WO2018152713A1 - 一种用于血压测量装置的数据处理方法 - Google Patents

一种用于血压测量装置的数据处理方法 Download PDF

Info

Publication number
WO2018152713A1
WO2018152713A1 PCT/CN2017/074439 CN2017074439W WO2018152713A1 WO 2018152713 A1 WO2018152713 A1 WO 2018152713A1 CN 2017074439 W CN2017074439 W CN 2017074439W WO 2018152713 A1 WO2018152713 A1 WO 2018152713A1
Authority
WO
WIPO (PCT)
Prior art keywords
blood pressure
point
pulse wave
value
wave
Prior art date
Application number
PCT/CN2017/074439
Other languages
English (en)
French (fr)
Inventor
张跃
潘俊俊
张拓
Original Assignee
清华大学深圳研究生院
深圳市岩尚科技有限公司
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 清华大学深圳研究生院, 深圳市岩尚科技有限公司 filed Critical 清华大学深圳研究生院
Priority to CN201780002087.8A priority Critical patent/CN107995981B/zh
Priority to PCT/CN2017/074439 priority patent/WO2018152713A1/zh
Publication of WO2018152713A1 publication Critical patent/WO2018152713A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/02—Preprocessing
    • 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
    • A61B5/021—Measuring pressure in heart or blood vessels
    • A61B5/02108—Measuring pressure in heart or blood vessels from analysis of pulse wave characteristics
    • 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
    • A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
    • 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
    • A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching
    • G06F2218/16—Classification; Matching by matching signal segments
    • G06F2218/20—Classification; Matching by matching signal segments by applying autoregressive analysis

Definitions

  • the present invention relates to a blood pressure measuring device, and more particularly to a data processing method for a blood pressure measuring device.
  • Blood pressure is one of the important physiological parameters of the human body and can reflect the cardiovascular condition of the human body. It is an important basis for clinical diagnosis of diseases, surgical indicators and monitoring of vital signs.
  • the blood pressure measuring device is divided into two types: invasive and non-invasive.
  • the invasive measuring device adopts the arterial intubation method, and the obtained blood pressure is the most accurate, but its technical requirements are high, and it is only suitable for clinical use because it causes damage to the human body. On critically ill patients and so on.
  • the non-invasive blood pressure device is further divided into an intermittent blood pressure measuring device and a continuous blood pressure measuring device.
  • Intermittent measuring devices including Koch's listening blood pressure measuring device, oscillometric blood measuring device, and constant volume measuring blood pressure device, require a cuff, cannot be used for long-term measurement, and can only measure a specific The blood pressure at the moment cannot meet the clinical needs.
  • a device for applying blood pressure measurement using a pulse wave is proposed.
  • the measurement of pulse wave velocity requires at least two points of acquisition, and the device is complicated and inconvenient to carry around.
  • the pulse wave can be collected at a single point on the wrist or on the finger to perform the measurement.
  • the blood pressure measuring device using the pulse wave characteristic parameter basically establishes a single one or more regression equations according to the collected data. Since the regression method has randomness, it is difficult to ensure that the established regression equation can be achieved on the fitted data set. Globally optimal.
  • the actual measured data has a certain randomness, including the randomness of the collected data and the randomness of the regression equation. This randomness adversely affects the utility of the blood pressure measuring device.
  • the present invention proposes a blood pressure measuring device and a data processing method therefor.
  • a blood pressure measuring device The data processing method includes the steps of: establishing a plurality of blood pressure estimation regression equations and calculating a blood pressure value, wherein the step of establishing a plurality of blood pressure estimation regression equations comprises: S11, acquiring a plurality of pulse waves and corresponding blood pressure value sets; S12, preprocessing Obtaining a pulse wave; S13, extracting a pulse wave feature point from the preprocessed pulse wave, and calculating a pulse wave global feature parameter value; S14, randomly selecting a pulse wave global feature parameter value and a corresponding blood pressure value set Establishing a plurality of sets of regression test sets, wherein the regression test set is composed of a test set and a training set; S15, obtaining a global optimal regression equation for each set of regression test sets; S16, evaluating and screening out a regression equation with high accuracy; S17, The high-accuracy regression equation is given corresponding weights; the step of calculating the blood pressure value includes
  • a data processing method for a blood pressure measuring device comprising the steps of establishing a plurality of blood pressure estimation regression equations and calculating blood pressure values.
  • the steps of establishing a plurality of blood pressure estimation regression equations include acquiring a pulse wave and a corresponding blood pressure value set, pre-processing the acquired pulse wave, extracting pulse wave feature points, calculating a pulse wave global characteristic parameter value, and using the acquired pulse wave global characteristic parameter value and Corresponding blood pressure value sets are established by randomly selecting multiple sets of regression test sets, and the global optimal regression equations of each set of regression test sets are obtained, and the regression equations with high accuracy are evaluated and screened, and the corresponding weights are assigned.
  • the step of calculating the blood pressure value includes substituting the acquired global characteristic parameters of the pulse wave into the selected regression equation, removing the abnormal value from the calculated result to obtain a plurality of blood pressure values, and finally weighting the average to improve the accuracy and stability of the blood pressure measurement method. Sex.
  • FIG. 1 is a typical waveform diagram of a periodic pulse wave marked with a feature point provided by the present invention.
  • FIG. 2 is a flow chart of establishing a plurality of blood pressure estimation regression equations provided by the present invention.
  • FIG. 3 is a flow chart of calculating blood pressure values provided by the present invention.
  • FIG. 4 is a block diagram showing the structure of a blood pressure measuring device provided by the present invention.
  • FIG. 1 is a typical waveform diagram of a periodic pulse wave marked with a feature point
  • B represents an aortic valve open point
  • C represents a maximum systolic pressure point
  • F represents a re-pulsation wave start point
  • G represents a maximum beat wave maximum pressure point.
  • the present invention also provides a data processing method for the above blood pressure processing apparatus, comprising the steps of establishing a plurality of blood pressure estimation regression equations and calculating blood pressure values, as shown in FIG. 2, the steps of establishing a plurality of blood pressure estimation regression equations include:
  • a plurality of sets of regression test sets are established by using a random selection manner, where the regression test set is composed of a test set and a training set;
  • the step of calculating a blood pressure value includes:
  • the corresponding blood pressure value acquired in step 101 is to establish a plurality of regression equations.
  • the global optimal regression equation is stepped back in step 105.
  • the regression analysis method acquires the training set in a global traversal manner.
  • the pre-processing pulse wave includes removing baseline drift and low-pass filtering, and the training data in the training set is not less than 8 groups.
  • the removal of the baseline drift can be processed using a median filter, the formula is as follows:
  • P is the original pulse wave data
  • B is the baseline drift of the original pulse wave
  • Pa is the pulse wave data after the baseline drift is removed.
  • Len is the data length of P.
  • m is a constant, and the number of data points in one cycle of the pulse wave is num, then the best value of m is ( Round up the symbol).
  • the low-pass filter mainly filters out high-frequency noise and power frequency interference.
  • the Chebyshev I-type filter can be used to set the corresponding parameters: the passband boundary frequency is about 10Hz, the stopband boundary frequency is about 12Hz, and the passband is the largest. Attenuation is 0.1dB and the stopband has a minimum attenuation of 20dB.
  • the pulse wave data after Pa filtering is Pb.
  • the pulse wave feature points described in step 103 or step 203 include an aortic valve open point, a systolic maximum pressure point, a severe beat wave origin, and a severe beat wave highest pressure point; Extracting the pulse wave feature points includes smoothing processing of the pulse wave, and the smoothing processing adopts a three-point line smoothing process.
  • the formula for the three-line smoothing method is: Among them, Pc represents the beat wave data point, and Pb represents the pulse wave data after the filtering process.
  • the plurality of sets of regression test sets are established by using the obtained n pairs of pulse wave global feature parameters and corresponding blood pressure values as 1 to n groups of sample data, and grouping each group of data by 1 to n.
  • the random number in the range of 1 to n is generated by generating a random number, and then the data corresponding to the corresponding random number is selected and repeated until the k group data is selected.
  • the selected k sets of data are used as the training set, and the remaining n-k sets of data are used as test sets.
  • Such training sets and test sets constitute a regression test set.
  • the steps of extracting the aortic valve opening point and the systolic maximum pressure point include:
  • T1 obtaining all generalized extreme points of the pulse wave, and obtaining an extreme point set; specifically, traversing each data point of the pulse wave from beginning to end, and determining whether the point is an extreme point according to the extreme point judgment condition;
  • T2 determining a threshold value of the difference between the aortic valve opening point B and the systolic maximum pressure point C; specifically, sorting the extreme point points obtained in the previous step in ascending order by magnitude to obtain the sorted extreme point set extS , respectively, taking thn extreme points from both ends of the sorted extreme point set (ie, the minimum data part and the largest data part), and determining the degree of dispersion of the thn points, if the discrete program is less than the acceptable standard degree da, Then, these thn points are used as the representative point set of point B or point C of the pulse wave data of the present section.
  • the representative point set of point B is the point that satisfies the requirement found from the minimum data part
  • the representative point set of point C is the point that satisfies the requirement found from the largest data part. More specifically, thn takes 2 to 5, and da takes 0.003 to 0.01.
  • the threshold formula for the difference between B and C is:
  • Thd ⁇ *(dC-dB), where the threshold coefficient is 0.5 to 0.7.
  • the extreme point judgment condition is: (Pc[i] - Pc[i-1]) * (Pc[i+1] - Pc[i]) ⁇ 0, and Pc represents a beat wave. data point.
  • step T3 the discriminant according to the threshold value for determining and extracting the pulse opening point and the systolic maximum pressure point is ext[i+1]-ext[i]>thd, 1 ⁇ i ⁇ len-1, Ext[i] represents the ith extreme point, ext[i+1] represents the i+1th extreme point, thd represents the threshold of the aortic valve opening point, the highest pressure point difference during systole, and len represents the extreme value. The number of points.
  • the steps of extracting the starting point F of the re-pulsation wave and the highest pressure point G of the re-pulsation wave include:
  • the period interval of the divided pulse wave; specifically, the B point extracted from above is a boundary to divide each segment of the pulse wave.
  • E2 Determine the pulse wave period interval where the peak of the re-pulsation wave and the highest pressure point of the re-pulsation wave are located; specifically, the ratio of the length between the C and F points in the period rcf, the length between the C and the G point in each period The ratio rgg in the cycle.
  • the interval interFG in which F and G are located in each cycle is determined by rcf, rcg, and the extracted points B and C.
  • rcf is from 0.08 to 0.12
  • rcg is from 0.45 to 0.55.
  • angChg[i] atan((p1-p2)/ts)-atan((p2-p3)/ts);
  • ts is the pulse wave data sampling period and atan is the inverse tangent value function. Take the point corresponding to the maximum value of angChg as point F, and take the point corresponding to the minimum value of angChg as point G.
  • acquiring the pulse wave global feature parameter value in step 103 or step 203 includes removing the outlier value and averaging each feature parameter set for removing the outlier value;
  • the pulse wave global feature parameter value includes a global Systolic time ratio, global main wave height, global relative gorge relative height, global tremor wave relative height, global systolic area ratio, global main wave rising slope and global K value; The Weiler method is carried out.
  • the beat wave global feature parameter value is obtained by calculating a feature parameter of each period of the pulse wave as a feature parameter set, and then acquiring from the feature parameter set.
  • the calculation of the single-cycle characteristic parameter value includes periodically segmenting the pulse wave according to the feature point extracted as described above, and acquiring the pulse wave data periodData of each single cycle, and then pairing the single cycle.
  • the pulse wave data is processed and the characteristic parameters are calculated.
  • the single-cycle data waveform is normalized: the baseline portion of the single-cycle data is removed, and the linear equation is passed. Fit the baseline and subtract the baseline from the original data to obtain the normalized data nomalData.
  • the formula is as follows:
  • romalData[i] periodData[i]-(kBase*i+bBase);1 ⁇ i ⁇ lenP.
  • periodData is the extracted waveform data
  • nomalData is the normalized waveform data
  • kBase is the baseline slope of the single-cycle waveform
  • bBase is the baseline intercept of the single-cycle waveform
  • i is the data point of the waveform The ith.
  • the amplitude of point B in a single cycle is zero.
  • single-cycle feature parameter calculations include:
  • the pulse wave from point B to point F is the pulse wave systole, and sInT is the ratio of the length of the BF segment to the entire period. Since the sampling rate is constant, the time is proportional to the sample length. Here, the value of the BF segment data length is compared with the value of the current cycle data length as sInT.
  • the main wave height corresponds to the amplitude of the C point, where the normalized C point amplitude is taken as mainH.
  • the height of the lower middle gorge corresponds to the amplitude of the F point
  • the gorgeInMainH refers to the height ratio of the height of the lower middle gorge relative to the main wave.
  • the ratio of the F-point amplitude to the mainH after normalization is taken as gorgeInMainH.
  • the beat wave height corresponds to the G point amplitude
  • the repeatInMianH refers to the height ratio of the beat wave height to the main wave.
  • the ratio of the G-point amplitude to the mainH after normalization is taken as repeatInMianH.
  • sysInArea refers to the ratio of the systolic area of the pulse wave waveform to the area of the entire period waveform.
  • the area calculation uses the following formula:
  • end corresponds to the position of the F point.
  • end lenP.
  • slopeMain is the average slope from point B to point C.
  • the formula is as follows:
  • lenBC is the number of sampling points of the BC segment pulse wave.
  • S period is the area of the entire periodic waveform.
  • param is the set of final selection parameters
  • coef is the set of corresponding parameters of each parameter
  • cont is a constant term
  • lenParam is the number of selected parameters
  • BPest is the estimated blood pressure value.
  • the accuracy of the regression equation is evaluated using the regression equation obtained above and the test set corresponding to the regression test set.
  • the assessment method is as follows:
  • each pulse wave global characteristic parameter data in the test set into the regression equation obtained by the above formula the corresponding estimated blood pressure value BPest is obtained, and BPest is compared with the blood pressure value BPval corresponding to the test set to obtain an estimation error of each data ⁇ BPest- BPval ⁇ , and the error value is divided into three paragraphs, ⁇ 5, ⁇ 10, ⁇ 15, calculate the proportion of test data in each segment.
  • the accuracy level of the regression equation is divided into four levels: A, B, C, and D.
  • the evaluation criteria are as follows:
  • the regression equation with lower accuracy level is removed, and the remaining regression equations are given different weights according to the accuracy level. The higher the level, the higher the weight.
  • the regression equation with the accuracy level below B is removed, the weight (w) corresponding to the regression equation of the level A is set to a, and the weight corresponding to the regression equation of the grade B is b.
  • the calculation formula of the weighting coefficient in step 205 is:
  • lenW is the number of blood pressure estimation values after the abnormal value is removed
  • wR[i] is the weighting coefficient corresponding to the i-th blood pressure estimation value.
  • the calculation formula of the blood pressure measurement value in step 206 is:
  • BPest is a set of blood pressure estimation values after the abnormal value is removed
  • wR[i] is a weighting coefficient corresponding to the i-th blood pressure estimation value
  • BP is a blood pressure measurement value.
  • the present invention also provides the use of the data processing method for a blood pressure measuring device according to any of the above, in a blood pressure measuring device or a physiological multi-parameter monitoring device.
  • the data processing method for the blood pressure measuring device is not limited to the application in the blood pressure measuring device or the physiological multi-parameter monitoring device, and the method is applicable to other devices having a blood pressure data processing function, the device including a mobile phone, a wristband, Rings, foot rings, computers, tablets, etc.
  • the blood pressure measuring device includes a plurality of regression equation establishing modules and a blood pressure estimating module, and the blood pressure estimating regression equation establishing module includes:
  • a template data acquisition module configured to acquire a plurality of pulse waves and corresponding blood pressure values
  • a data preprocessing module configured to preprocess a pulse wave acquired from the template data acquisition module
  • a template feature parameter calculation module configured to extract a pulse wave feature point from a pulse wave preprocessed by the preprocessing module, and calculate a pulse wave global feature parameter value by using the feature point;
  • the regression test set establishing module establishes a plurality of sets of regression test sets by randomly selecting a sample set consisting of the pulse wave global feature parameter value obtained from the template feature parameter calculation module and the corresponding blood pressure value obtained from the template data acquisition module
  • the regression test set consists of a test set and a training set;
  • a preliminary regression equation building module is used to obtain a globally optimal regression equation for each set of regression test sets
  • the regression equation evaluation and selection module is used for evaluating and screening out the regression equation with high accuracy in the regression equation building module, and assigning the corresponding weight to the accuracy equation with high accuracy according to the accuracy level;
  • the blood pressure estimation module includes:
  • a pulse wave acquisition module for acquiring a pulse wave
  • a data preprocessing module for preprocessing the pulse wave collected from the pulse wave acquisition module
  • a characteristic parameter calculation module configured to extract a pulse wave feature point from a pulse wave preprocessed in the preprocessing module, and calculate a pulse wave global feature parameter value by using the feature point;
  • the preliminary blood pressure value estimation module is used to calculate the pulse wave global feature parameters acquired in the feature parameter calculation module into multiple sets of regression equations in the plurality of regression equation building modules, and remove the abnormal values in the calculation results to obtain multiple blood pressure estimates. value;
  • the weighting coefficient calculation module calculates a corresponding weighting coefficient by using weights of regression equations corresponding to the plurality of blood pressure values obtained in the preliminary blood pressure value estimation module;
  • the blood pressure estimation output module performs weighted averaging on the plurality of blood pressure estimation values obtained in the preliminary blood pressure value estimation module by using the weighting coefficient obtained in the weighting coefficient calculation module to obtain a blood pressure measurement value.
  • the blood pressure value corresponding to the pulse wave in the template data collection module may be acquired by using a cuff type mercury sphygmomanometer or an electronic sphygmomanometer.
  • the physiological multi-parameter monitoring device processes blood pressure data by acquiring a pulse wave signal to obtain a blood pressure measurement value.
  • the physiological multi-parameter monitoring device is a patch-type physiological multi-parameter monitoring device, and includes a shell of a flat flexible material suitable for attaching to human skin, and the front surface of the shell can be attached to the human body through a sticker.
  • a pulse wave sensor is disposed on the front side and/or the back side of the casing, and is integrally formed with the casing by liquid silicone injection molding or solid silicone molding, and the casing is sealed for processing the collected physiological a circuit module of parameter data, wherein the circuit module is connected to the pulse wave sensor; the pulse wave sensor performs pulse wave signal acquisition, and transmits the collected data to a circuit module in the housing.
  • the acquired pulse wave signal may calculate at least one physiological parameter including blood pressure, blood oxygen saturation, heart rate, respiratory rate, and maximum oxygen uptake, and the pulse wave sensor preferably uses a photoelectric pulse wave sensor.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • Molecular Biology (AREA)
  • General Health & Medical Sciences (AREA)
  • Veterinary Medicine (AREA)
  • Public Health (AREA)
  • Animal Behavior & Ethology (AREA)
  • Physiology (AREA)
  • Surgery (AREA)
  • Medical Informatics (AREA)
  • Biophysics (AREA)
  • Pathology (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Psychiatry (AREA)
  • Cardiology (AREA)
  • Power Engineering (AREA)
  • Theoretical Computer Science (AREA)
  • Vascular Medicine (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)

Abstract

一种用于血压测量装置的数据处理方法,包括建立多个血压估算回归方程的步骤和计算血压值的步骤。其中,建立多个血压估算回归方程的步骤包括获取脉搏波及对应的血压值（101）；预处理获取的脉搏波（102）；提取脉搏波特征点，获取脉搏波全局特征参数值（103）；利用获取的脉搏波全局特征参数值和对应的血压值，通过随机选取的方式建立多组回归测试集（104）；获取每组回归测试集中全局最优的回归方程（105）；评估并筛选出准确度高的回归方程（106）；赋予相应的权值（107）。计算血压值的步骤包括将获取的脉搏波全局特征参数代入准确度高的回归方程中，对计算结果去除异常值得到多个血压估计值，最后加权取平均。该血压测量装置的数据处理方法提高了血压测量值的准确性和稳定性。

Description

一种用于血压测量装置的数据处理方法 技术领域
本发明涉及血压测量装置，特别是涉及一种用于血压测量装置的数据处理方法。
背景技术
本项研究工作得到了中国国家自然科学基金资助(项目批准号：61571268)。
血压是人体的重要生理参数之一，能够反映人体的心血管状况，是临床上疾病诊断，手术指标考量和生命体征监控的重要依据。血压测量装置分为有创式和无创式两大类，有创式测量装置采用动脉插管法，所得血压最准确，但其技术要求高，且由于会对人体造成损伤，所以仅适用于临床上危重患者等。无创测量血压装置又分为间歇式测量血压装置和连续式测量血压装置。间歇式测量装置，包括柯氏听音法测量血压装置、示波法测量血压装置、恒定容积法测量血压装置，需要用到袖带，无法应用于长时间测量，且只能测得某一特定时刻的血压值，不能满足临床的需要。在此背景下，应用脉搏波进行血压测量的装置被提出。
一种利用脉搏波传播速度建立与血压相关的数据关系的数据处理方法中，脉搏波速的测定至少需要两点采集，设备复杂，不便随身携带。利用单点脉搏波形的特征参数建立关于血压的回归方程，只需要在手腕上或手指上单点采集脉搏波，就可以进行测量。目前，利用脉搏波特征参数的血压测量装置基本是根据采集的数据建立单一的一个或多个回归方程，由于回归方法具有随机性，因此，难以保证建立的回归方程在拟合数据集上能达到全局最优。同时，实际测量的数据具有一定的随机性，包括采集数据的随机性及回归方程建立的随机性。这种随机性会对血压测量装置的实用性产生不利的影响。
发明内容
为了解决因采集数据过程和建立回归方程过程的随机性、噪声干扰而导致的血压测量装置处理数据不准确的技术问题，本发明提出了一种血压测量装置以及用于该装置的数据处理方法。
本发明的技术问题通过以下的技术方案予以解决：一种用于血压测量装置的 数据处理方法，包括建立多个血压估算回归方程的步骤和计算血压值的步骤，所述建立多个血压估算回归方程的步骤包括：S11、获取多段脉搏波及对应的血压值集合；S12、预处理获取的脉搏波；S13、从预处理过的脉搏波中提取脉搏波特征点，计算脉搏波全局特征参数值；S14、在脉搏波全局特征参数值和对应的血压值集合中通过随机选取的方式建立多组回归测试集，所述回归测试集由测试集和训练集构成；S15、获取每组回归测试集中全局最优的回归方程；S16、评估并筛选出准确度高的回归方程；S17、对准确度高的回归方程赋予相应的权值；所述计算血压值的步骤包括：S21、采集脉搏波；S22、预处理采集的脉搏波；S23、从预处理过的脉搏波中提取脉搏波特征点，计算脉搏波全局特征参数值；S24、将获取的脉搏波全局特征参数代入步骤S16中所述的回归方程中，对计算结果去除异常值得到多个血压估计值；S25、对得到的血压估计值根据步骤S17中所述对应的权值计算出其加权系数；S26、对得到的血压估计值加权平均，得到血压测量值。本发明还提供了上述任一所述的用于血压测量装置的数据处理方法在血压测量装置或生理多参数监测设备中的应用。
本发明与现有技术对比的有益效果包括：一种用于血压测量装置的数据处理方法，该方法包括建立多个血压估算回归方程的步骤和计算血压值的步骤。建立多个血压估算回归方程的步骤包括获取脉搏波及对应的血压值集合，预处理获取的脉搏波，提取脉搏波特征点，计算脉搏波全局特征参数值，利用获取的脉搏波全局特征参数值和对应的血压值集合通过随机选取的方式建立多组回归测试集，获取每组回归测试集中全局最优的回归方程，评估并筛选出准确度高的回归方程，赋予相应的权值。计算血压值的步骤包括将获取的脉搏波全局特征参数代入选出的回归方程中，对计算的结果剔除异常值得到多个血压值，最后加权取平均，提高了血压测量方法的准确性和稳定性。
附图说明
图1是本发明提供的标有特征点的周期脉搏波的典型波形图。
图2是本发明提供的建立多个血压估算回归方程的流程图。
图3是本发明提供的计算血压值的流程图。
图4是本发明提供的血压测量装置的结构框图。
具体实施方式
下面对照附图并结合优选的实施方式对本发明作进一步说明。
图1为标有特征点的周期脉搏波的典型波形图，B表示主动脉瓣开放点，C表示收缩期最高压力点，F表示重搏波起点，G表示重搏波最高压力点。
本发明还提供了一种用于上述血压处理装置的数据处理方法，包括建立多个血压估算回归方程的步骤和计算血压值的步骤，如图2所示，建立多个血压估算回归方程的步骤包括：
101、获取多段脉搏波及对应的血压值集合；
102、预处理获取的脉搏波；
103、从预处理过的脉搏波中提取脉搏波特征点，利用特征点计算脉搏波全局特征参数值；
104、在获取的脉搏波全局特征参数值和对应的血压值集合中，采用随机选取的方式建立多组回归测试集，所述回归测试集由测试集和训练集构成；
105、获取每组回归测试集中全局最优的回归方程；
106、评估并筛选出准确度高的回归方程；
107、对选出的回归方程赋予相应的权值；
如图3所示，所述计算血压值的步骤包括：
201、采集脉搏波；
202、预处理采集的脉搏波；
203、从预处理过的脉搏波中提取脉搏波特征点，利用特征点计算脉搏波全局特征参数值；
204、将获取的脉搏波全局特征参数代入步骤106中所述的回归方程中，对计算结果去除异常值得到多个血压估计值；
205、对得到的血压估计值根据步骤107中所述的权值计算出其加权系数；
206、对得到的血压估计值加权平均，得到血压测量值。
需要说明的是，在步骤101中获取的对应的血压值是为了建立多个回归方程。
在本具体实施方式中，在步骤105中所述全局最优的回归方程是采用逐步回 归分析法获取；所述逐步回归分析法是以全局遍历的方式对所述训练集进行分析。
具体地，预处理脉搏波包括去除基线漂移和低通滤波，所述训练集中的训练数据不少于8组。
进一步地，基线漂移的去除可采用中值滤波器进行处理，公式如下：
Figure PCTCN2017074439-appb-000001
Figure PCTCN2017074439-appb-000002
其中，P为原始脉搏波数据，B为原始脉搏波的基线漂移，Pa为去除基线漂移之后的脉搏波数据。len为P的数据长度。m为常数，设脉搏波一个周期的数据点数为num，则m的最佳取值为
Figure PCTCN2017074439-appb-000003
(
Figure PCTCN2017074439-appb-000004
为向下取整符号)。
低通滤波器主要滤除高频噪声以及工频干扰等，可采用切比雪夫Ⅰ型滤波器，设定相应的参数为：通带边界频率10Hz左右，阻带边界频率12Hz左右，通带最大衰减0.1dB，阻带最小衰减20dB。Pa经过滤波处理后的脉搏波数据为Pb。
在某些优选的实施例中，在步骤103或步骤203中所述的脉搏波特征点包括主动脉瓣开放点、收缩期最高压力点、重搏波起点、重搏波最高压力点；所述提取脉搏波特征点包括对脉搏波的平滑处理，所述平滑处理采用三点线平滑处理的方式。
具体地，三线平滑处理方式的公式为：
Figure PCTCN2017074439-appb-000005
其中，Pc表示搏波数据点，Pb表示经过滤波处理后的脉搏波数据。
在本具体实施方式中，多组回归测试集的建立是通过对得到的n对脉搏波全局特征参数和对应的血压值作为样本数据的1～n组，将各组数据按1～n进行编号， 通过生成随机数的方式生成1～n范围内的随机数，之后选择相应的随机数对应的数据，重复进行，直到选出k组数据。将选出的k组数据作为训练集，其余的n-k组数据作为测试集，这样的训练集和测试集构成一个回归测试集。
重复上述步骤，得到h组不同的回归测试集。优选地，8＜n，8＜k＜n，5＜h。
在本具体实施方式中，所述主动脉瓣开放点、收缩期最高压力点的提取步骤包括：
T1、获取脉搏波的所有广义极值点，并得到极值点集；具体地，从头到尾遍历脉搏波各数据点，根据极值点判断条件，判断该点是否为极值点；
T2、确定主动脉瓣开放点B、收缩期最高压力点C差值的阈值；具体地，将上一步中得到的极值点集按幅值大小进行升序排序得到排序后的极值点集extS，分别从排序后的极值点集的两端(即最小数据部分和最大数据部分)连续取thn个极值点，判断这thn个点的离散程度，若离散程序小于可接受标准程度da，则将这thn个点作为本段脉搏波数据的B点或C点的代表点集。否则，按从两端到中间的方向依次取下一组thn个点，继续判断，直到找到B点和C点的代表点集。B点的代表点集为从最小数据部分找到的满足要求的点，C点的代表点集为从最大数据部分找到的满足要求的点。更具体地，thn取2～5，da取0.003～0.01。
T3、根据阈值判断并提取出所述脉瓣开放点、收缩期最高压力点；具体地，将B点代表点取均值得到该段脉搏波的全局B点代表点幅值dB，将C点代表点取均值得到全局C点代表点幅值dC。则B、C差值的阈值公式为：
thd＝ζ*(dC-dB)，其中阈值系数ζ取值为0.5～0.7。
具体地，在步骤T1中所述极值点判断条件为：(Pc[i]-Pc[i-1])*(Pc[i+1]-Pc[i])≤0，Pc表示搏波数据点。
在步骤T3中所述根据阈值判断并提取出所述脉瓣开放点、收缩期最高压力点的判别式为ext[i+1]-ext[i]＞thd,1≤i≤len-1，ext[i]表示第i个极值点，ext[i+1]表示第i+1个极值点，thd表示主动脉瓣开放点、收缩期最高压力点差值的阈值，len表示极值点的个数。
在本具体实施方式中，所述重搏波起点F、重搏波最高压力点G的提取步骤包括：
E1、分割脉搏波的周期区间；具体地，由上面提取到的B点为边界，来分割脉搏波的各段区间。
E2、确定重搏波起点、重搏波最高压力点所在的脉搏波周期区间；具体地，设各周期中，C、F点之间长度在周期中的比例rcf，C、G点之间长度在周期中的比例rcg。通过rcf、rcg，以及提取出的B、C点确定各周期中F、G所在的区间interFG。优选地，rcf取0.08～0.12，rcg取0.45～0.55。
E3、通过计算所述重搏波起点、重搏波最高压力点所在的脉搏波周期区间内各点的平均斜率角变化指标值的集合提取出所述的重搏波起点、重搏波最高压力点。具体地，计算公式如下：
angChg[i]＝atan((p1-p2)/ts)-atan((p2-p3)/ts)；
p1＝interFG[i+2]+inter[i+1]+interFG[i]；
p2＝interFG[i]+inter[i-1]+interFG[i-2]；
p3＝interFG[i-2]+inter[i-3]+interFG[i-4]；
其中，ts为脉搏波数据采样周期，atan为取反正切值函数。取angChg中最大值对应的点为F点，取angChg中最小值对应的点为G点。
在本具体实施方式中，在步骤103或步骤203中获取脉搏波全局特征参数值包括对异常值的去除及对去除异常值的各特征参数集取平均；所述脉搏波全局特征参数值包括全局收缩期时间占比、全局主波高度、全局降中峡相对高度、全局重搏波相对高度、全局收缩期面积占比、全局主波上升斜率和全局K值；所述异常值的判别采用肖维勒方法进行。
具体地，所述搏波全局特征参数值是通过计算出该段脉搏波各周期的特征参数作为特征参数集，再从特征参数集中获取。
需要说明的是，单周期特征参数值的计算包括根据上述提取到的特征点，以特征点B为界对这段脉搏波进行周期分割，获取各个单周期的脉搏波数据periodData，再对单周期的脉搏波数据进行处理和特征参数的计算。
首先，单周期数据波形标准化：去除单周期数据的基线部分，通过线性方程 拟合出基线，再用原数据减去基线获得标准化之后的数据nomalData。公式如下：
romalData[i]＝periodData[i]-(kBase*i+bBase)；1≤i≤lenP.
Figure PCTCN2017074439-appb-000006
bBase＝periodData[1]-kBase
其中，lenP为periodData的数据长度，periodData为提取出的波形数据；nomalData为标准化后的波形数据；kBase是单周期波形的基线斜率；bBase是单周期波形的基线截距；i为波形的数据点的第i个。特别的，经过标准化之后，单周期内的B点幅值为0。
之后，单周期特征参数计算，包括：
i.收缩期时间占比sInT：
脉搏波从B点到F点的这一段为脉搏波收缩期，sInT就是BF段时长在整个周期的比值。由于采样率不变，所以时间与样本长度成正比，这里用BF段数据长度比上本周期数据长度的值作为sInT。
ii.主波高度mainH：
主波高度对应于C点幅值，这里取标准化之后的C点幅值作为mainH。
iii.降中峡相对高度gorgeInMainH：
降中峡高度对应于F点幅值，gorgeInMainH是指降中峡高度相对于主波的高度比。这里取标准化之后的F点幅值与mainH的比值作为gorgeInMainH。
iv.重搏波相对高度repeatInMianH：
重搏波高度对应于G点幅值，repeatInMianH是指重搏波高度相对于主波的高度比。这里取标准化之后的G点幅值与mainH的比值作为repeatInMianH。
v.收缩期面积占比sysInArea：
sysInArea是指脉搏波波形的收缩期面积占整个周期波形面积的比值。面积计算采用如下公式：
Figure PCTCN2017074439-appb-000007
当计算收缩期面积时，end对应于F点所在位置。计算整个周期波形面积时，end＝lenP。
vi.主波上升斜率slopeMain：
slopeMain是从B点到C点的平均斜率，计算公式如下：
slopeMain＝mairH/(lenBC*ts)
其中，lenBC为BC段脉搏波的采样点数。
vii.K值的计算公式如下：
K＝Speriod/((lenP-1)*mainH)
其中，Speriod为整个周期波形的面积。
在本具体实施方式中，在步骤105中所述全局最优的回归方程是以调整后的决定系数值作为回归方程拟合效果的判断标准，所述调整后的决定系数值的计算公式为adjR2＝1-rmse2/var(BPval)，其中RMSE为回归方程拟合的均方误差，var(BPval)为血压值样本的均方差，adjR2表示调整后的决定系数。
在本具体实施方式中，上述任一方法中所述的回归方程的表达式为
Figure PCTCN2017074439-appb-000008
其中param为最终选入参数的集合，coef为各参数对应系数的集合，cont为常数项，lenParam为选入参数的个数，BPest为估计血压值。
在本具体实施方式中，利用上述得到的回归方程和与之对应的回归测试集中的测试集，进行该回归方程的准确度评估。评估方法如下：
将测试集中的各个脉搏波全局特征参数数据代入上式得到的回归方程中，得到相应的估计血压值BPest，将BPest与测试集对应的血压值BPval进行比较，得到各个数据的估计误差│BPest-BPval│，并将误差值按三段要求进行划分，≤5，≤10，≤15，计算在各段的测试数据比例。并根据评估标准将回归方程准确等级划分为A、B、C、D四个等级，评估标准如下：
Figure PCTCN2017074439-appb-000009
Figure PCTCN2017074439-appb-000010
将准确度等级较低的回归方程去除，并对剩下的回归方程按准确度等级赋予不同的权值，等级越高，权值越高。
在本具体实施方式中，将准确等级为B级以下的回归方程予以去除，设定等级A的回归方程对应的权值(w)为a，等级为B的回归方程对应的权值为b，优选地，取a＝5，b＝2。
在本具体实施方式中，在步骤205中的加权系数的计算公式为：
Figure PCTCN2017074439-appb-000011
其中，lenW为去除异常值后的血压估计值的个数，wR[i]为第i个血压估计值对应的加权系数。
在本具体实施方式中，在步骤206中血压测量值的计算公式为：
Figure PCTCN2017074439-appb-000012
其中，BPest为去除异常值后的血压估计值集合，wR[i]为第i个血压估计值对应的加权系数，BP为血压测量值。
本发明还提供了上述任一所述的用于血压测量装置的数据处理方法在血压测量装置或生理多参数监测设备中的应用。
所述用于血压测量装置的数据处理方法不局限于在血压测量装置或生理多参数监测设备中的应用，该方法适用于其他具有血压数据处理功能的设备，所述设备包括手机、手环、指环、脚环、电脑、平板等。
所述血压测量装置，如图4所示，包括多个回归方程建立模块和血压估算模块，所述血压估算回归方程建立模块包括：
模板数据采集模块，用于获取多段脉搏波及对应的血压值；
数据预处理模块，用于预处理从模板数据采集模块中获取的脉搏波；
模板特征参数计算模块，用于从预处理模块预处理过的脉搏波中提取脉搏波特征点，利用特征点计算脉搏波全局特征参数值；
回归测试集建立模块，对从模板特征参数计算模块中获取的脉搏波全局特征参数值和从模板数据采集模块中获取的对应的血压值组成的样本集合应用随机选取的方式建立多组回归测试集，所述回归测试集由测试集和训练集构成；
初步回归方程建立模块，用于获取各组回归测试集中全局最优的回归方程；
回归方程评估与选择模块，用于评估并筛选出回归方程建立模块中的准确度高的回归方程，对所述准确度高的回归方程按准确度等级赋予相应的权值；
所述血压估算模块包括：
脉搏波采集模块，用于采集脉搏波；
数据预处理模块，用于预处理从脉搏波采集模块中采集的脉搏波；
特征参数计算模块，用于从预处理模块中预处理过的脉搏波中提取脉搏波特征点，利用特征点计算脉搏波全局特征参数值；
初步血压值估算模块，用于将特征参数计算模块中获取的脉搏波全局特征参数代入多组回归方程建立模块中的多组回归方程进行计算，去除计算结果中的异常值，得到多个血压估计值；
加权系数计算模块，利用初步血压值估算模块中得出的多个血压值所对应的回归方程的权重计算出对应的加权系数；
血压估算输出模块，利用加权系数计算模块中得到的加权系数对初步血压值估算模块中得到的多个血压估计值进行加权平均，得到血压测量值。
需要说明的是，在模板数据采集模块中脉搏波所对应的血压值的获取方式可以是利用袖带式汞柱血压计或电子血压计进行采集。
具体地，所述生理多参数监测设备通过采集脉搏波信号处理血压数据，获得血压测量值。
具体地，所述生理多参数监测设备为贴片式生理多参数监测设备，包括适于贴附到人体皮肤上的扁平状柔性材料的壳体，所述壳体正面可以通过胶贴贴至人体皮肤上，所述壳体的正面和/或背面设置有脉搏波传感器，通过液态硅胶注射成型方式或固态硅胶模压成型方式与壳体成型为一体，所述壳体内密封装有用于处理所采集生理参数数据的电路模块，所述电路模块与所述脉搏波传感器相连；所述脉搏波传感器进行脉搏波信号的采集，并将采集的数据传输给壳体内的电路模块相连。采集的脉搏波信号可以计算包括血压、血氧饱和度、心率、呼吸频率、最大摄氧量中至少一种生理参数，脉搏波传感器优选用光电式脉搏波传感器。
以上内容是结合具体的优选实施方式对本发明所作的进一步详细说明，不能认定本发明的具体实施只局限于这些说明。对于本发明所属技术领域的技术人员来说，在不脱离本发明构思的前提下，还可以做出若干等同替代或明显变型，而 且性能或用途相同，都应当视为属于本发明的保护范围。

Claims (10)

  1. 一种用于血压测量装置的数据处理方法，其特征在于：包括建立多个血压估算回归方程的步骤和计算血压值的步骤，所述建立多个血压估算回归方程的步骤包括：
    S11、获取多段脉搏波及对应的血压值集合；
    S12、预处理获取的脉搏波；
    S13、从预处理过的脉搏波中提取脉搏波特征点，计算脉搏波全局特征参数值；
    S14、在脉搏波全局特征参数值和对应的血压值集合中通过随机选取的方式建立多组回归测试集，所述回归测试集由测试集和训练集构成；
    S15、获取每组回归测试集中全局最优的回归方程；
    S16、评估并筛选出准确度高的回归方程；
    S17、对准确度高的回归方程赋予相应的权值；
    所述计算血压值的步骤包括：
    S21、采集脉搏波；
    S22、预处理采集的脉搏波；
    S23、从预处理过的脉搏波中提取脉搏波特征点，计算脉搏波全局特征参数值；
    S24、将获取的脉搏波全局特征参数代入步骤S16中所述的回归方程中，对计算结果去除异常值得到多个血压估计值；
    S25、对得到的的血压估计值根据步骤S17中所述的对应的权值计算出其加权系数；
    S26、对得到的血压估计值加权平均，得到血压测量值。
  2. 如权利要求1所述的用于血压测量装置的数据处理方法，其特征在于：在步骤S15中所述全局最优的回归方程是采用逐步回归分析法获取；所述逐步回归分析法是以全局遍历的方式对所述训练集进行分析。
  3. 如权利要求1所述的用于血压测量装置的数据处理方法，其特征在于：在步骤S13或步骤S23中所述的脉搏波特征点包括主动脉瓣开放点、收缩期最高压力点、重搏波起点、重搏波最高压力点；所述提取脉搏波特征点包括对脉搏波的平滑处理，所述平滑处理采用三点线平滑处理的方式。
  4. 如权利要求3所述的用于血压测量装置的数据处理方法，其特征在于： 所述主动脉瓣开放点、收缩期最高压力点的提取步骤包括：
    T1、获取脉搏波的所有广义极值点，并得到极值点集；
    T2、确定主动脉瓣开放点、收缩期最高压力点差值的阈值；
    T3、根据阈值判断并提取出所述脉瓣开放点、收缩期最高压力点。
  5. 如权利要求4所述的用于血压测量装置的数据处理方法，其特征在于：在步骤T1中所述极值点判断条件为：(Pc[i]-Pc[i-1])*(Pc[i+1]-Pc[i]≤0，Pc表示搏波数据点；在步骤T3中所述根据阈值判断并提取出所述脉瓣开放点、收缩期最高压力点的判别式为ext[i+1]-ext[i]＞thd，1≤i≤len-1，ext[i]表示第i个极值点，ext[i+1]表示第i+1个极值点，thd表示主动脉瓣开放点、收缩期最高压力点差值的阈值，len表示极值点的个数。
  6. 如权利要求3所述的用于血压测量装置的数据处理方法，其特征在于：所述重搏波起点、重搏波最高压力点的提取步骤包括：
    E1、分割脉搏波的周期区间；
    E2、确定重搏波起点、重搏波最高压力点所在的脉搏波周期区间；
    E3、通过计算所述重搏波起点、重搏波最高压力点所在的脉搏波周期区间内各点的平均斜率角变化指标值的集合提取出所述的重搏波起点、重搏波最高压力点。
  7. 如权利要求1所述的用于血压测量装置的数据处理方法，其特征在于：在步骤S13或步骤S23中获取脉搏波全局特征参数值包括对异常值的去除及对去除异常值的各特征参数集取平均；所述脉搏波全局特征参数值包括全局收缩期时间占比、全局主波高度、全局降中峡相对高度、全局重搏波相对高度、全局收缩期面积占比、全局主波上升斜率和全局K值；所述异常值的判别采用肖维勒方法进行。
  8. 如权利要求1所述的用于血压测量装置的数据处理方法，其特征在于：在步骤S15中所述全局最优的回归方程是以调整后的决定系数值作为回归方程拟合效果的判断标准，所述调整后的决定系数值的计算公式为adjR2＝1-rmse2/var(BPval)，其中RMSE为回归方程拟合的均方误差，var(BPval)为血压值样本的均方差，adjR2表示调整后的决定系数。
  9. 如权利要求1-8任一所述的用于血压测量装置的数据处理方法，其特征在于：所述的回归方程的表达式为
    Figure PCTCN2017074439-appb-100001
    其中param为最终选入参数的集合，coef为各参数对应系数的集合，cont为常数项，lenParam为选入参数的个数，BPest为估计血压值。
  10. 如权利要求1-9任一所述的用于血压测量装置的数据处理方法在血压测量装置或生理多参数监测设备中的应用。
PCT/CN2017/074439 2017-02-22 2017-02-22 一种用于血压测量装置的数据处理方法 WO2018152713A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201780002087.8A CN107995981B (zh) 2017-02-22 2017-02-22 一种用于血压测量装置的数据处理方法
PCT/CN2017/074439 WO2018152713A1 (zh) 2017-02-22 2017-02-22 一种用于血压测量装置的数据处理方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2017/074439 WO2018152713A1 (zh) 2017-02-22 2017-02-22 一种用于血压测量装置的数据处理方法

Publications (1)

Publication Number Publication Date
WO2018152713A1 true WO2018152713A1 (zh) 2018-08-30

Family

ID=62039958

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/074439 WO2018152713A1 (zh) 2017-02-22 2017-02-22 一种用于血压测量装置的数据处理方法

Country Status (2)

Country Link
CN (1) CN107995981B (zh)
WO (1) WO2018152713A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113261924A (zh) * 2021-04-15 2021-08-17 北京雪扬科技有限公司 一种脑卒中智能预警系统及方法
CN116712049A (zh) * 2023-08-09 2023-09-08 临沂大学 一种运动数据采集处理方法及系统

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109350025A (zh) * 2018-10-23 2019-02-19 苏州鱼跃医疗科技有限公司 一种动态血压分析方法
CN110179449A (zh) * 2019-06-18 2019-08-30 广东乐心医疗电子股份有限公司 血压计以及血压测量方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102178518A (zh) * 2011-05-31 2011-09-14 北京新兴阳升科技有限公司 用于用脉搏波连续测量估算动脉血压的个体化校正方法及装置
US20130296723A1 (en) * 2012-05-03 2013-11-07 Samsung Electronics Co., Ltd. Portable blood pressure measuring apparatus and blood pressure measuring method in portable terminal
CN103976721A (zh) * 2014-04-22 2014-08-13 辛勤 血压测量方法以及用于实现该方法的嵌入式装置
CN104116503A (zh) * 2014-07-16 2014-10-29 华中科技大学 一种无创连续血压的测量方法及装置
CN106413534A (zh) * 2015-08-08 2017-02-15 深圳先进技术研究院 连续血压测量装置、测量模型建立方法和系统

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007014684A (ja) * 2005-07-11 2007-01-25 Motoharu Hasegawa 動脈硬化度評価装置および動脈硬化指数算出プログラム
US20110270051A1 (en) * 2010-05-03 2011-11-03 Morteza Naghavi Cardiovascular Disease Screening Method and Apparatus
CN104688193A (zh) * 2015-04-08 2015-06-10 常晓年 一种心血管参数的检测方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102178518A (zh) * 2011-05-31 2011-09-14 北京新兴阳升科技有限公司 用于用脉搏波连续测量估算动脉血压的个体化校正方法及装置
US20130296723A1 (en) * 2012-05-03 2013-11-07 Samsung Electronics Co., Ltd. Portable blood pressure measuring apparatus and blood pressure measuring method in portable terminal
CN103976721A (zh) * 2014-04-22 2014-08-13 辛勤 血压测量方法以及用于实现该方法的嵌入式装置
CN104116503A (zh) * 2014-07-16 2014-10-29 华中科技大学 一种无创连续血压的测量方法及装置
CN106413534A (zh) * 2015-08-08 2017-02-15 深圳先进技术研究院 连续血压测量装置、测量模型建立方法和系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
MIAO CHANGYUN ET AL.: "Research on the Method of Blood Pressure Monitoring Based on Multiple Parameters of pulse wave", JOURNAL OF BIOMDEDICAL ENGINEERING, vol. 5, no. 32, 31 October 2015 (2015-10-31), pages 1113 - 1117 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113261924A (zh) * 2021-04-15 2021-08-17 北京雪扬科技有限公司 一种脑卒中智能预警系统及方法
CN116712049A (zh) * 2023-08-09 2023-09-08 临沂大学 一种运动数据采集处理方法及系统
CN116712049B (zh) * 2023-08-09 2023-10-20 临沂大学 一种运动数据采集处理方法及系统

Also Published As

Publication number Publication date
CN107995981B (zh) 2021-07-23
CN107995981A (zh) 2018-05-04

Similar Documents

Publication Publication Date Title
US20190295729A1 (en) Universal non-invasive blood glucose estimation method based on time series analysis
CN108185996B (zh) 动脉血管年龄估算模型构建方法和装置
Hou et al. A real-time QRS detection method based on phase portraits and box-scoring calculation
US11006885B2 (en) Apparatus for determining blood pressure
CN108498089A (zh) 一种基于深度神经网络的无创连续血压测量方法
US10537254B2 (en) Blood pressure calculation method based on pulse return wave transmission time, and blood pressure monitor
EP3474749B1 (en) Dynamic calibration of a blood pressure measurement device
CN112274126A (zh) 一种基于多路脉搏波的无创连续血压检测方法、装置
CN105662345B (zh) 心跳信号处理方法、装置和系统
JP5132003B2 (ja) 血圧推定装置及び血圧推定方法
Motin et al. PPG derived heart rate estimation during intensive physical exercise
WO2018152713A1 (zh) 一种用于血压测量装置的数据处理方法
CN108354597A (zh) 一种基于最优波提取的快速血压计算方法
CN105105734A (zh) 一种基于心音信号的无创连续血压测量方法、装置及系统
KR101696791B1 (ko) 흉부임피던스를 이용한 폐기능 모니터링 장치 및 방법
CN112274120A (zh) 一种基于单路脉搏波的无创动脉硬化检测方法、装置
CN111000544B (zh) 基于ppg波形的混合式连续血压测量模型构建方法及系统
CN210095711U (zh) 一种无创连续血压测量设备
CN114145725B (zh) 一种基于无创连续血压测量的ppg采样率估算方法
KR20180067348A (ko) Ppg를 이용한 맥박수 추정방법 및 장치
Zhang et al. Non-invasive blood glucose detection using NIR based on GA and SVR
CN110265150A (zh) 血压计算模型建立方法及穿戴式监测装置与血压测量方法
Khan et al. A highly integrated computing platform for continuous, non-invasive bp estimation
Chen et al. A developed algorithm for oscillometric blood pressure measurement
CN117292821B (zh) 一种基于知识图谱的辅助诊疗系统

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: 17898289

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 1205A DATED 24.01.2020)

122 Ep: pct application non-entry in european phase

Ref document number: 17898289

Country of ref document: EP

Kind code of ref document: A1