frequency domain signal waveform obtained by time-frequency transform of the time domain signal waveform of FIG. 5.
  • Figure 11 is a large-scale body motion when the subject occurs in the middle of a certain number of heartbeat cycles, destroying the signal waveform of the time domain signal during the time period, but the entire window data for the time-frequency transform still contains multiple Effective heart beat cycle data, when the energy of the effective signal caused by the heart beat is still at a higher level in the frequency domain signal, the time-domain search wave calculation result of the heart shock signal is schematic.
  • FIG. 12 is a functional block diagram of a vital sign signal analysis and processing apparatus according to Embodiment 2 of the present invention.
  • FIG. 13 is a schematic structural diagram of a vital sign monitoring device according to Embodiment 4 of the present invention.
  • Embodiment 1 is a diagrammatic representation of Embodiment 1:
  • Embodiment 1 is a flowchart of a method for analyzing and analyzing a vital sign signal according to Embodiment 1 of the present invention. It should be noted that, if substantially the same result is obtained, the method of the present invention is not in the sequence of the flow shown in FIG. 1. limit.
  • the method can include, but is not limited to, the following steps:
  • the senor may be an acceleration sensor, a pressure sensor, a displacement sensor, or a sensor that converts physical quantity equivalence based on acceleration, pressure, and displacement (such as an electrostatic charge sensitive sensor, an inflatable micro sensor, One or more of fiber optic sensors, etc.).
  • the sensor When the sensor collects the original signal, the sensor can be generally placed in various ways such as the standing subject, the sitting subject under the hip, and the lying subject. Since the sensor senses the vibration signal of the body, the original signals collected include the respiratory signal component of the subject and the heartbeat signal component, as well as the environmental microvibration, the disturbance caused by the body motion of the subject, and the noise signal of the circuit itself. .
  • the raw signal collected by the sensor can be continuously acquired, and the subject is continuously acquired for a certain period of time.
  • the sensor may also have an interactive interface, and the user may enter information such as age, weight, height, past medical history, and the like of the subject.
  • Figure 2 shows a schematic diagram of the original signal waveform.
  • the large contour of the original signal is the respiratory signal envelope generated by human breathing, and the heart beat and other noise are superimposed on the respiratory signal envelope curve.
  • the vital sign time domain signal includes a BCG time domain signal and/or a respiratory time domain signal.
  • the original signal contains a plurality of sub-signals, such as a subject's respiratory signal, a heart beat signal, an environmental micro-vibration signal, a subject's body motion signal, and a noise signal of the sensor circuit itself
  • filters of different bandwidth ranges are designed, and the original signal is After filtering, the signal of interest can be separated.
  • the signal of interest is a BCG signal
  • the other signals are noise signals, which can be removed by filter filtering.
  • the generating the vital sign time domain signal based on the original signal includes:
  • the original signal of the filter denoising is scaled according to the dynamic range of the signal to obtain a BCG time domain signal
  • the generating the vital sign time domain signal based on the original signal comprises:
  • the filtered denoised original signal is scaled according to the dynamic range of the signal to obtain a respiratory time domain signal.
  • the filter may adopt one or more of an IIR filter, an FIR filter, a wavelet filter, a zero-phase bidirectional filter, etc., and the filter may perform at least one filtering process on the original signal.
  • FIG. 3 is a schematic diagram of a BCG time domain signal waveform generated based on an original signal.
  • the original signal is collected, the subject remains stationary, and the original signal has almost no disturbance in the physiological frequency bandwidth range, and the obtained high-quality BCG time domain signal waveform is obtained.
  • Each waveform has obvious features, regular cycle, clear outline and stable baseline.
  • FIG. 4 is a schematic diagram of a BCG time domain signal waveform generated based on another original signal.
  • the subject had occasional subtle body shake.
  • Each waveform feature is obvious, the cycle is regular, the profile is basically clear, and the baseline is stable. The details of the partial waves change, but do not affect the contour recognition of the waveforms one by one.
  • Figure 5 is a schematic diagram of a BCG time domain signal waveform generated based on another original signal.
  • the subject experienced a short period of substantial body motion in the middle of a certain number of heartbeat cycles.
  • the time domain waveform at this time is relatively disordered, and there is a large irregular fluctuation due to the body motion of the subject.
  • the vital sign time domain signal calculates, according to the vital sign time domain signal, a first vital sign parameter, where the first vital sign parameter includes a first heart rate and/or a first respiration rate; and the time-frequency transform of the vital sign time domain signal of the preset duration is obtained.
  • the vital sign frequency domain signal is calculated, and the second vital sign parameter is calculated based on the vital sign frequency domain signal; the vital sign frequency domain signal includes a BCG frequency domain signal and/or a respiratory frequency domain signal, and the second vital sign parameter includes Two heart rate and / or second breathing rate.
  • S104 Calculate a final vital sign parameter based on the first vital sign parameter and the second vital sign parameter, where the final vital sign parameter includes a final heart rate and/or a final respiratory rate.
  • the BCG time domain signal generated according to the original signal is as shown in FIG.
  • the first vital sign parameter is a first heart rate
  • the vital sign frequency domain signal is a BCG frequency domain signal
  • the second vital sign parameter is a second heart rate
  • the calculating based on the vital sign time domain signal A vital sign parameter, the time-frequency transform of the vital sign time domain signal of the preset duration is obtained to obtain the vital sign frequency domain signal
  • the second vital sign parameter is calculated based on the vital sign frequency domain signal, which may specifically include the following steps:
  • S1011 Search for the characteristic peaks and valleys of each period of the time domain waveform according to the waveform characteristics of the BCG time domain signal, calculate the beat-by-beat heart rate, and calculate the average heart rate as the first heart rate according to the preset duration or the preset number of beats.
  • the average heart rate calculated according to the preset duration or the preset number of beats can be:
  • FIG. 6 is a schematic diagram of the calculation result of the BCG time domain signal search wave of FIG. 3, which has 7 complete waveforms, and the respective beat-by heart beats are 66, 67, 68, 64, 65, 66, 63 ( Unit bpm), the average value can be calculated to be 65.571 bpm, and the average value after removing the maximum and minimum values can be calculated as 65.6 bpm, which is the first heart rate.
  • S1032 Re-sampling the BCG time domain signal of the preset duration, determining the number of time-frequency transform points according to the re-sampling rate, performing time-frequency transform to obtain a corresponding BCG frequency domain signal, and searching the BCG frequency domain signal according to the base time
  • the frequency attribute identifies a reasonable main peak frequency to calculate the second heart rate. Note that S1032 and S1031 are parallel.
  • the time domain calculation is the more accurate the calculation if the number of points is longer in the unit time.
  • the signal sampling rate generally needs to be more accurate at 500 Hz or above, and the time-domain waveform needs to be downsampled and resampled (ie, resampled), for example, 500 Hz is extracted into 100 Hz, 62.5 Hz, 50Hz and so on.
  • the appropriate time-frequency transform points are determined according to the computing resources and capabilities.
  • time domain waveform used for time-frequency transform can contain two or more periodic waveforms. For example, if the minimum value of the heart rate measurement range is declared to be 30 bpm, time domain data containing at least 4 seconds or longer is required. Then, combined with the resampling rate, the number of points of the time-frequency transform can be determined.
  • the time-frequency transform method may employ a Fourier transform, a wavelet transform, or the like. As shown in FIG. 7, the frequency domain signal waveform obtained by time-frequency transform is selected for the time domain signal of FIG. At this time, it is also a high-quality BCG frequency domain waveform.
  • each effective peak has obvious features, the outline is clear and erect, and the fundamental frequency multiplication characteristic is obvious.
  • 66bpm is the main peak, that is, the fundamental frequency peak, and the subsequent distinct peaks are respectively double frequency, triple frequency, and quadruple frequency.
  • the reason why the fundamental frequency peak energy is not the highest at this time is related to the filter characteristics.
  • the low-frequency interference can be filtered or depressed, and the low-frequency interference may be filtered out. Depress the main peak energy. From this, the second heart rate can be calculated to be 66 bpm.
  • the frequency domain signal quality evaluation method can be combined with each peak shape (width, height), combined with interference peaks, and combined with the base frequency group.
  • the quality of the current BCG frequency domain signal can be considered as 100.
  • S104 may specifically be: calculating the output final heart rate according to the quality of the BCG time domain signal and the BCG frequency domain signal, combining the first heart rate and the second heart rate.
  • the time domain signal waveform at this time is a high-quality waveform
  • each waveform has obvious features, periodic regularity, clear contour, and stable baseline
  • the frequency domain signal waveform is also a high-quality waveform
  • each effective peak characteristic is obvious
  • the contour is clear and upright
  • the base frequency is The characteristics are obvious. Therefore, the first heart rate calculated by the time domain signal or the second heart rate calculated by the frequency domain signal has high reliability and accuracy.
  • the first heart rate is rounded up to 66 bpm, which is completely consistent with the second heart rate, that is, the average heart rate calculated by the current window time is 66bpm.
  • the final heart rate can be calculated based on the quality of the BCG time domain signal and the quality of the BCG frequency domain signal:
  • the vital sign frequency domain signal is a respiratory frequency domain signal
  • the second vital sign parameter is a second respiration rate
  • the biometric time domain signal is calculated
  • the first vital sign parameter is obtained, and the vital sign time domain signal of the preset duration is time-frequency transformed to obtain the vital sign frequency domain signal, and the second vital sign parameter is calculated based on the vital sign frequency domain signal, including:
  • the waveform characteristics of the respiratory time domain signal searching for the characteristic peaks and valleys of each period of the time domain waveform, calculating the beat rate per beat, and calculating the average breathing rate as the first breathing rate according to the preset duration or the preset number of beats;
  • Re-sampling the respiratory time domain signal of the preset duration determining the number of time-frequency transform points according to the re-sampling rate, performing time-frequency transform to obtain a corresponding respiratory frequency domain signal, and searching for the respiratory frequency domain signal according to the base frequency doubling attribute Identifying a reasonable main peak frequency to calculate a second respiration rate;
  • the calculating the final vital sign parameter based on the first vital sign parameter and the second vital sign parameter comprises:
  • the output final respiratory rate is calculated in conjunction with the first respiratory rate and the second respiratory rate.
  • the BCG time domain signal generated according to the original signal is as shown in FIG. 4, when the first vital sign parameter is the first Heart rate, the vital sign frequency domain signal is a BCG frequency domain signal, and when the second vital sign parameter is a second heart rate, the first vital sign parameter is calculated based on the vital sign time domain signal, and the life of the preset duration is The time-domain signal of the sign is time-frequency transformed to obtain the vital sign frequency domain signal, and the second vital sign parameter is calculated based on the vital sign frequency domain signal, which may specifically include the following steps:
  • S1033 Search for the characteristic peaks and valleys of each period of the time domain waveform according to the waveform characteristics of the BCG time domain signal, calculate the beat-by-beat heart rate, and calculate the average heart rate as the first heart rate according to the preset duration or the preset number of beats.
  • the average heart rate calculated according to the preset duration or the preset number of beats can be:
  • FIG. 8 is a schematic diagram of the calculation result of the BCG time domain signal search wave of FIG. 4, the window (the duration of the window is not consistent with that in S1031, generally the window duration is fixed for subsequent time-frequency conversion when the actual project is implemented,
  • the main steps of the eight waveforms are 68, 73, 76, 71, 68, 70, 76, 72 (units bpm), and the average value can be calculated to be 71.75 bpm (or the calculation is removed).
  • the average value after the maximum and minimum values is 71.667 bpm), which is the first heart rate.
  • the relationship between the number of theoretical effective waveforms and the actual number of waveforms is explained.
  • S1034 performing time-frequency transform of the BCG time domain signal waveform of the preset duration to obtain a corresponding BCG frequency domain signal waveform, searching for a main peak near the first heart rate in the BCG frequency domain signal waveform, and verifying according to the base frequency characteristic, The frequency corresponding to the verified main peak is taken as the second heart rate.
  • the frequency domain signal waveform calculated by time-frequency transform of the time domain signal of Fig. 8 is as shown in Fig. 9.
  • the main peak is annihilated in the interference peak, and there are interference peaks with large energy on both the left and right sides.
  • the first heart rate is calculated in combination with the time domain, and the main peak of the auxiliary search is near 71.75 bpm, and there is exactly the main peak with the third energy ranking.
  • the second frequency, the triple frequency, and the quadruple frequency which have obvious characteristics and clear outlines, can be searched, so that the second heart rate can be calculated to be 72 bpm.
  • the maximum effective multiplication of the fundamental frequency peak of 1.2 Hz is near the frequency of 4.8 Hz.
  • the quality of the signal is 100. If you need to consider the signal characteristics more deeply, you can find that the maximum and second energy peaks are pseudo peaks (interference peaks), and the signal quality can be corrected.
  • the signal quality is reduced by X (subtraction correction), or the signal quality is multiplied by Y% (multiplication correction), and both X and Y are reasonable values defined by empirical coefficients.
  • X subtraction correction
  • Y% multiplication correction
  • S104 may specifically be: calculating the output final heart rate according to the quality of the BCG time domain signal and the BCG frequency domain signal, combining the first heart rate and the second heart rate.
  • the time domain signal waveform is close to the high quality waveform, and each waveform has obvious characteristics, periodicity, clear outline, and stable baseline. Although there are changes in the detailed features of the partial waves, it does not affect the contour recognition of the waveforms one by one. Although there are interference peaks in the frequency domain signal waveform, the main peak based on the first heart rate reference is still clear, and the characteristics of each frequency doubling peak are obvious, the contour is clear and erect, and the fundamental frequency doubling characteristic is obvious. Therefore, the first heart rate calculated by the time domain signal has high reliability and accuracy, and the second heart rate calculated based on the first heart rate reference in the frequency domain signal also has high reliability and accuracy.
  • the first heart rate is 72bpm after rounding, which is completely consistent with the second heart rate, that is, the average heart rate calculated by the current window time is 72bpm.
  • the vital sign frequency domain signal is a respiratory frequency domain signal
  • the second vital sign parameter is a second respiration rate
  • the biometric time domain signal is calculated
  • the first vital sign parameter is obtained, and the vital sign time domain signal of the preset duration is time-frequency transformed to obtain the vital sign frequency domain signal, and the second vital sign parameter is calculated based on the vital sign frequency domain signal, including:
  • the waveform characteristics of the respiratory time domain signal searching for the characteristic peaks and valleys of the time domain waveform, calculating the beat rate per beat, and calculating the average breathing rate as the first breathing rate according to the preset duration or the preset number of beats;
  • Performing a time-frequency transform on the respiratory time domain signal to obtain a corresponding respiratory frequency domain signal searching for a main peak near the first respiratory rate in the respiratory frequency domain signal, and verifying according to the fundamental frequency characteristic, the frequency corresponding to the verified main peak is taken as Second respiratory rate;
  • the calculating the final vital sign parameter based on the first vital sign parameter and the second vital sign parameter comprises:
  • the output final respiratory rate is calculated in conjunction with the first respiratory rate and the second respiratory rate.
  • the subject when the original signal is collected, the subject experiences a short-term large body motion in the middle of a certain number of heartbeat cycles, which destroys the signal waveform of the time domain signal during the time period, and is generated according to the original signal.
  • the BCG time domain signal is as shown in FIG. 5
  • the first vital sign parameter is a first heart rate
  • the vital sign frequency domain signal is a BCG frequency domain signal
  • the second vital sign parameter is a second heart rate
  • the first vital sign parameter is calculated based on the vital sign time domain signal
  • the vital sign time domain signal of the preset duration is time-frequency transformed to obtain the vital sign frequency domain signal
  • the second vital sign is calculated based on the vital sign frequency domain signal.
  • the parameters may specifically include the following steps:
  • S1035 re-sampling the BCG time domain signal waveform of the preset duration, determining the number of time-frequency transform points according to the re-sampling rate, performing time-frequency transform to obtain a corresponding BCG frequency domain signal waveform, and searching the BCG frequency domain signal waveform.
  • the second heart rate is calculated by identifying a reasonable main peak frequency based on the base frequency characteristic.
  • the frequency domain signal waveform obtained by time-frequency transform of the time domain signal waveform of FIG. 5 is as shown in FIG.
  • the first energy peak is 231 bpm
  • the second energy peak is 152 bpm
  • the third and fourth energy peaks are all suspected main peaks.
  • 77bpm is a reasonable main peak. Its double frequency and triple frequency characteristics are obvious, and the outline is clear and erect, so the second heart rate can be calculated to be 77 bpm.
  • the fundamental frequency multiplication since we only care about the frequency within 5 Hz, the maximum effective frequency multiplication of the fundamental frequency peak of 1.28 Hz is near the triple frequency of 3.84 Hz.
  • S1036 Calculate a single cycle average time width according to the second heart rate, set a reasonable upper and lower width threshold line with a single cycle average time width, and search for the effective wave and the approximate effective wave in the window with the waveform matching or the characteristic peak-to-valley threshold to obtain the first The range of heart rates.
  • the number of waveforms that can be recognized in the window by the search method according to the waveform matching or the reasonable formulation of the characteristic peak-to-valley threshold is limited.
  • the single-cycle average time width calculated by the second heart rate 77bpm is 779.22ms, and the reasonable upper and lower width threshold lines are set according to the width, and the waveform matching or the characteristic peak-valley threshold can be reasonably formulated to search for effective in the window.
  • Waves and approximate effective waves may be a heartbeat cycle or a pseudowave). As shown in FIG.
  • the beat-by-heart rate of the valid wave (or approximate effective wave) recognized at this time is 69, 79, 81, 77, 72 (unit bpm), and the three waves of 69, 79, and 81 are approximately effective.
  • Wave there is a certain shape variation.
  • the beat-by-beat heart rate of the intermediate wave It can only be estimated that the average heart rate may fluctuate within the range of 69-81 bpm (may also exceed this range). Therefore, it is difficult to determine the exact value of the first heart rate HR1.
  • the window duration is about 10 seconds.
  • S104 may specifically be: calculating the output final heart rate according to the quality of the BCG time domain signal and the BCG frequency domain signal, combined with the range of the first heart rate and the second heart rate.
  • the second heart rate has higher reliability and accuracy.
  • the time domain signal waveform is disordered, it is difficult to search and determine all the beat-by-shot waveforms in the window, and the first heart rate calculated based on the second heart rate in the time domain signal can only determine the approximate heart rate range and cannot determine the accurate value.
  • the second heart rate 77bpm is highly reliable on the one hand, and falls within the maximum possible range of the first heart rate from 69 to 81 bpm on the other hand. It can be considered that the average heart rate calculated by the current window time is 77 bpm.
  • the final heart rate can be calculated based on the quality of the BCG time domain signal and the quality of the BCG frequency domain signal:
  • the vital sign frequency domain signal is a respiratory frequency domain signal
  • the second vital sign parameter is a second respiration rate
  • the biometric time domain signal is calculated
  • the first vital sign parameter is obtained, and the vital sign time domain signal of the preset duration is time-frequency transformed to obtain the vital sign frequency domain signal, and the second vital sign parameter is calculated based on the vital sign frequency domain signal, including:
  • Re-sampling the respiratory time domain signal determining the number of time-frequency transform points according to the re-sampling rate, performing time-frequency transform to obtain the corresponding respiratory frequency domain signal, searching for the respiratory frequency domain signal, and identifying a reasonable main peak frequency according to the base frequency characteristic To calculate the second respiratory rate;
  • the calculating the final vital sign parameter based on the first vital sign parameter and the second vital sign parameter comprises:
  • the output final respiratory rate is calculated in conjunction with the range of the first respiratory rate and the second respiratory rate.
  • the calculating the final respiratory rate according to the quality of the respiratory time domain signal and the respiratory frequency domain signal, in combination with the first respiratory rate range and the second respiratory rate, is specifically:
  • Minimum respiratory rate (first respiratory rate minimum * respiratory time domain signal quality + second respiratory rate * respiratory frequency domain signal quality) / (breathing time domain signal quality + respiratory frequency domain signal quality);
  • the highest final respiratory rate (first respiratory rate maximum * respiratory time domain signal quality + second respiratory rate * respiratory frequency domain signal quality) / (breathing time domain signal quality + respiratory frequency domain signal quality);
  • the final respiration rate is equal to the average of the lowest final respiratory rate and the highest final respiratory rate.
  • the time domain calculation and the frequency domain calculation can be mutually referenced as mutual assistance, and the time domain calculation and the frequency domain calculation can be based on the same time window data.
  • the time domain calculation result of the current window data may not be limited to the frequency domain calculation reference for the current window data, and may also be used for the time domain calculation reference and the frequency domain calculation reference of the next window data.
  • the frequency domain calculation result of the current window data may not be limited to the time domain calculation reference for the current window data, and may also be used for the frequency domain calculation reference and the time domain calculation reference of the next window data.
  • Embodiment 2 is a diagrammatic representation of Embodiment 1:
  • the vital sign signal analysis and processing apparatus provided in the second embodiment of the present invention includes:
  • the obtaining module 11 is configured to acquire an original signal collected by the sensor
  • the vital sign time domain signal generating module 12 is configured to generate a vital sign time domain signal based on the original signal, where the vital sign time domain signal comprises a BCG time domain signal and/or a respiratory time domain signal;
  • the first/two vital sign parameter calculation module 13 is configured to calculate a first vital sign parameter based on the vital sign time domain signal, where the first vital sign parameter includes a first heart rate and/or a first respiration rate;
  • the vital sign time domain signal is time-frequency transformed to obtain a vital sign frequency domain signal, and the second vital sign parameter is calculated based on the vital sign frequency domain signal;
  • the vital sign frequency domain signal includes a BCG frequency domain signal and/or a respiratory frequency domain a signal, the second vital sign parameter comprising a second heart rate and/or a second breathing rate;
  • the final vital sign parameter calculation module 14 is configured to calculate a final vital sign parameter based on the first vital sign parameter and the second vital sign parameter, the final vital sign parameter including a final heart rate and/or a final respiratory rate.
  • the vital sign signal analysis and processing device provided by the second embodiment of the present invention and the vital sign signal analysis and processing method provided by the first embodiment of the present invention belong to the same concept, and the specific implementation process thereof is detailed in the specification, and details are not described herein again.
  • Embodiment 3 is a diagrammatic representation of Embodiment 3
  • the third embodiment of the present invention further provides a computer readable storage medium, where the computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement vital sign analysis according to the first embodiment of the present invention.
  • the steps of the processing method are not limited to a computer readable storage medium, where the computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement vital sign analysis according to the first embodiment of the present invention. The steps of the processing method.
  • Embodiment 4 is a diagrammatic representation of Embodiment 4:
  • FIG. 13 is a block diagram showing a specific structure of a vital sign monitoring device according to Embodiment 4 of the present invention.
  • a vital sign monitoring device 100 includes:
  • One or more processors 101 are One or more processors 101;
  • One or more computer programs wherein the one or more computer programs are stored in the memory 102 and configured to be executed by the one or more processors 101, the processor 101 executing the computer
  • the steps of the vital sign signal analysis processing method provided in the first embodiment of the present invention are implemented in the program.
  • the vital sign time domain signal of the preset duration is time-frequency transformed to obtain the vital sign frequency domain signal, and is calculated based on the vital sign frequency domain signal.
  • the second vital sign parameter; the final vital sign parameter is calculated based on the first vital sign parameter and the second vital sign parameter.
  • the two methods are processed in parallel and refer to the auxiliary calculations to enhance the anti-interference ability while satisfying the real-time performance, which can greatly improve the accuracy and reliability of the calculation results.
  • the heart rate and the respiration rate can be measured under the condition that the BCG signal has certain interference and the quality of the collected signal is not high, and the performance is stable and the result is accurate.

Abstract

一种生命体征信号分析处理方法、装置和生命体征监测设备（100）。该方法包括：获取由传感器采集的原始信号（S101）；基于原始信号生成生命体征时域信号（S102）；基于生命体征时域信号计算得到第一生命体征参数；将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数（S103）；基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数，该最终生命体征参数包括最终心率和/或最终呼吸率（S104）。该方法在满足实时性的同时增强抗干扰能力，可大大提升计算结果的准确性和可靠性。

Description

一种生命体征信号分析处理方法和生命体征监测设备 技术领域
本发明属于信号处理领域，尤其涉及一种生命体征信号分析处理方法和生命体征监测设备。
 背景技术
人体呼吸率、心率等生命体征参数监测对睡眠质量监测、呼吸疾病预防及诊断、心血管疾病预防及诊断等具有非常重要意义。
心电图(Electrocardiogram，ECG)是一种比较有效的监测方式，其能够精确捕捉到人体心脏每一次搏动的电活动信号，进而计算逐拍心率、每博波形特征，基于此还能分析心率变异性 (Heart Rate Variability，HRV)、波群幅值或宽度的特征信息、R波幅度特征提取得到的呼吸率等参数结果。但心电图监测需要在受试者身上粘贴电极片，对受试者日常活动产生一定约束，对身体额外粘贴上传感器的心理感受可能会有一定疑虑或排斥甚至产生心理负担，尤其当使用多导联需粘贴多个电极片的情形下会更加强烈。
基于光电容积描记波信号 (Photoplethysmograph，PPG)也能监测呼吸率、心率（此时为脉率）生命体征参数。但PPG信号采集容易受到多种因素干扰，如传感器佩戴位置、佩戴者的皮肤、佩戴部位容易发生的运动造成的伪影等均可能造成信号干扰，导致每搏脉率检测不准确；同时基于PPG计算的呼吸率需要较长时间稳定的优质信号，需要从优质PPG信号中提取低频的呼吸轮廓来计算，非常容易受到细微的低频干扰而影响计算；且PPG仍需要佩戴传感器（指夹式、耳夹式、鼻夹式等），尽管相比ECG粘贴电极片的排斥感会稍微薄弱，但仍然不能将受试者完全从佩戴传感器的心理束缚中脱离出来。
心冲击图 (Ballistocardiography，BCG)技术，可以非侵入式地测量人体由于心脏每次搏动所喷射的血液对于人体运动的作用，获得相应的BCG波形信号。BCG信号采集技术可以非侵入式地测量人体由于心跳和呼吸引起的身体细微振动信号，因此能够非接触式的对患者进行呼吸率、心率监测。相较于ECG技术，BCG反映的是心脏机械活动，同样能够反映逐拍心率、每博波形特征，且受试者不需要任何接触到身体的传感器；相较于PPG技术，BCG能够更加精细地反映每博信号波形特征，其波形携带的信息量比PPG要丰富很多，在明确波形各个波群代表的生理含义的前提下能够比PPG具备更有利的诊断条件和优势，且受试者不需要任何接触到身体的传感器。
随着现代传感器技术的高速发展，越来越多优秀的传感器被用于检测获取优质BCG信号，能够更加优秀地反映心脏每博运动的真实信号，使得BCG技术的研究与发展又重获诸多研究机构与企业的青睐。然而由于各种BCG信号采集方法所使用的传感器迥异，各方法采集得到的BCG信号会具备不同的波形时域形态细节特征；而且，不同受试者由于生理结构差异，同样的采集设备得到的BCG信号也会具备不同的波形时域形态特征；且，同样的个体不同垂直面（头-脚竖直面、腹-背前后面、左右侧面等），得到的BCG信号也会具备不同的波形时域形态特征；甚至，对应同一个个体同样的垂直面，在不同时间段由于生理状况不一样也会有不同的时域形态特征。因此，现有技术的BCG技术，当BCG信号存在一定干扰和采集信号质量不高的情况下测量心率和呼吸率的计算结果误差大。这对于基于BCG技术的长期、实时呼吸率、心率监测提出了重大挑战。
 技术问题
本发明的目的在于提供一种生命体征信号分析处理方法、计算机可读存储介质和生命体征监测设备，旨在解决现有技术的BCG技术，当BCG信号存在一定干扰和采集信号质量不高的情况下测量心率和呼吸率的计算结果误差大的问题。
 技术解决方案
第一方面，本发明提供了一种生命体征信号分析处理方法，所述方法包括：
获取由传感器采集的原始信号；
基于原始信号生成生命体征时域信号，所述生命体征时域信号包括BCG时域信号和/或呼吸时域信号；
基于生命体征时域信号计算得到第一生命体征参数，所述第一生命体征参数包括第一心率和/或第一呼吸率；将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数；所述生命体征频域信号包括BCG频域信号和/或呼吸频域信号，所述第二生命体征参数包括第二心率和/或第二呼吸率；
基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数，所述最终生命体征参数包括最终心率和/或最终呼吸率。
第二方面，本发明提供了一种计算机可读存储介质，所述计算机可读存储介质存储有计算机程序，所述计算机程序被处理器执行时实现如上述的生命体征信号分析处理方法的步骤。
第三方面，本发明提供了一种生命体征监测设备，包括：
一个或多个处理器；
存储器；以及
一个或多个计算机程序，其中所述一个或多个计算机程序被存储在所述存储器中，并且被配置成由所述一个或多个处理器执行，所述处理器执行所述计算机程序时实现如上述的生命体征信号分析处理方法的步骤。
 有益效果
在本发明中，由于基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数；基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数。两种方法并行处理，相互参考辅助计算，在满足实时性的同时增强抗干扰能力，可大大提升计算结果的准确性和可靠性。能够在BCG信号存在一定干扰和采集信号质量不高的情况下测量心率和呼吸率，性能稳定，结果精确。
 附图说明
图1是本发明实施例一提供的生命体征信号分析处理方法流程图。
图2是由BCG传感器采集的原始信号波形示意图。
图3是一种BCG时域信号波形示意图。
图4是另一种BCG时域信号波形示意图。
图5是另一种BCG时域信号波形示意图。
图6是在采集原始信号时受试者保持静止，几乎无生理频率带宽范围内扰动时，心冲击信号时域搜波计算结果示意图。
图7是选取图6时域信号进行时频变换后计得到的频域信号波形示意图。
图8是当受试者存在偶发的细微身体抖动，可能未破坏时域信号波形的逐个波形特性，但使得各个波形之间的细节一致性发生变化时，心冲击信号时域搜波计算结果示意图。
图9是选取图8时域信号进行时频变换后计得到的频域信号波形示意图。
图10是选取图5时域信号波形进行时频变换后得到的频域信号波形示意图。
图11是当受试者在某几个心跳周期中间发生短时的大幅体动，破坏了时域信号在该时间段内的信号波形，但是用于时频变换的整个窗口数据仍然包含多个有效的心脏搏动周期数据，在频域信号上心脏搏动引起的有效信号的能量仍然处于较优质水平时，心冲击信号时域搜波计算结果示意图。
图12是本发明实施例二提供的生命体征信号分析处理装置的功能模块框图。
图13是本发明实施例四提供的生命体征监测设备结构示意图。
 本发明的最佳实施方式
为了使本发明的目的、技术方案及有益效果更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。
为了说明本发明所述的技术方案，下面通过具体实施例来进行说明。
实施例一：
请参阅图1，是本发明实施例一提供的生命体征信号分析处理方法流程图，需注意的是，若有实质上相同的结果，本发明的方法并不以图1所示的流程顺序为限。所述方法可以包括但不限于以下步骤：
S101、获取由传感器采集的原始信号。
在本发明实施例一中，传感器可以是加速度传感器、压力传感器、位移传感器、或者以加速度、压力、位移为基础将物理量等效性转换的传感器（如静电荷敏感传感器、充气式微动传感器、光纤传感器等）中的一种或多种。
传感器采集原始信号时，一般可将传感器放置于站姿受试者脚下、坐姿受试者臀下、躺姿受试者背下等多种部位多种方式进行测量。由于传感器感应的是身体的震动信号，因此所采集得到的原始信号均包含受试者呼吸信号成分和心脏搏动信号成分，以及环境微震动、受试者体动引起的干扰和电路自身的噪声信号。
传感器采集原始信号可以是连续采集的，在一定的时间段内对受试者进行连续的采集信号。在一些实施例中，传感器还可以具有交互界面，用户可以输入信息，例如，受试者的年龄、体重、身高、既往病史等相关信息。
图2所示为一种原始信号波形示意图，此时的原始信号大轮廓即为人体呼吸产生的呼吸信号包络，心脏搏动与其它噪声则叠加在呼吸信号包络曲线上。
S102、基于原始信号生成生命体征时域信号，所述生命体征时域信号包括BCG时域信号和/或呼吸时域信号。
由于原始信号中包含多种子信号，如受试者呼吸信号、心脏搏动信号、环境微震动信号、受试者体动信号和传感器电路自身的噪声信号，设计不同带宽范围的滤波器，对原始信号进行滤波后即可分离出感兴趣的信号，例如感兴趣的信号是BCG信号，则其他信号则为噪声信号，可用滤波器滤波去除。
因此，在本发明实施例一中，当所述生命体征时域信号是BCG时域信号时，所述基于原始信号生成生命体征时域信号包括：
通过心率计算所需生理频率带宽范围的滤波器对采集的原始信号进行滤波去噪；
根据信号动态范围对滤波去噪后的原始信号进行缩放得到BCG时域信号；
当所述生命体征时域信号是呼吸时域信号时，所述基于原始信号生成生命体征时域信号包括：
通过呼吸率计算所需生理频率带宽范围的滤波器对采集的原始信号进行滤波去噪；
根据信号动态范围对滤波去噪后的原始信号进行缩放得到呼吸时域信号。
其中，滤波器可采用IIR滤波器、FIR滤波器、小波滤波器、零相位双向滤波器等中的一种或多种，滤波器可以对原始信号进行至少一次滤波处理。
如图3所示为基于一种原始信号生成的BCG时域信号波形示意图。采集此种原始信号时受试者保持静止，原始信号几乎没有生理频率带宽范围内扰动，得到的优质的BCG时域信号波形示意图。每个波形特征明显，周期规律，轮廓清晰，基线平稳。
图4是基于另一种原始信号生成的BCG时域信号波形示意图。采集此种原始信号时，受试者存在偶发的细微身体抖动。每个波形特征明显，周期规律，轮廓基本清晰，基线平稳。部分波的细节特征发生变化，但是并不影响逐个波形的轮廓识别。
图5是基于另一种原始信号生成的BCG时域信号波形示意图。采集此种原始信号时，受试者在某几个心跳周期中间发生短时的大幅体动。此时的时域波形比较紊乱，存在由于受试者体动引起的极不规律的大幅波动。
S103、基于生命体征时域信号计算得到第一生命体征参数，所述第一生命体征参数包括第一心率和/或第一呼吸率；将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数；所述生命体征频域信号包括BCG频域信号和/或呼吸频域信号，所述第二生命体征参数包括第二心率和/或第二呼吸率。
S104、基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数，所述最终生命体征参数包括最终心率和/或最终呼吸率。
在本发明实施例一中，当在采集原始信号时受试者保持静止，几乎无生理频率带宽范围内扰动时，依据原始信号生成的BCG时域信号如图3所示时。当所述第一生命体征参数是第一心率，所述生命体征频域信号是BCG频域信号，所述第二生命体征参数是第二心率时，所述基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数具体可以包括以下步骤：
S1031、根据BCG时域信号波形特征，搜索识别时域波形的各周期特征峰谷，计算逐拍心率，并根据预设时长或者预设节拍数计算得到平均心率作为第一心率。
根据预设时长或者预设节拍数计算得到平均心率具体可以是：
计算所有预设时长或者所有预设节拍数的平均心率，或者，计算去除预设时长或者预设节拍数中的最大值和最小值以后的平均心率。
如图6所示是对图3的BCG时域信号搜波计算结果示意图，该窗口内有7个完整波形，各自的逐拍心率依次为66，67，68，64，65，66，63（单位bpm），可以计算得到均值为65.571bpm，也可以计算去除最大最小值后的平均值为65.6bpm，此即为第一心率。时域信号质量的评估方法有多种，可以根据窗口内理论有效波形数跟实际波形数的关系，也可根据逐拍波形的相关性或者匹配程度，或者每个波形特征峰谷（例如J峰）与其它峰阈值合理性程度。以理论有效波形数跟实际波形数的关系来说明，此时窗口时长为6.5秒左右，理论上能够包含的有效波数为7个，此时实际恰好能够有7个波，实际波数/理论波数*100%=100%，去掉“%”单位，可以认为当前BCG时域信号的质量为100。
S1032、将预设时长的BCG时域信号进行重采样，根据重采样率确定时频变换的点数，进行时频变换得到相应的BCG频域信号，对BCG频域信号进行搜波，根据基倍频属性识别合理主峰频率来计算得到第二心率。需注意S1032与S1031是并行的。
由于频域计算运算量大，需要一定时长数据，因此在单位时长内如果点数越多运算量也将越大。而时域计算则是单位时长内如果点数越多，计算越精确，一般地可以理解为原始信号采样率越高越好。针对心率计算的场景，信号采样率一般需要在500Hz或以上可得到较为准确的逐拍心率，需要对时域波形进行降采样重新抽样（即重采样），比如将500Hz抽成100Hz、62.5Hz、50Hz等。确定重采样率之后，根据运算资源和能力确定合适的时频变换点数，一般来说点数越多越精确，但点数越多需要的原始数据长度也越长。合理的设计最好能够保证用于时频变换的时域波形能够包含两个或者以上的周期波形，例如假设心率测量范围的最小值宣称为30bpm，则需要至少包含4秒以上时长的时域数据，再结合重采样率可确定时频变换的点数。时频变换方法可采用傅立叶变换、小波变换等。如图7所示，为选取图6时域信号进行时频变换后计得到的频域信号波形。此时也为优质的BCG频域波形，每个有效峰特征明显，轮廓清晰直立，基倍频特性明显。此时66bpm为主峰即基频峰，后续的各个明显峰分别为其二倍频、三倍频、四倍频。注意此时基频峰能量不是最高的原因与滤波器特性有关，在由时域信号进行时频变换的同时欲抑制低频干扰可以对低频信号进行滤除或压低，在滤除低频干扰的同时可能压低主峰能量。由此可计算得到第二心率为66bpm。频域信号质量的评估方法可结合每个峰自身形状（宽度，高度），也可结合干扰峰，还可以结合基倍频组。例如以基倍频来考虑，由于我们只关心5Hz以内频率，此时基频峰1.1Hz的最大有效倍频为四倍频4.4Hz附近。此时我们恰好能够找到其二、三、四倍频，因此实际倍频数/理论倍频数*100%=100%，去掉“%”单位，可以认为当前BCG频域信号的质量为100。
此时，S104具体可以为：根据BCG时域信号和BCG频域信号的质量，结合第一心率和第二心率，计算输出最终心率。
由于此时的时域信号波形为优质波形，每个波形特征明显，周期规律，轮廓清晰，基线平稳；频域信号波形也为优质波形，每个有效峰特征明显，轮廓清晰直立，基倍频特性明显。因此无论时域信号计算得到的第一心率，还是频域信号计算得到的第二心率均具有高可靠性与准确性。实际上对第一心率进行四舍五入之后为66bpm，与第二心率是完全一致的，即当前窗口时间计算得到的平均心率最终输出为66bpm。实际上，可根据BCG时域信号的质量和BCG频域信号的质量计算最终心率：
最终心率= (第一心率*BCG时域信号的质量 + 第二心率*BCG频域信号的质量)/( BCG时域信号的质量+ BCG频域信号的质量) = (66*100 + 66*100)/(100 + 100) = 66(bpm)。对于最终呼吸率参数，采用如下公式：最终呼吸率= (第一呼吸率*呼吸时域信号的质量+第二呼吸率*呼吸频域信号的质量)/(呼吸时域信号的质量+呼吸频域信号的质量)。
当所述第一生命体征参数是第一呼吸率，所述生命体征频域信号是呼吸频域信号，所述第二生命体征参数是第二呼吸率时，所述基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数包括：
根据呼吸时域信号波形特征，搜索识别时域波形的各周期特征峰谷，计算逐拍呼吸率，并根据预设时长或者预设节拍数计算得到平均呼吸率作为第一呼吸率；
将预设时长的呼吸时域信号进行重采样，根据重采样率确定时频变换的点数，进行时频变换得到相应的呼吸频域信号，对呼吸频域信号进行搜波，根据基倍频属性识别合理主峰频率来计算得到第二呼吸率；
当所述最终生命体征参数是最终呼吸率时，所述基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数包括：
根据呼吸时域信号和呼吸频域信号的质量，结合第一呼吸率和第二呼吸率，计算输出最终呼吸率。
在本发明实施例一中，当采集原始信号时受试者存在偶发的细微身体抖动，依据原始信号生成的BCG时域信号如图4所示时，当所述第一生命体征参数是第一心率，所述生命体征频域信号是BCG频域信号，所述第二生命体征参数是第二心率时，所述基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数具体可以包括以下步骤：
S1033、根据BCG时域信号波形特征，搜索识别时域波形的各周期特征峰谷，计算逐拍心率，并根据预设时长或者预设节拍数计算得到平均心率作为第一心率。
根据预设时长或者预设节拍数计算得到平均心率具体可以是：
计算所有预设时长或者所有预设节拍数的平均心率，或者，计算去除预设时长或者预设节拍数中的最大值和最小值以后的平均心率。
如图8所示是对图4的BCG时域信号搜波计算结果示意图，该窗口（该窗口时长与S1031中并非一致，一般地在实际工程实现时会固定窗口时长来做后续时频变换，这里主要为阐述实施步骤因此并未限定）内8个波形的逐拍心率为68，73，76，71，68，70，76，72（单位bpm），可以计算均值为71.75bpm（或者计算去除最大最小值后的平均值为71.667bpm），此即为第一心率。同样地以理论有效波形数跟实际波形数的关系来说明，此时窗口时长为7秒左右，理论上能够包含的有效波数为8个，此时实际恰好能够有8个波，实际波数/理论波数*100%=100%，去掉“%”单位，可以认为当前BCG时域信号的质量为100。
S1034、将预设时长的BCG时域信号波形进行时频变换得到相应的BCG频域信号波形，搜索BCG频域信号波形中在第一心率附近的主峰，并根据基倍频属性验证后，将通过验证的主峰对应的频率作为第二心率。
选取图8时域信号进行时频变换后计算得到的频域信号波形如图9所示。此时主峰被湮没在干扰峰当中，其左右两侧均存在能量较大的干扰峰。此时结合时域计算结果第一心率，辅助搜索主峰为71.75bpm附近，恰好存在能量排序为第三的主峰。进一步计算可靠性的同时，可以搜索到其存在特征明显，轮廓清晰直立的二倍频、三倍频、四倍频，因此可以计算得到第二心率为72bpm。同样地以基倍频来考虑，由于我们只关心5Hz以内频率，此时基频峰1.2Hz的最大有效倍频为四倍频4.8Hz附近。此时我们恰好能够找到其二、三、四倍频，因此实际倍频数/理论倍频数*100%=100%，去掉“%”单位，可以认为仅以基倍频考虑得到的当前BCG频域信号的质量为100。而如果需要更加深入考虑信号特性，可发现此时最大与第二大能量峰均为伪峰（干扰峰），此时可以对信号质量做一定修正。比如存在一个非有效主峰的较大能量峰则信号质量降低X（减法修正），或者说信号质量乘以Y%（乘法修正），X与Y均为根据经验系数定义的合理值。不妨这里取乘法修正为例，Y为95%，则此时的BCG频域信号的质量为100*95%*95%=90。
此时，S104具体可以为：根据BCG时域信号和BCG频域信号的质量，结合第一心率和第二心率，计算输出最终心率。
此时的时域信号波形接近优质波形，每个波形特征明显，周期规律，轮廓基本清晰，基线平稳。尽管存在部分波的细节特征发生变化，但是并不影响逐个波形的轮廓识别。频域信号波形尽管存在干扰峰伪峰，但基于第一心率参考的主峰仍然较为清晰，并且各个倍频峰特征明显，轮廓清晰直立，基倍频特性明显。因此时域信号计算得到的第一心率具有高可靠性与准确性，且基于第一心率参考在频域信号计算得到的第二心率也具有高可靠性与准确性。实际上第一心率进行四舍五入之后为72bpm，与第二心率是完全一致的，即当前窗口时间计算得到的平均心率最终输出为72bpm。实际上，可根据两者信号质量计算最终心率：最终心率= (第一心率*BCG时域信号的质量+第二心率*BCG频域信号的质量)/( BCG时域信号的质量+BCG频域信号的质量) = (72*100 + 72*90)/(100 + 90) = 72 (bpm)。对于最终呼吸率参数，采用如下公式：最终呼吸率= (第一呼吸率*呼吸时域信号的质量+第二呼吸率*呼吸频域信号的质量)/(呼吸时域信号的质量+呼吸频域信号的质量)。
当所述第一生命体征参数是第一呼吸率，所述生命体征频域信号是呼吸频域信号，所述第二生命体征参数是第二呼吸率时，所述基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数包括：
根据呼吸时域信号波形特征，搜索识别时域波形的特征峰谷，计算逐拍呼吸率，并根据预设时长或者预设节拍数计算得到平均呼吸率作为第一呼吸率；
对呼吸时域信号进行时频变换得到相应的呼吸频域信号，搜索呼吸频域信号中在第一呼吸率附近的主峰，并根据基倍频属性验证后，将通过验证的主峰对应的频率作为第二呼吸率；
当所述最终生命体征参数是最终呼吸率时，所述基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数包括：
根据呼吸时域信号和呼吸频域信号的质量，结合第一呼吸率和第二呼吸率，计算输出最终呼吸率。
在本发明实施例一中，当采集原始信号时受试者在某几个心跳周期中间发生短时的大幅体动，会破坏时域信号在该时间段内的信号波形，依据原始信号生成的BCG时域信号如图5所示时，当所述第一生命体征参数是第一心率，所述生命体征频域信号是BCG频域信号，所述第二生命体征参数是第二心率时，所述基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数具体可以包括以下步骤：
S1035、将预设时长的BCG时域信号波形进行重采样，根据重采样率确定时频变换的点数，进行时频变换得到相应的BCG频域信号波形，对BCG频域信号波形进行搜波，根据基倍频属性识别合理主峰频率来计算得到第二心率。
选取图5时域信号波形进行时频变换后得到的频域信号波形如图10所示。此时第一能量峰为231bpm，第二能量峰为152bpm，剩下第三、第四能量峰均为疑似主峰。而显然根据基倍频属性，77bpm为合理主峰。其二倍频、三倍频特征明显，轮廓清晰直立，因此可以计算得到第二心率为77bpm。同样地以基倍频来考虑，由于我们只关心5Hz以内频率，此时基频峰1.28Hz的最大有效倍频为三倍频3.84Hz附近。此时我们恰好能够找到其二、三倍频，因此实际倍频数/理论倍频数*100%=100%，去掉“%”单位，可以认为仅以基倍频考虑得到的当前BCG频域信号质量为100。而如果需要更加深入考虑信号特性，可发现此时第三大能量峰均为伪峰（干扰峰），高于基频峰能量，此时可以对信号质量做一定修正。同样地不妨这里取乘法修正为例，Y为95%，则此时的BCG频域信号质量为100*95%=95。
S1036、根据第二心率计算得到单周期平均时间宽度，以单周期平均时间宽度设置合理上下宽度阈值线，配合波形匹配或者特征峰谷阈值搜索出窗口内的有效波和近似有效波，得到第一心率的范围。
此时由于信号干扰较大，根据波形匹配或者特征峰谷阈值的合理制定等搜索方法在窗口内可以识别到的波形数量有限。结合频域计算结果，第二心率77bpm计算得到的单周期平均时间宽度为779.22ms，以此宽度设置合理上下宽度阈值线，配合波形匹配或者特征峰谷阈值的合理制定可以搜索出窗口内的有效波和近似有效波（可能是一次心跳周期也可能是伪波）。如图11所示，此时识别出的有效波（或者近似有效波）的逐拍心率为69，79，81，77，72（单位bpm），且69，79和81三个波为近似有效波，存在一定的形状变异。但是由于中间存在较多干扰导致无法知道中间波的逐拍心率，只能估计平均心率可能位于69~81bpm附近范围波动（也可能超出该范围）。因此较难决定第一心率HR1的准确值。同样地以理论有效波形数跟实际波形数的关系来说明，此时窗口时长为10秒左右，理论上能够包含的有效波数为12个，此时仅有5个波，实际波数/理论波数*100%=42%，去掉“%”单位，可以认为当前信号质量为42。但实际上有三个波为近似波，以此为标准的实际有效波只有2个，实际波数/理论波数*100%=17%，去掉“%”单位，可以认为当前信号质量为17。进一步地，还可以考虑干扰时长占比，如此时干扰时长约占窗口时长50%，则在前述信号质量上再乘以50%，或者其他调整方式以降低时域结果对最终结果的影响。不妨以2个实际有效峰为结果的信号质量来进行计算，并乘以50%的干扰时长百分比，即实际波数/理论波数*100%*50%=8%，去掉“%”单位，得到当前BCG时域信号质量为8。
此时，S104具体可以为：根据BCG时域信号和BCG频域信号的质量，结合第一心率的范围和第二心率，计算输出最终心率。
此时的频域信号波形尽管存在伪峰，但是结合基倍频属性其二倍频、三倍频特征明显，轮廓清晰直立，基倍频特性明显。因此第二心率具有较高可靠性与准确性。而时域信号波形紊乱，较难搜索确定窗口内全部逐拍波形，且基于第二心率在时域信号计算得到的第一心率只能确定大概的心率范围无法确定其准确值。但第二心率77bpm一方面本身可靠性较高，另一方面落在第一心率的极大可能范围69~81bpm内，可认为当前窗口时间计算得到的平均心率最终输出为77bpm。实际上，可根据BCG时域信号的质量和BCG频域信号的质量计算最终心率：
最终心率最低值 = (第一心率最小值*BCG时域信号质量+第二心率*BCG频域信号质量)/( BCG时域信号质量+ BCG频域信号质量)= (69*8 + 77*95)/(8 + 95) = 76.38 (bpm)
最终心率最高值= (第一心率最大值*BCG时域信号质量+第二心率*BCG频域信号质量)/( BCG时域信号质量+ BCG频域信号质量) = (81*8 + 77*95)/(8 + 95) = 77.17 (bpm)，则最终心率等于最终心率最低值和最终心率最高值的平均值，即（76.38+ 77.17）/2= 77bpm。
当所述第一生命体征参数是第一呼吸率，所述生命体征频域信号是呼吸频域信号，所述第二生命体征参数是第二呼吸率时，所述基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数包括：
对呼吸时域信号进行重采样，根据重采样率确定时频变换的点数，进行时频变换得到相应的呼吸频域信号，对呼吸频域信号进行搜波，根据基倍频属性识别合理主峰频率来计算得到第二呼吸率；
根据第二呼吸率计算得到单周期平均时间宽度，以单周期平均时间宽度设置合理上下宽度阈值线，配合波形匹配或者特征峰谷阈值搜索出窗口内的有效波和近似有效波，得到第一呼吸率的范围；
当所述最终生命体征参数是最终呼吸率时，所述基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数包括：
根据呼吸时域信号和呼吸频域信号的质量，结合第一呼吸率的范围和第二呼吸率，计算输出最终呼吸率。
所述根据呼吸时域信号和呼吸频域信号的质量，结合第一呼吸率的范围和第二呼吸率，计算输出最终呼吸率具体为：
最终呼吸率最低值 = (第一呼吸率最小值*呼吸时域信号质量+第二呼吸率*呼吸频域信号质量)/(呼吸时域信号质量+呼吸频域信号质量)；
最终呼吸率最高值= (第一呼吸率最大值*呼吸时域信号质量+第二呼吸率*呼吸频域信号质量)/(呼吸时域信号质量+呼吸频域信号质量)；
则最终呼吸率等于最终呼吸率最低值和最终呼吸率最高值的平均值。
在本发明实施例一中，依据BCG信号计算最终心率时，时域计算与频域计算可以互为参考互为辅助，时域计算与频域计算可以依据同一时间窗口数据。在另一些实施例中，当前窗口数据的时域计算结果可以不局限于用于当前窗口数据的频域计算参考，还可以用于下一窗口数据的时域计算参考和频域计算参考。同样地，当前窗口数据的频域计算结果也可以不局限于用于当前窗口数据的时域计算参考，还可以用于下一窗口数据的频域计算参考和时域计算参考。本领域普通技术人员可对其进行简单地熟知地演化推导。
实施例二：
请参阅图12，本发明实施例二提供的生命体征信号分析处理装置包括：
获取模块11，用于获取由传感器采集的原始信号；
生命体征时域信号生成模块12，用于基于原始信号生成生命体征时域信号，所述生命体征时域信号包括BCG时域信号和/或呼吸时域信号；
第一/二生命体征参数计算模块13，用于基于生命体征时域信号计算得到第一生命体征参数，所述第一生命体征参数包括第一心率和/或第一呼吸率；将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数；所述生命体征频域信号包括BCG频域信号和/或呼吸频域信号，所述第二生命体征参数包括第二心率和/或第二呼吸率；
最终生命体征参数计算模块14，用于基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数，所述最终生命体征参数包括最终心率和/或最终呼吸率。
本发明实施例二提供的生命体征信号分析处理装置及本发明实施例一提供的生命体征信号分析处理方法属于同一构思，其具体实现过程详见说明书全文，此处不再赘述。
实施例三：
本发明实施例三还提供了一种计算机可读存储介质，所述计算机可读存储介质存储有计算机程序，所述计算机程序被处理器执行时实现如本发明实施例一提供的生命体征信号分析处理方法的步骤。
实施例四：
图13示出了本发明实施例四提供的生命体征监测设备的具体结构框图，一种生命体征监测设备100，包括：
一个或多个处理器101；
存储器102；以及
一个或多个计算机程序，其中所述一个或多个计算机程序被存储在所述存储器102中，并且被配置成由所述一个或多个处理器101执行，所述处理器101执行所述计算机程序时实现如本发明实施例一提供的生命体征信号分析处理方法的步骤。
在本发明中，由于基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数；基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数。两种方法并行处理，相互参考辅助计算，在满足实时性的同时增强抗干扰能力，可大大提升计算结果的准确性和可靠性。能够在BCG信号存在一定干扰和采集信号质量不高的情况下测量心率和呼吸率，性能稳定，结果精确。
以上所述仅为本发明的较佳实施例而已，并不用以限制本发明，凡在本发明的精神和原则之内所作的任何修改、等同替换和改进等，均应包含在本发明的保护范围之内。

Claims (12)

  1. 一种生命体征信号分析处理方法，其特征在于，所述方法包括：
    获取由传感器采集的原始信号；
    基于原始信号生成生命体征时域信号，所述生命体征时域信号包括BCG时域信号和/或呼吸时域信号；
    基于生命体征时域信号计算得到第一生命体征参数，所述第一生命体征参数包括第一心率和/或第一呼吸率；将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数；所述生命体征频域信号包括BCG频域信号和/或呼吸频域信号，所述第二生命体征参数包括第二心率和/或第二呼吸率；
    基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数，所述最终生命体征参数包括最终心率和/或最终呼吸率。
  2. 如权利要求1所述的方法，其特征在于，当所述生命体征时域信号是BCG时域信号时，所述基于原始信号生成生命体征时域信号包括：
    通过心率计算所需生理频率带宽范围的滤波器对采集的原始信号进行滤波去噪；
    根据信号动态范围对滤波去噪后的原始信号进行缩放得到BCG时域信号；
    当所述生命体征时域信号是呼吸时域信号时，所述基于原始信号生成生命体征时域信号包括：
    通过呼吸率计算所需生理频率带宽范围的滤波器对采集的原始信号进行滤波去噪；
    根据信号动态范围对滤波去噪后的原始信号进行缩放得到呼吸时域信号。
  3. 如权利要求1述的方法，其特征在于，
    当所述第一生命体征参数是第一心率，所述生命体征频域信号是BCG频域信号，所述第二生命体征参数是第二心率时，所述基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数包括：
    根据BCG时域信号波形特征，搜索识别时域波形的各周期特征峰谷，计算逐拍心率，并根据预设时长或者预设节拍数计算得到平均心率作为第一心率；
    将预设时长的BCG时域信号进行重采样，根据重采样率确定时频变换的点数，进行时频变换得到相应的BCG频域信号，对BCG频域信号进行搜波，根据基倍频属性识别合理主峰频率来计算得到第二心率；
    当所述最终生命体征参数是最终心率时，所述基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数包括：
    根据BCG时域信号和BCG频域信号的质量，结合第一心率和第二心率，计算输出最终心率。
  4. 如权利要求1述的方法，其特征在于，当所述第一生命体征参数是第一呼吸率，所述生命体征频域信号是呼吸频域信号，所述第二生命体征参数是第二呼吸率时，所述基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数包括：
    根据呼吸时域信号波形特征，搜索识别时域波形的各周期特征峰谷，计算逐拍呼吸率，并根据预设时长或者预设节拍数计算得到平均呼吸率作为第一呼吸率；
    将预设时长的呼吸时域信号进行重采样，根据重采样率确定时频变换的点数，进行时频变换得到相应的呼吸频域信号，对呼吸频域信号进行搜波，根据基倍频属性识别合理主峰频率来计算得到第二呼吸率；
    当所述最终生命体征参数是最终呼吸率时，所述基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数包括：
    根据呼吸时域信号和呼吸频域信号的质量，结合第一呼吸率和第二呼吸率，计算输出最终呼吸率。
  5. 如权利要求1所述的方法，其特征在于，
    当所述第一生命体征参数是第一心率，所述生命体征频域信号是BCG频域信号，所述第二生命体征参数是第二心率时，所述基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数包括：
    根据BCG时域信号波形特征，搜索识别时域波形的各周期特征峰谷，计算逐拍心率，并根据预设时长或者预设节拍数计算得到平均心率作为第一心率；
    将预设时长的BCG时域信号进行时频变换得到相应的BCG频域信号，搜索BCG频域信号中在第一心率附近的主峰，并根据基倍频属性验证后，将 通过验证的主峰对应的频率作为第二心率；
    当所述最终生命体征参数是最终心率时，所述基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数包括：
    根据BCG时域信号和BCG频域信号的质量，结合第一心率和第二心率，计算输出最终心率。
  6. 如权利要求1所述的方法，其特征在于，当所述第一生命体征参数是第一呼吸率，所述生命体征频域信号是呼吸频域信号，所述第二生命体征参数是第二呼吸率时，所述基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数包括：
    根据呼吸时域信号波形特征，搜索识别时域波形的各周期特征峰谷，计算逐拍呼吸率，并根据预设时长或者预设节拍数计算得到平均呼吸率作为第一呼吸率；
    将预设时长的呼吸时域信号进行时频变换得到相应的呼吸频域信号，搜索呼吸频域信号中在第一呼吸率附近的主峰，并根据基倍频属性验证后，将通过验证的主峰对应的频率作为第二呼吸率；
    当所述最终生命体征参数是最终呼吸率时，所述基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数包括：
    根据呼吸时域信号和呼吸频域信号的质量，结合第一呼吸率和第二呼吸率，计算输出最终呼吸率。
  7. 如权利要求3至6任一项所述的方法，其特征在于，
    所述根据BCG时域信号和BCG频域信号的质量，结合第一心率和第二心率，计算输出最终心率具体为：
    最终心率= (第一心率*BCG时域信号的质量+第二心率*BCG频域信号的质量)/( BCG时域信号的质量+ BCG频域信号的质量)；
    所述根据呼吸时域信号和呼吸频域信号的质量，结合第一呼吸率和第二呼吸率，计算输出最终呼吸率具体为：
    最终呼吸率= (第一呼吸率*呼吸时域信号的质量+第二呼吸率*呼吸频域信号的质量)/(呼吸时域信号的质量+呼吸频域信号的质量)。
  8. 如权利要求1所述的方法，其特征在于，
    当所述第一生命体征参数是第一心率，所述生命体征频域信号是BCG频域信号，所述第二生命体征参数是第二心率时，所述基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数包括：
    将预设时长的BCG时域信号进行重采样，根据重采样率确定时频变换的点数，进行时频变换得到相应的BCG频域信号，对BCG频域信号进行搜波，根据基倍频属性识别合理主峰频率来计算得到第二心率；
    根据第二心率计算得到单周期平均时间宽度，以单周期平均时间宽度设置合理上下宽度阈值线，配合波形匹配或者特征峰谷阈值搜索出窗口内的有效波和近似有效波，得到第一心率的范围；
    当所述最终生命体征参数是最终心率时，所述基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数包括：
    根据BCG时域信号和BCG频域信号的质量，结合第一心率的范围和第二心率，计算输出最终心率。
  9. 如权利要求1所述的方法，其特征在于，当所述第一生命体征参数是第一呼吸率，所述生命体征频域信号是呼吸频域信号，所述第二生命体征参数是第二呼吸率时，所述基于生命体征时域信号计算得到第一生命体征参数，将预设时长的生命体征时域信号进行时频变换得到生命体征频域信号，并基于生命体征频域信号计算得到第二生命体征参数包括：
    将预设时长的呼吸时域信号进行重采样，根据重采样率确定时频变换的点数，进行时频变换得到相应的呼吸频域信号，对呼吸频域信号进行搜波，根据基倍频属性识别合理主峰频率来计算得到第二呼吸率；
    根据第二呼吸率计算得到单周期平均时间宽度，以单周期平均时间宽度设置合理上下宽度阈值线，配合波形匹配或者特征峰谷阈值搜索出窗口内的有效波和近似有效波，得到第一呼吸率的范围；
    当所述最终生命体征参数是最终呼吸率时，所述基于第一生命体征参数和第二生命体征参数，计算得到最终生命体征参数包括：
    根据呼吸时域信号和呼吸频域信号的质量，结合第一呼吸率的范围和第二呼吸率，计算输出最终呼吸率。
  10. 如权利要求8或9所述的方法，其特征在于，
    所述根据BCG时域信号和BCG频域信号的质量，结合第一心率的范围和第二心率，计算输出最终心率具体为：
    最终心率最低值 = (第一心率最小值*BCG时域信号质量+第二心率*BCG频域信号质量)/( BCG时域信号质量+ BCG频域信号质量)；
    最终心率最高值= (第一心率最大值*BCG时域信号质量+第二心率*BCG频域信号质量)/( BCG时域信号质量+ BCG频域信号质量)；
    则最终心率等于最终心率最低值和最终心率最高值的平均值；
    所述根据呼吸时域信号和呼吸频域信号的质量，结合第一呼吸率的范围和第二呼吸率，计算输出最终呼吸率具体为：
    最终呼吸率最低值 = (第一呼吸率最小值*呼吸时域信号质量+第二呼吸率*呼吸频域信号质量)/(呼吸时域信号质量+呼吸频域信号质量)；
    最终呼吸率最高值= (第一呼吸率最大值*呼吸时域信号质量+第二呼吸率*呼吸频域信号质量)/(呼吸时域信号质量+呼吸频域信号质量)；
    则最终呼吸率等于最终呼吸率最低值和最终呼吸率最高值的平均值。
  11.  一种计算机可读存储介质，所述计算机可读存储介质存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现如权利要求1至10任一项所述的生命体征信号分析处理方法的步骤。
  12. 　一种生命体征监测设备，包括：
    一个或多个处理器；
    存储器；以及
    一个或多个计算机程序，其中所述一个或多个计算机程序被存储在所述存储器中，并且被配置成由所述一个或多个处理器执行，其特征在于，所述处理器执行所述计算机程序时实现如权利要求1至10任一项所述的生命体征信号分析处理方法的步骤。
PCT/CN2018/115490 2017-11-14 2018-11-14 一种生命体征信号分析处理方法和生命体征监测设备 WO2019096175A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711124781.5A CN108056769B (zh) 2017-11-14 2017-11-14 一种生命体征信号分析处理方法、装置和生命体征监测设备
CN201711124781.5 2017-11-14

Publications (1)

Publication Number Publication Date
WO2019096175A1 true WO2019096175A1 (zh) 2019-05-23

Family

ID=62134856

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/115490 WO2019096175A1 (zh) 2017-11-14 2018-11-14 一种生命体征信号分析处理方法和生命体征监测设备

Country Status (2)

Country Link
CN (1) CN108056769B (zh)
WO (1) WO2019096175A1 (zh)

Families Citing this family (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108056769B (zh) * 2017-11-14 2020-10-16 深圳市大耳马科技有限公司 一种生命体征信号分析处理方法、装置和生命体征监测设备
CN109222928B (zh) * 2018-08-03 2022-08-23 深圳市大耳马科技有限公司 呼吸信号的提取方法、装置、处理设备和系统
CN109009023A (zh) * 2018-08-03 2018-12-18 深圳市大耳马科技有限公司 一种呼吸信号的提取方法、装置、处理设备和系统
WO2020051741A1 (zh) * 2018-09-10 2020-03-19 深圳市大耳马科技有限公司 心脏生理参数测量方法、设备、终端及计算机存储介质
CN110881956B (zh) * 2018-09-10 2022-06-14 深圳市大耳马科技有限公司 心脏生理参数测量方法、设备、终端及计算机存储介质
US20220039692A1 (en) * 2018-09-11 2022-02-10 Shenzhen Darma Technology Co., Ltd. Method and processing device for assessing volume responsiveness
WO2020061887A1 (zh) * 2018-09-27 2020-04-02 深圳市伊欧乐科技有限公司 心率测量方法、装置及计算机可读存储介质
CN113242716B (zh) * 2019-02-01 2022-11-11 深圳市大耳马科技有限公司 一种心律失常事件的监测方法、设备和系统
CN109965885A (zh) * 2019-04-24 2019-07-05 中国科学院电子学研究所 一种基于去噪自动编码器的bcg信号降噪方法及装置
WO2020232604A1 (zh) * 2019-05-20 2020-11-26 深圳市大耳马科技有限公司 一种心脏舒张功能评估方法、设备和系统
WO2020232606A1 (zh) * 2019-05-20 2020-11-26 深圳市大耳马科技有限公司 一种心脏舒张功能评估方法、设备和系统
TWI750504B (zh) * 2019-08-29 2021-12-21 鉅怡智慧股份有限公司 活體偵測的方法及相關裝置
CN111481173B (zh) * 2020-04-15 2023-06-06 上海贝氪若宝健康科技有限公司 身体体征信号的检测方法、介质、设备及系统
CN114246579B (zh) * 2020-09-23 2024-03-15 深圳绿米联创科技有限公司 心率值确定方法及装置、终端设备及存储介质
CN114533010A (zh) * 2020-11-25 2022-05-27 安徽华米健康科技有限公司 心率检测方法和装置
CN113261933A (zh) * 2021-06-21 2021-08-17 淮北师范大学 一种基于Mesh网的矿工心电信号无线监测方法及检测系统
CN113729678B (zh) * 2021-10-12 2023-11-21 南京润楠医疗电子研究院有限公司 一种基于毫米波雷达的无线体征测量系统
CN114145735B (zh) * 2021-12-03 2023-10-20 华南师范大学 基于bcg信号的运动规划方案生成方法、系统以及设备
CN114795168B (zh) * 2022-06-24 2022-09-30 昂科信息技术（上海）股份有限公司 生命体征参数心率计算方法及系统
CN115486833B (zh) * 2022-08-22 2023-06-06 华南师范大学 呼吸状态检测方法、装置、计算机设备以及存储介质
CN115316985B (zh) * 2022-10-13 2023-04-18 华南师范大学 基于生理信号的心脏信息检测方法、装置以及设备

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102988036A (zh) * 2012-12-26 2013-03-27 中国科学院自动化研究所 测量脉率的方法
CN105816163A (zh) * 2016-05-09 2016-08-03 安徽华米信息科技有限公司 检测心率的方法、装置及可穿戴设备
CN106606358A (zh) * 2015-10-23 2017-05-03 富港电子(昆山)有限公司 测量心率的方法及其装置
EP3181038A1 (en) * 2015-12-14 2017-06-21 Cheng Uei Precision Industry Co., Ltd. Heart rate measurement method and heart rate measurement device applying the same
CN106880351A (zh) * 2015-12-15 2017-06-23 德州仪器公司 减少光体积描记(ppg)信号中运动诱发的伪影
US9724016B1 (en) * 2009-10-16 2017-08-08 Masimo Corp. Respiration processor
CN107233086A (zh) * 2016-03-28 2017-10-10 三星电子株式会社 使用低功率传感器进行心率和呼吸率估计的方法和装置
CN108056769A (zh) * 2017-11-14 2018-05-22 深圳市大耳马科技有限公司 一种生命体征信号分析处理方法、装置和生命体征监测设备

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100210962A1 (en) * 2009-02-13 2010-08-19 Jingping Xu Respiratory signal detection and time domain signal processing method and system
CN101843489A (zh) * 2009-03-26 2010-09-29 深圳市理邦精密仪器有限公司 一种呼吸信号处理方法
EP2434947A4 (en) * 2009-05-29 2015-07-29 Univ Yale APPARATUS, SYSTEMS AND METHODS USING PLYTHYSMOGRAPHIC DATA
CN105662375B (zh) * 2016-03-17 2019-01-29 广州中科新知科技有限公司 非接触式检测生命体征信号的方法和装置

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9724016B1 (en) * 2009-10-16 2017-08-08 Masimo Corp. Respiration processor
CN102988036A (zh) * 2012-12-26 2013-03-27 中国科学院自动化研究所 测量脉率的方法
CN106606358A (zh) * 2015-10-23 2017-05-03 富港电子(昆山)有限公司 测量心率的方法及其装置
EP3181038A1 (en) * 2015-12-14 2017-06-21 Cheng Uei Precision Industry Co., Ltd. Heart rate measurement method and heart rate measurement device applying the same
CN106880351A (zh) * 2015-12-15 2017-06-23 德州仪器公司 减少光体积描记(ppg)信号中运动诱发的伪影
CN107233086A (zh) * 2016-03-28 2017-10-10 三星电子株式会社 使用低功率传感器进行心率和呼吸率估计的方法和装置
CN105816163A (zh) * 2016-05-09 2016-08-03 安徽华米信息科技有限公司 检测心率的方法、装置及可穿戴设备
CN108056769A (zh) * 2017-11-14 2018-05-22 深圳市大耳马科技有限公司 一种生命体征信号分析处理方法、装置和生命体征监测设备

Also Published As

Publication number Publication date
CN108056769B (zh) 2020-10-16
CN108056769A (zh) 2018-05-22

Similar Documents

Publication Publication Date Title
CN108056769B (zh) 一种生命体征信号分析处理方法、装置和生命体征监测设备
Paalasmaa et al. Adaptive heartbeat modeling for beat-to-beat heart rate measurement in ballistocardiograms
CN107198516B (zh) 一种无束缚检测呼吸率心率的方法及智能床
KR101656611B1 (ko) 무구속적으로 측정한 생체신호를 이용하여 산소탈포화지수를 획득하는 방법
Boyle et al. Automatic detection of respiration rate from ambulatory single-lead ECG
CN105662375B (zh) 非接触式检测生命体征信号的方法和装置
CN108882867A (zh) 生物体信息分析装置、系统以及程序
CN105662345B (zh) 心跳信号处理方法、装置和系统
Rosales et al. Heart rate monitoring using hydraulic bed sensor ballistocardiogram 1
WO2020019358A1 (zh) 一种生理参数测量系统及配备该测量系统的智能座椅
Landreani et al. Smartphone accelerometers for the detection of heart rate
Sun et al. PPG signal motion artifacts correction algorithm based on feature estimation
CN109222928B (zh) 呼吸信号的提取方法、装置、处理设备和系统
CN109009023A (zh) 一种呼吸信号的提取方法、装置、处理设备和系统
CN113598728B (zh) 生理信号的降噪方法、监测方法、监测装置及可穿戴设备
Cao et al. Guard Your Heart Silently: Continuous Electrocardiogram Waveform Monitoring with Wrist-Worn Motion Sensor
Hsiao et al. Design and implementation of auscultation blood pressure measurement using vascular transit time and physiological parameters
TW201838587A (zh) 分析心衝擊信號用來計算短期心率值的方法
Berset et al. Robust heart rhythm calculation and respiration rate estimation in ambulatory ECG monitoring
Kansara et al. Heart Rate Measurement
Shilvya et al. Obstructive Sleep Apnea Detection from ECG Signals with Deep Learning
AU2020346386A1 (en) Method for determining respiratory rate
Akbulut et al. Estimation of Beat-to-Beat Interval from Wearable Photoplethysmography Sensor on Different Measurement Sites During Daily Activities
CN116369888B (zh) 一种非接触式心率变异性数据获取方法和装置
Tian et al. SleepABP: Noninvasive Ambulatory Blood Pressure Monitoring Based on Ballistocardiogram in Sleep State

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18878062

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 25.09.2020)

122 Ep: pct application non-entry in european phase

Ref document number: 18878062

Country of ref document: EP

Kind code of ref document: A1