CN104127194A - Depression evaluating system and method based on heart rate variability analytical method - Google Patents

Depression evaluating system and method based on heart rate variability analytical method Download PDF

Info

Publication number
CN104127194A
CN104127194A CN201410334899.0A CN201410334899A CN104127194A CN 104127194 A CN104127194 A CN 104127194A CN 201410334899 A CN201410334899 A CN 201410334899A CN 104127194 A CN104127194 A CN 104127194A
Authority
CN
China
Prior art keywords
data
interval
module
heart rate
parameter
Prior art date
Application number
CN201410334899.0A
Other languages
Chinese (zh)
Other versions
CN104127194B (en
Inventor
杨荣骞
吕瑞雪
Original Assignee
华南理工大学
深圳市是源医学科技有限公司
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 华南理工大学, 深圳市是源医学科技有限公司 filed Critical 华南理工大学
Priority to CN201410334899.0A priority Critical patent/CN104127194B/en
Publication of CN104127194A publication Critical patent/CN104127194A/en
Application granted granted Critical
Publication of CN104127194B publication Critical patent/CN104127194B/en

Links

Abstract

The invention discloses a depression evaluating system and method based on a heart rate variability analytical method, particularly a depression evaluating system and method in a perinatal period. The depression evaluating system comprises a data storage module used for storing electrocardio and pulse wave signals, a peak detecting module used for acquiring a peak point sequence of recording data, a data correcting module used for acquiring sinus beat NN interval sequence, a heart rate variability curve acquiring module used for acquiring a heart rate variability curve, an HRV analyzing module conducting time domain analysis, frequency domain analysis and nonlinear analysis, a feature parameter selecting module used for selecting a feature parameter from HRV parameters, a modeling module used for obtaining a perinatal period depression classification model, and a model applying module used for inputting the data of a testee into the classification module to obtain a depression degree. By means of the depression evaluating system and method based on the heart rate variability analytical method, degree quantitative evaluation of perinatal period depression is achieved, a scientific research method of the depression based on the technical field of physiological information examination is enriched, a test is simple and practicable, medical resources can be effectively saved, and good clinical practicality is achieved.

Description

A kind of evaluating system of the depression based on heart rate variance analyzing method and method
Technical field
The present invention relates to a kind of computer-aided diagnosis technology, particularly a kind of evaluating system of the depression based on heart rate variance analyzing method and method
Background technology
Gestation and childbirth are women's great experience in life, aspect physiology, will experience each vitals of health and gonadal hormone and a series of great change of associated hormone pregnancy period to puerperal; At psychological aspects, adapt to the variation of the relations such as work, family and colleague, be concerned about neonatal growing up healthy and sound, consider the arrangement of economic aspect etc.; The variation of these complexity and the stimulation of other undesirable elements just can be brought out perinatal stage depression.Pregnant and lying-in women's depressive emotion is usually in trimester of pregnancy, and especially 1 year late trimester of pregnancy to puerperal, the depressive emotion therefore this particular time being occurred calls perinatal stage depression.According to the research at CDC mother and child care center, show, in China, term depression prevalence up to 10-20%.Since 2003, the annual Population Birth number of China basicly stable with 1,600 ten thousand left and right, wherein, within 2010, Population Birth number is 1,588 ten thousand, within 2011, Population Birth number is 1,604 ten thousand, and Population Birth number in 2012 is 1,635 ten thousand (Data Sources: the 6th census in the whole nation).China's perinatal stage depression is a problem that is worth attaching great importance to.Main Basis medical history, mental symptom inspection are determined in the diagnosis of perinatal stage depression, and consider in conjunction with the rule of course advancement.At present clinical common depression assessment scale has: anxiety, self rating depressive scale, Edinburgh postnatal depression scale and self-defined scale.Domestic conventional postpartum depression self evaluation scale is Edinburgh postnatal depression scale (Edinburgh Postnatal Depression Scale, EDPS).According to the injury of international disease and Clasification standard the tenth edition, China's patients with postpartum depression great majority are slight depression, and minority is moderate depressive patients, rarely found severe depression patient.The survey showed that after 6 weeks in childbirth to take Edinburg depression scale: postpartum depression incidence rate is 22%, and wherein patients with mild is 90.6%, and moderate is 9.4%.Different from other medical domains, almost place one's entire reliance upon patient's medical history of mental sickness states to assess its order of severity whether ill and disease.This assessment mode is subject to many-sided impact, and as patient's mental status, ability to express, the situation that even exists oneself to cover up, prior art exists following shortcoming with not enough:
1, scale test is low with psychological inquiring efficiency, can not meet current huge and ever-increasingly treat examination crowd.
2, scale appraisal result can not accurately reflect tester's psychologic status, may have the subjective situation that conceals the state of an illness.
3, the scoring of single rating scale can not be used for making a definite diagnosis depression.
4, the diagnostic result of shrink can be subject to the impact of self subjective factors and actual clinical experience.
Summary of the invention
Primary and foremost purpose of the present invention is that the shortcoming that overcomes prior art, with not enough, provides a kind of evaluating system of the depression based on heart rate variance analyzing method, and this evaluating system is a kind of perinatal stage depression evaluating system.
The shortcoming that another object of the present invention is to overcome prior art, with not enough, provides a kind of appraisal procedure of evaluating system of the depression based on heart rate variance analyzing method, and this appraisal procedure can realize fast, objective diagnosis perinatal stage depression grade.
Primary and foremost purpose of the present invention is achieved through the following technical solutions: a kind of evaluating system of the depression based on heart rate variance analyzing method, comprise: data are preserved module, peak detection block, Data correction module, obtain heart rate variability curve module, HRV analysis module, selects characteristic parameter module, model building module, model application module.
Data are preserved module for recording test process pregnant and lying-in women's electrocardiosignal and pulse wave signal, and electrocardiosignal is preserved into electrocardiosignal file and pulse wave signal file with pulse wave signal with binary data form.Described test process comprises: quiescent condition, deep breathing state, Valsalva maneuver and standing activities one of four states, record electrocardio and pulse wave data under different conditions, each state recording duration 2 minutes.Described quiescent condition refers to that tester sits quietly 2 minutes, keeps eupnea; Described deep breathing state be take 5 second aspiratory action exhale action in 5 second be a breathing cycle, repeat 12 times; Described Valsalva maneuver is with deeply air-breathing, holds one's breath for 15 seconds, and then forced expiration loosens 15 seconds is an action cycle, repeats four times.Described standing state refers to that tester becomes erect-position from seat, keeps standing state 2 minutes, and eupnea, avoids body action.
Peak detection block is for obtaining the peak point sequence of electrocardiosignal and pulse wave signal, and described peak point refers to the peak of each heart beat cycle electrocardiosignal and pulse wave signal waveform, and the interval of adjacent peak point is a heart beat cycle.
Data correction module is for proofreading and correct the data of peak value point sequence, removes non-hole cardiac signal.Described non-hole cardiac signal comprises: the short RR interval producing due to ectopic beat and follow Long RR interval thereafter; When peak value detects, due to the undetected Long RR interval causing (being actually the RR interval sum of two above heartbeats); While detecting due to artifact or peak value, detect the too low ghost peak producing of threshold value, this will make a normal hole IBI be divided into two NN intervals, thus the rub-out signal of generation.The process that Data correction module is carried out Data correction is: while there is continuously 3 hole IBIs, a hole IBI in the middle of choosing is as the NN data of interval.
Obtain the electrocardiosignal of heart rate variability curve module after for calculation correction and the hole heartbeat NN interval series of pulse wave signal, and the NN interval series data Replica obtaining is become to duration is the NN interval series of 10 minutes, the duration of each NN interval of take in sequence is the longitudinal axis, take the time that NN interval occurs is transverse axis, obtains heart rate variability linearity curve.
HRV analysis module, for calculating the parameter of heart rate variability linearity curve, comprises time domain parameter, frequency domain parameter and nonlinear parameter.
Select characteristic parameter module for calculating the characteristic parameter of all HRV parameters under 4 kinds of test modes, according to the characteristic parameter collection of best prioritization criteria and CfsSubsetEval attribute appraisal procedure acquisition heart rate variability analysis.1, selecting a heart rate variability parameter is initial attribute collection R1, the scoring of calculating current property set; 2, the heart rate variability analysis parameter that increases property set, obtains second property set R 2, the scoring of calculating second property set; 3, repeating step 2, obtain i (i is positive integer) property set R i; 4, compare property set R iwith R i ?1score, if the inferior R of N continuous (N is positive integer) iscoring is higher than R i ?1, get back to step 3; Otherwise calculate, finish, property set Ri is characteristic parameter collection.
Model building module carries out category of model training for use characteristic parameter, based on AdaBoost algorithm, obtains can be used in the perinatal stage depression disaggregated model of diagnosis perinatal stage depression grade.Initialization sample space is that the distribution probability of n sample is q by N the even distribution and constitution of sample n0, through training for the first time, obtain Weak Classifier h 1, the accuracy rate of classification results is α 1, then revise sample space, the weight of the correct sample that reduces to classify, the weight of the sample of extend assortment mistake, obtains second sample space X1, through training for the second time, obtains Weak Classifier h 2, the accuracy rate of classification results is α 2, through t repetition training, obtain t Weak Classifier h i, the accuracy rate of classification results is α τ, described h i={ h i| i=1,2,3 ..., t}, t is positive integer, the expression formula of described Degree of Depression quantitative appraisement model is: wherein, h trepresent t Weak Classifier, α tthe weight that represents t Weak Classifier; p iweight for each grader of calculating according to the accuracy rate of grader.
Model application module, for tester's data input perinatal stage depression disaggregated model being detected, obtains current tester's depression grade.
Data are preserved module and according to data acquisition time, are divided four sections of preservations by the electrocardiosignal of collection and pulse wave signal, the corresponding test mode of every segment data, and the electrocardiosignal of preserving and pulse wave signal are inputted as the data of peak detection block; Peak detection block is obtained the peak point sequence of electrocardiosignal and pulse wave signal, and the input using it as Data correction module; Data correction module is proofreaied and correct peak point sequence, removes non-hole cardiac signal and obtains hole heartbeat NN interval series, as the input of heart rate variability linearity curve; Obtaining heart rate variability curve module, hole heartbeat NN interval series is copied into duration is the NN interval series of 10 minutes, the duration of each NN interval of take in sequence is the longitudinal axis, take the time that NN interval occurs is transverse axis, obtain heart rate variability linearity curve, input as HRV analysis module, HRV analysis module carries out time-domain analysis, frequency-domain analysis and nonlinear analysis to HRV curve, obtains HRV parameter, as the date processing basis of selecting characteristic parameter module; Select characteristic parameter module to assess the HRV parameter under four test modes, in all HRV parameters from one of four states, select characteristic parameter, the data basis of setting up as model; Model building module use characteristic parameter is carried out category of model training, based on AdaBoost algorithm, obtains can be used in the perinatal stage depression disaggregated model of diagnosis perinatal stage depression grade; Model application module is tester's data input perinatal stage depression disaggregated model to be detected, obtains current tester's depression grade.
The evaluating system of the perinatal stage depression based on heart rate variance analyzing method, also comprises with lower module: building database module, model correcting module.
Building database module is for management to all data, can complete according to keywords retrieve data, by particular community sorting data with by particular community garbled data.
Model correcting module is used for proofreading and correct the test result of current data, and result is saved in data base, the abundant sample size of setting up model use.
HRV analysis module, comprises time domain parameter computing unit, nonlinear parameter computing unit, time-frequency converting unit and frequency domain parameter computing unit.
Time domain parameter computing unit for according to the NN interval series of HRV curve, calculates following five time domain parameter: MEAN, SDNN, RMSSD, pNN50 and SDSD, and described MEAN is all hole NN average of interval; Described SDNN is all hole NN standard deviation of interval, and described RMSSD is the root-mean-square of adjacent NN interval difference, and described pNN50 is the ratios of the above adjacent NN interval difference of 50 ms intervals, and described SDSD is the standard deviation of adjacent NN between interval.
The data object of nonlinear parameter computing unit is HRV curve, for calculating following nonlinear parameter: SD1, SD2, α 1and α 2; Described SD2 is the scatterplot of the HRV curve the longest distance between two points in scatterplot region in X=Y direction, and described SD1 is the distance perpendicular to the longest point-to-point transmission in scatterplot region in X=Y direction, described α 1for HRV curve removes the slope of trend fluction analysis first (1st~11 points) fitting a straight line, described α 2for HRV curve removes the slope of trend fluction analysis second portion (the 12nd point is to last point) fitting a straight line.
Time-frequency converting unit is for resampling to HRV curve, then section length and overlapped data length are set, every segment data is carried out to windowing process, described window function is Hamming window, calculate the power spectrum of every segment data, the power spectrum of each segment data is integrated into the power spectrum of all HRV curves.
Frequency domain parameter computing unit is used for calculating frequency domain parameter, and described frequency domain parameter comprises: VLF, LF, HF, TP, pVLF, pLF, pHF, nLF, nHF and LF/HF.Described VLF is the power of extremely low frequency composition 0.0033~0.04Hz, and described LF is the power of low-frequency component 0.04~0.15Hz; Described HF is the power of radio-frequency component 0.15~0.4Hz; Described TP is general power, described pVLF is the percentage ratio of heart rate variability linearity curve extremely low frequency composition, described pLF is the percentage ratio of heart rate variability linearity curve low-frequency component, described pHF is the percentage ratio of heart rate variability linearity curve radio-frequency component, described nLF is normalized low frequency power, described nHF is normalized radio-frequency component, and described LF/HF is the ratio of low-frequency component and radio-frequency component.
Another object of the present invention is achieved through the following technical solutions: the appraisal procedure of the perinatal stage depression based on heart rate variance analyzing method, comprises the following steps:
Step 1, data are preserved.Record electrocardiosignal and the pulse wave signal of pregnant and lying-in women in test process, and electrocardiosignal is preserved into electrocardiosignal file and pulse wave signal file with pulse wave signal with binary data form.Described test process comprises: quiescent condition, deep breathing state, Valsalva maneuver and standing activities one of four states, record electrocardio and pulse wave data under different conditions, each state recording duration 2 minutes.Described quiescent condition refers to that tester sits quietly 2 minutes, keeps eupnea; Described deep breathing state be take 5 second aspiratory action exhale action in 5 second be a breathing cycle, repeat 12 times; Described Valsalva maneuver is with deeply air-breathing, holds one's breath for 15 seconds, and then forced expiration loosens 15 seconds is an action cycle, repeats four times.Described standing state refers to that tester becomes erect-position from seat, keeps standing state 2 minutes, and eupnea, avoids body action.
Step 2, peak value detect.Obtain the peak point sequence of electrocardiosignal and pulse wave signal, described peak point refers to the peak of each heart beat cycle electrocardiosignal and pulse wave signal waveform, and the interval of adjacent peak point is a heart beat cycle.
Step 3, Data correction.Proofread and correct the data of peak value point sequence, remove non-hole cardiac signal.Described non-hole cardiac signal comprises: the short RR interval producing due to ectopic beat and follow Long RR interval thereafter; When peak value detects, due to the undetected Long RR interval causing (being actually the RR interval sum of two above heartbeats); While detecting due to artifact or peak value, detect the too low ghost peak producing of threshold value, this will make a normal hole IBI be divided into two NN intervals, thus the rub-out signal of generation.The process that Data correction module is carried out Data correction is: while there is continuously 3 hole IBIs, a hole IBI in the middle of choosing is as the NN data of interval.
Step 4, obtain heart rate variability linearity curve.The hole heartbeat NN interval series of the electrocardiosignal after calculation correction and pulse wave signal, and the NN interval series data Replica obtaining is become to duration is the NN interval series of 10 minutes, the duration of each NN interval of take in sequence is the longitudinal axis, take the time that NN interval occurs is transverse axis, obtains heart rate variability linearity curve.
Step 5, HRV analyze.The parameter of calculating heart rate variability linearity curve, comprises time domain parameter, frequency domain parameter and nonlinear parameter.
Step 6, selection characteristic parameter.Calculate the characteristic parameter of all HRV parameters under 4 kinds of test modes, according to the characteristic parameter collection of best prioritization criteria and CfsSubsetEval attribute appraisal procedure acquisition heart rate variability analysis.1, selecting a heart rate variability parameter is initial attribute collection R1, the scoring of calculating current property set; 2, the heart rate variability analysis parameter that increases property set, obtains second property set R 2, the scoring of calculating second property set; 3, repeating step 2, obtain i (i is positive integer) property set R i; 4, compare property set R iwith R i ?1score, if the inferior R of N continuous (N is positive integer) iscoring is higher than R i ?1, get back to step 3; Otherwise calculate, finish, property set Ri is characteristic parameter collection.
Step 7, model are set up.Use characteristic parameter is carried out category of model training, based on AdaBoost algorithm, obtains can be used in the perinatal stage depression disaggregated model of diagnosis perinatal stage depression grade.Initialization sample space is that the distribution probability of n sample is q by N the even distribution and constitution of sample n0, through training for the first time, obtain Weak Classifier h 1, the accuracy rate of classification results is α 1, then revise sample space, the weight of the correct sample that reduces to classify, the weight of the sample of extend assortment mistake, obtains second sample space X1, through training for the second time, obtains Weak Classifier h 2, the accuracy rate of classification results is α 2, through t repetition training, obtain t Weak Classifier h i, the accuracy rate of classification results is α τ, described hi={hi|i=1,2,3 ..., t}, t is positive integer, the expression formula of described Degree of Depression quantitative appraisement model is: wherein, h trepresent t Weak Classifier, α tthe weight that represents t Weak Classifier; p iweight for each grader of calculating according to the accuracy rate of grader.
Step 8, model application.To tester's data input perinatal stage depression disaggregated model be detected, obtain current tester's depression grade.
Step 9, building database.Data management, can complete according to keywords retrieve data, by particular community sorting data with by particular community garbled data.The conveniently management to all data in system of data base, practicality and the operability of increase system.
Step 10, model correction.Proofread and correct the test result of current data, and result is saved in data base.Model tuning can increase the training sample of perinatal stage depression model, can enrich and set up the sample size that model is used, and increases the feature of different brackets data in model, is conducive to improve the accuracy rate of category of model.
Calculate heart rate variability analysis parameter, comprise the following steps:
Step 51, time domain parameter calculate.According to the NN interval series of HRV curve, calculate following five time domain parameter: MEAN, SDNN, RMSSD, pNN50 and SDSD, described MEAN is all hole NN average of interval; Described SDNN is all hole NN standard deviation of interval, and described RMSSD is the root-mean-square of adjacent NN interval difference, and described pNN50 is the ratios of the above adjacent NN interval difference of 50 ms intervals, and described SDSD is the standard deviation of adjacent NN between interval.
Step 52, nonlinear parameter calculate.Calculate following nonlinear parameter: SD1, SD2, α 1and α 2; Described SD2 is the scatterplot of the HRV curve the longest distance between two points in scatterplot region in X=Y direction, and described SD1 is the distance perpendicular to the longest point-to-point transmission in scatterplot region in X=Y direction, described α 1for HRV curve removes the slope of trend fluction analysis first (1st~11 points) fitting a straight line, described α 2for HRV curve removes the slope of trend fluction analysis second portion (the 12nd point is to last point) fitting a straight line.
Step 53, time-frequency conversion.HRV curve is resampled, then section length and overlapped data length are set, every segment data is carried out to windowing process, described window function is Hamming window, calculate the power spectrum of every segment data, the power spectrum of each segment data is integrated into the power spectrum of all HRV curves.
Step 54, frequency domain parameter calculate.Calculate frequency domain parameter, described frequency domain parameter comprises: VLF, LF, HF, TP, pVLF, pLF, pHF, nLF, nHF and LF/HF.Described VLF is the power of extremely low frequency composition 0.0033~0.04Hz, and described LF is the power of low-frequency component 0.04~0.15Hz; Described HF is the power of radio-frequency component 0.15~0.4Hz; Described TP is general power, described pVLF is the percentage ratio of heart rate variability linearity curve extremely low frequency composition, described pLF is the percentage ratio of heart rate variability linearity curve low-frequency component, described pHF is the percentage ratio of heart rate variability linearity curve radio-frequency component, described nLF is normalized low frequency power, described nHF is normalized radio-frequency component, and described LF/HF is the ratio of low-frequency component and radio-frequency component.
The present invention has following advantage and effect with respect to prior art:
1, realize perinatal stage Degree of Depression quantitative evaluation, avoided pregnant and lying-in women to use subjectivity and the polytropy of scale test.
2, enrich depression and based on physiologic information, checked the scientific research methods of technical field.
3, only need to obtain the electrocardiograph pulse wave datum of testee, test simple and easy to doly, facilitate pregnant and lying-in women to test examination, can effectively save medical resource, can have good Clinical practicability; Can realize quick, objective diagnosis perinatal stage depression grade.
Accompanying drawing explanation
Fig. 1 is system module figure.
Fig. 2 is peak value overhaul flow chart.
Fig. 3 is time-frequency flow path switch figure.
Fig. 4 characteristic parameter and model Establishing process figure.
The specific embodiment
Below in conjunction with embodiment and accompanying drawing, the present invention is described in further detail, but embodiments of the present invention are not limited to this.
Embodiment
The present embodiment is the electrocardio under different conditions and pulse wave data with 92 routine pregnant and lying-in women, as the training sample of perinatal stage depression disaggregated model.Different conditions refers to quiescent condition, deep breathing state, Valsalva maneuver and standing activities one of four states, each state recording duration 2 minutes.Described different conditions refers to quiescent condition, deep breathing state, Valsalva maneuver and standing activities one of four states, records electrocardio and pulse wave data under different conditions, each state recording duration 2 minutes.Quiescent condition refers to that tester sits quietly 2 minutes, keeps eupnea; Deep breathing state be take 5 second aspiratory action exhale action in 5 second be a breathing cycle, repeat 12 times; Valsalva maneuver is with deeply air-breathing, holds one's breath for 15 seconds, and then forced expiration loosens 15 seconds is an action cycle, repeats four times.Standing state refers to that tester becomes erect-position from seat, keeps standing state 2 minutes, and eupnea, avoids body action.
A kind of evaluating system of the depression based on heart rate variance analyzing method as shown in Figure 1, comprise data preservation module, peak detection block, Data correction module, obtain heart rate variability curve module, HRV analysis module, selects characteristic parameter module, model building module, model application module; Also comprise building database module, model correcting module.The concrete implementation step of appraisal procedure of the depression based on heart rate variance analyzing method of this system is as follows:
Step 1, data are preserved.Record electrocardiosignal and the pulse wave signal of pregnant and lying-in women in test process, and electrocardiosignal is preserved into electrocardiosignal file and pulse wave signal file with pulse wave signal with binary data form.
Step 2, peak value detect.Extract the electrocardio under different conditions, the peak point of pulse wave signal, the flow chart that peak value detects as shown in Figure 2.The single order differential F ' that calculates electrocardiosignal or pulse wave signal (p), second-order differential F " (p); gather n2 the zero point of calculating second-order differential; near (n) the set n1 of >K of F ' in set of computations n2; K is slope threshold value; electrocardiosignal K=100; pulse wave signal K=20, searches for F ' (n1+i) n1, makes F ' (n1+i)=0 invocation point be peak point.
Step 3, Data correction.Electrocardio under different conditions, pulse wave signal are carried out to Data correction, remove non-hole cardiac signal.Due to Data Detection or human body self physiological reason, can cause in electrocardio and pulse waveform and may be mixed with expense hole cardiac signal, the electrocardiograph pulse wave train of analyzing for HRV, must be by the heart beat cycle sequence due to hole heartbeat, therefore to carry out Data correction to image data, remove non-hole heartbeat.The non-hole cardiac signal of generally, sneaking in NN interval data has: the short RR interval that ectopic beat produces and follow Long RR interval thereafter; When QRS detects, due to the undetected Long RR interval causing (being actually the RR interval sum of two above heartbeats); While detecting due to artifact or QRS, detect the too low false QRS producing of threshold value, this will make the RR interval of a normal hole heartbeat be divided into two short RR intervals, thus the rub-out signal producing.Described Data correction comprises: while 1, there is continuously 3 hole heartbeats, the RR interval of a heartbeat in the middle of choosing is as the NN data of interval; 2: the NN interval data of choosing by above-mentioned rule occur continuously, and while reaching certain quantity, as a data segment that can be used for carrying out HRV analysis, otherwise this segment data is given up.According to above-mentioned correction rule, obtain NN interval series.
Step 4, obtains heart rate variability linearity curve, by NN interval series long data when total that is copied into 10 minutes, obtains 10 minutes heart rate variability linearity curves.Original data sequence is 2 minute datas of detection record under different conditions, the total duration of NN interval series after peak value detection and Data correction is less than or equals 2 minutes, on this basis NN interval data sequence is copied into the data sequence of 10 minutes durations, NN interval series with 10 minutes obtains heart rate variability linearity curve, the basis of processing as follow-up data.
Step 5, HRV analyzes.Calculate heart rate variability analysis parameter.Heart rate variability analysis parameter comprises time domain parameter, frequency domain parameter, nonlinear parameter.
Time domain parameter comprises: MEAN, SDNN, RMSSD, pNN50 and SDSD, and described MEAN is all hole NN average of interval; SDNN is all hole NN standard deviation of interval, n is the number of NN interval in 10 minutes HRV curves, and i is the sequence number of order; Described RMSSD is the root-mean-square of adjacent NN interval difference, Δ NNi is the poor of two adjacent NN intervals, and described pNN50 is the ratios of the above adjacent NN interval difference of 50 ms intervals, and SDSD is the standard deviation of adjacent NN between interval, mean Δ NN is the meansigma methods of all adjacent NN interval differences;
NN interval data sequence is resampled and time-frequency conversion, its flow chart as shown in Figure 3: 1, the sample frequency that time-frequency changes being set is Fs=N/600, and wherein N is the number of NN interval series in 10 minutes HRV curves; 2, to the HRV curve segmentation of 10 minutes, segment length L=512, overlap length nov=400, segments n is the positive integer that is not more than N/L, every segment data is designated as X i; 3, to data segment X iwindowing process, obtains Xwi=X i* W, take Hamming window as example, window function expression formula W=0.54 ?0.46*cos (2*pi*x), x={i/ (m ?1) | i is positive integer, i<L/2}; 4, to each data segment rated output spectrum and each section of spectrum component, must power spectrum be all sections of power spectrum and.Frequency domain parameter comprises: VLF, LF, HF, TP, pVLF, pLF, pHF, nLF, nHF and LF/HF.VLF is the power of extremely low frequency composition 0.0033~0.04Hz, and LF is the power of low-frequency component 0.04~0.15Hz; HF is the power of radio-frequency component 0.15~0.4Hz; TP is general power, pVLF is the percentage ratio of heart rate variability linearity curve extremely low frequency composition, pLF is the percentage ratio of heart rate variability linearity curve low-frequency component, pHF is the percentage ratio of heart rate variability linearity curve radio-frequency component, nLF is normalized low frequency power, nLF=LF/ (TP ?VLF) * 100%, and nHF is normalized radio-frequency component, nHF=HF/ (TP ?VLF) * 100%, LF/HF is the ratio of low-frequency component and radio-frequency component;
Described nonlinear parameter comprises SD1, SD2, α 1and α 2; Described SD2 is the scatterplot region the longest distance between two points of scatterplot in X=Y direction, and SD1 is the distance perpendicular to the longest point-to-point transmission in scatterplot region in X=Y direction, described α 1for HRV curve removes the slope of trend fluction analysis first (1st~11 points) fitting a straight line, described α 2for HRV curve removes the slope of trend fluction analysis second portion (the 12nd point is to last point) fitting a straight line.
Step 6, selection characteristic parameter.According to the characteristic parameter collection of the greedy search expansion that can recall and CfsSubsetEval attribute appraisal procedure acquisition heart rate variability analysis.1, selecting a heart rate variability parameter is initial attribute collection R1, the scoring of calculating current property set; 2, the heart rate variability analysis parameter that increases property set, obtains second property set R 2, the scoring of calculating second property set; 3, repeating step 2, obtain i (i is positive integer) property set R i; 4, compare property set R iwith R i-1score, if the inferior R of N continuous (N is positive integer) iscoring is higher than R i-1, get back to step 3; Otherwise calculate, finish, property set Ri is characteristic parameter collection.92 sample sets of take are experimental example, and the characteristic parameter collection getting in said process is, tranquillization state parameter: pNN50, LF, HF, TP, α 1; Deep breathing state data: RMSSD, pNN50, VLF, LF, α 1; Wa Er Salva state parameter: pNN50, TP, pVLF, nLF, nHF; The state of standing parameter: Mean.
Step 7, model are set up.Initialization sample space is that the distribution probability of n sample is q by 92 even distribution and constitutions of sample n0, through training for the first time, obtain Weak Classifier h 1, the accuracy rate of classification results is α 1, then revise sample space, the weight of the correct sample that reduces to classify: q r1=A rq r0, r is the sample number of correct classification, the weight that Ar is r sample, Ar<1; The weight of the sample of extend assortment mistake: q w1=A wq w0, the sample number that w is misclassification, the weight that Aw is w sample, Aw>1; And obtain second sample space X1, through training for the second time, obtain Weak Classifier h 2, the accuracy rate of classification results is α 2, through t repetition training, obtain t Weak Classifier h i, the accuracy rate of classification results is α τ, described hi={hi|i=1,2,3 ..., t}, t is positive integer, the expression formula of described Degree of Depression quantitative appraisement model is:
H = &Sigma; i = 1 t h i p i , p i = &alpha; i / ( &alpha; 1 + &alpha; 2 + &CenterDot; &CenterDot; &CenterDot; + &alpha; t ) ,
Wherein, h trepresent t Weak Classifier, α tthe weight that represents t Weak Classifier; p iweight for each grader of calculating according to the accuracy rate of grader.
Select the detailed process of characteristic parameter and model foundation as shown in Figure 4.
Step 8, model application.Data input model, obtains diagnostic result.Data are inputted to grader H, obtain the classification results of data.
Step 9, building database.Data management, can complete according to keywords retrieve data, by particular community sorting data with by particular community garbled data.The conveniently management to all data in system of data base, practicality and the operability of increase system.
Step 10, model correction.Proofread and correct the test result of current data, and result is saved in data base.Model tuning can increase the training sample of perinatal stage depression model, can enrich and set up the sample size that model is used, and increases the feature of different brackets data in model, is conducive to improve the accuracy rate of category of model.
Implementation process has designed three groups of contrast tests, first group adopts 54 pregnant and lying-in women's sample data as the training sample set of perinatal stage depression model, its accuracy rate of the model obtaining is 63.2%, second group adopts 71 pregnant and lying-in women's sample data as the training sample set of perinatal stage depression model, its accuracy rate of the model obtaining is 78.87%, the 3rd group adopts 92 pregnant and lying-in women's sample data as the training sample set of perinatal stage depression model, and its accuracy rate of the model obtaining is 82.5%.As can be seen here, if continue to increase model training collection in follow-up test, the accuracy rate of model can continue to improve.
The evaluating system of a kind of depression based on heart rate variance analyzing method disclosed by the invention and method, a kind of new research method is proposed in perinatal stage depression diagnosis and detection field, the method energy science, objective assessment pregnant and lying-in women's depressive state grade, can effectively assist and clinical pregnant and lying-in women be carried out to the examination of perinatal stage depression, prevention, discovery perinatal stage depression are had to very large meaning, can effectively reduce the clinical onset rate of perinatal stage depression, reduce the injury that depression causes Perinatal Women and family thereof, there is generalization and Clinical practicability.
Above-described embodiment is preferably embodiment of the present invention; but embodiments of the present invention are not restricted to the described embodiments; other any do not deviate from change, the modification done under spirit of the present invention and principle, substitutes, combination, simplify; all should be equivalent substitute mode, within being included in protection scope of the present invention.

Claims (6)

1. the evaluating system of the depression based on heart rate variance analyzing method, it is characterized in that, comprise data preservation module, peak detection block, Data correction module, obtains heart rate variability curve module, HRV analysis module, select characteristic parameter module, model building module, model application module;
Data are preserved module for recording test process pregnant and lying-in women's electrocardiosignal and pulse wave signal, and preserve into file with binary data form; Described test process comprises: quiescent condition, deep breathing state, Valsalva maneuver and standing activities one of four states, each state recording duration 2 minutes; Described quiescent condition refers to that tester sits quietly 2 minutes, keeps eupnea; Described deep breathing state be take 5 second aspiratory action exhale action in 5 second be a breathing cycle, repeat 12 times; Described Valsalva maneuver is with deeply air-breathing, holds one's breath for 15 seconds, and then forced expiration loosens 15 seconds is an action cycle, repeats four times; Described standing state refers to that tester becomes erect-position from seat, keeps standing state 2 minutes, and eupnea, avoids body action;
Peak detection block is for obtaining the peak point sequence of electrocardiosignal and pulse wave signal, and described peak point refers to the peak of each heart beat cycle electrocardiosignal and pulse wave signal waveform, and the interval of adjacent peak point is a heart beat cycle;
The data of Data correction module for proofreading and correct peak value point sequence, remove non-hole cardiac signal; Described non-hole cardiac signal comprises:
(1) the short RR interval producing due to ectopic beat and follow Long RR interval thereafter;
(2) when peak value detects, due to the undetected Long RR interval causing;
(3) during due to artifact or peak value detection, detect the too low ghost peak producing of threshold value; The process that Data correction module is carried out Data correction is: while there is continuously 3 hole IBIs, a hole IBI in the middle of choosing is as the NN data of interval;
Obtain heart rate variability curve module for calculating hole heartbeat NN interval series, and the NN interval series obtaining is copied into duration is the NN interval series of 10 minutes, the duration of each NN interval of take in sequence is the longitudinal axis, and take the time that NN interval occurs is transverse axis, obtains heart rate variability linearity curve;
HRV analysis module, for calculating the parameter of heart rate variability linearity curve, comprises time domain parameter, frequency domain parameter and nonlinear parameter;
Select characteristic parameter module for calculating the characteristic parameter of all HRV parameters under 4 kinds of test modes, described characteristic parameter refers to the parameter of selecting through attribute appraisal procedure and best prioritization criteria iterative algorithm;
Model building module carries out category of model training for use characteristic parameter, based on AdaBoost algorithm, obtains can be used in the perinatal stage depression disaggregated model of diagnosis perinatal stage depression grade;
Model application module, for tester's data input perinatal stage depression disaggregated model being detected, obtains current tester's depression grade;
Data are preserved module and according to data acquisition time, are divided four sections of preservations by the electrocardiosignal of collection and pulse wave signal, and the corresponding test mode of every segment data, as the data input of peak detection block; Peak detection block is obtained peak point sequence, and the input using it as Data correction module; Data correction module is removed non-hole cardiac signal in peak point sequence and is obtained hole heartbeat NN interval series, as the input of heart rate variability linearity curve; Obtaining heart rate variability curve module, hole heartbeat NN interval series is copied into duration is the NN interval series of 10 minutes, and the duration of each NN interval of take in sequence is the longitudinal axis, and take the time that NN interval occurs is transverse axis, obtains heart rate variability linearity curve; HRV analysis module carries out time-domain analysis, frequency-domain analysis and nonlinear analysis to HRV curve, obtains HRV parameter; Select characteristic parameter module to assess the HRV parameter under four test modes, select characteristic parameter; Model building module use characteristic parameter is carried out category of model training, based on AdaBoost algorithm, obtains can be used in the perinatal stage depression disaggregated model of diagnosis perinatal stage depression grade; Model application module is tester's data input perinatal stage depression disaggregated model to be detected, obtains current tester's depression grade.
2. the evaluating system of the depression based on heart rate variance analyzing method according to claim 1, is characterized in that, also comprises with lower module: building database module, model correcting module;
Building database module is for management to all data, can complete according to keywords retrieve data, by particular community sorting data with by particular community garbled data;
Model correcting module is used for proofreading and correct the test result of current data, and result is saved in data base, the abundant sample size of setting up model use.
3. the evaluating system of the depression based on heart rate variance analyzing method according to claim 1, it is characterized in that, described HRV analysis module, comprises time domain parameter computing unit, nonlinear parameter computing unit, time-frequency converting unit and frequency domain parameter computing unit;
Time domain parameter computing unit for according to the NN interval series of HRV curve, calculates following five time domain parameter: MEAN, SDNN, RMSSD, pNN50 and SDSD, and described MEAN is all hole NN average of interval; Described SDNN is all hole NN standard deviation of interval, and described RMSSD is the root-mean-square of adjacent NN interval difference, and described pNN50 is the ratios of the above adjacent NN interval difference of 50 ms intervals, and described SDSD is the standard deviation of adjacent NN between interval;
The data object of nonlinear parameter computing unit is HRV curve, for calculating following nonlinear parameter: SD1, SD2, α 1and α 2; Described SD2 is the scatterplot of the HRV curve the longest distance between two points in scatterplot region in X=Y direction, and described SD1 is the distance perpendicular to the longest point-to-point transmission in scatterplot region in X=Y direction, described α 1for HRV curve removes the slope of trend fluction analysis first fitting a straight line, described α 2for HRV curve removes the slope of trend fluction analysis second portion fitting a straight line; , described first refers to 1st~11 points, described second portion refers to that the 12nd point is to last point;
Time-frequency converting unit is for resampling to HRV curve, then section length and overlapped data length are set, every segment data is carried out to windowing process, described window function is Hamming window, calculate the power spectrum of every segment data, the power spectrum of each segment data is integrated into the power spectrum of all HRV curves;
Frequency domain parameter computing unit is used for calculating frequency domain parameter, and described frequency domain parameter comprises: VLF, LF, HF, TP, pVLF, pLF, pHF, nLF, nHF and LF/HF; Described VLF is the power of extremely low frequency composition 0.0033~0.04Hz, and described LF is the power of low-frequency component 0.04~0.15Hz; Described HF is the power of radio-frequency component 0.15~0.4Hz; Described TP is general power, described pVLF is the percentage ratio of heart rate variability linearity curve extremely low frequency composition, described pLF is the percentage ratio of heart rate variability linearity curve low-frequency component, described pHF is the percentage ratio of heart rate variability linearity curve radio-frequency component, described nLF is normalized low frequency power, described nHF is normalized radio-frequency component, and described LF/HF is the ratio of low-frequency component and radio-frequency component.
4. an appraisal procedure for the evaluating system of the depression based on heart rate variance analyzing method described in claim 1, is characterized in that, comprises the following steps:
Step 1, data are preserved; Record electrocardiosignal and the pulse wave signal of pregnant and lying-in women in test process, and preserve into file with binary data form; Described test process comprises: quiescent condition, deep breathing state, Valsalva maneuver and standing activities one of four states, each state recording duration 2 minutes; Described quiescent condition refers to that tester sits quietly 2 minutes, keeps eupnea; Described deep breathing state be take 5 second aspiratory action exhale action in 5 second be a breathing cycle, repeat 12 times; Described Valsalva maneuver is with deeply air-breathing, holds one's breath for 15 seconds, and then forced expiration loosens 15 seconds is an action cycle, repeats four times; Described standing state refers to that tester becomes erect-position from seat, keeps standing state 2 minutes, and eupnea, avoids body action;
Step 2, peak value detect; Obtain the peak point sequence of electrocardiosignal and pulse wave signal, described peak point refers to the peak of each heart beat cycle electrocardiosignal and pulse wave signal waveform, and the interval of adjacent peak point is a heart beat cycle;
Step 3, Data correction; Proofread and correct the data of peak value point sequence, remove non-hole cardiac signal; Described non-hole cardiac signal comprises:
(1) the short RR interval producing due to ectopic beat and follow Long RR interval thereafter;
(2) when peak value detects, due to the undetected Long RR interval causing;
(3) during due to artifact or peak value detection, detect the too low ghost peak producing of threshold value; The process that Data correction module is carried out Data correction is: while there is continuously 3 hole IBIs, a hole IBI in the middle of choosing is as the NN data of interval;
Step 4, obtain heart rate variability linearity curve; The hole heartbeat NN interval series of the electrocardiosignal after calculation correction and pulse wave signal, and the NN interval series data Replica obtaining is become to duration is the NN interval series of 10 minutes, the duration of each NN interval of take in sequence is the longitudinal axis, take the time that NN interval occurs is transverse axis, obtains heart rate variability linearity curve;
Step 5, HRV analyze; The parameter of calculating heart rate variability linearity curve, the parameter of described heart rate variability linearity curve comprises time domain parameter, frequency domain parameter and nonlinear parameter;
Step 6, selection characteristic parameter; Select the characteristic parameter of all HRV parameters under 4 kinds of test modes, described characteristic parameter refers to the parameter of selecting through attribute appraisal procedure and best prioritization criteria iterative algorithm;
Step 7, model are set up; Use characteristic parameter is carried out category of model training, based on AdaBoost algorithm, obtains can be used in the perinatal stage depression disaggregated model of diagnosis perinatal stage depression grade;
Step 8, model application; To tester's data input perinatal stage depression disaggregated model be detected, obtain current tester's depression grade.
5. appraisal procedure according to claim 4, is characterized in that, further comprising the steps of:
Steps A, building database; Management to all data, can complete according to keywords retrieve data, by particular community sorting data with by particular community garbled data;
Step B, model correction; Proofread and correct the test result of current data, and result is saved in data base, the abundant sample size of setting up model use.
6. an appraisal procedure claimed in claim 4, is characterized in that, in step 5, the method for calculating the parameter of described heart rate variability linearity curve comprises the following steps:
Step 51, time domain parameter calculate; According to the NN interval series of HRV curve, calculate following five time domain parameter: MEAN, SDNN, RMSSD, pNN50 and SDSD, described MEAN is all hole NN average of interval, described SDNN is all hole NN standard deviation of interval, described RMSSD is the root-mean-square of adjacent NN interval difference, described pNN50 is the ratio of the above adjacent NN interval difference of 50 ms intervals, and described SDSD is the standard deviation of adjacent NN between interval;
Step 52, nonlinear parameter calculate; Calculate following nonlinear parameter: SD1, SD2, α 1and α 2; Described SD2 is the scatterplot of the HRV curve the longest distance between two points in scatterplot region in X=Y direction, and described SD1 is the distance perpendicular to the longest point-to-point transmission in scatterplot region in X=Y direction, described α 1for HRV curve removes the slope of trend fluction analysis first fitting a straight line, described α 2for HRV curve removes the slope of trend fluction analysis second portion fitting a straight line; Described first refers to 1st~11 points, and described second portion refers to that the 12nd point is to last point;
Step 53, time-frequency conversion; HRV curve is resampled, then section length and overlapped data length are set, every segment data is carried out to windowing process, described window function is Hamming window, calculate the power spectrum of every segment data, the power spectrum of each segment data is integrated into the power spectrum of all HRV curves;
Step 54, frequency domain parameter calculate; Calculate frequency domain parameter, described frequency domain parameter comprises: VLF, LF, HF, TP, pVLF, pLF, pHF, nLF, nHF and LF/HF; Described VLF is the power of extremely low frequency composition 0.0033~0.04Hz, and described LF is the power of low-frequency component 0.04~0.15Hz; Described HF is the power of radio-frequency component 0.15~0.4Hz; Described TP is general power, described pVLF is the percentage ratio of heart rate variability linearity curve extremely low frequency composition, described pLF is the percentage ratio of heart rate variability linearity curve low-frequency component, described pHF is the percentage ratio of heart rate variability linearity curve radio-frequency component, described nLF is normalized low frequency power, described nHF is normalized radio-frequency component, and described LF/HF is the ratio of low-frequency component and radio-frequency component.
CN201410334899.0A 2014-07-14 2014-07-14 A kind of evaluating system of the depression based on heart rate variance analyzing method CN104127194B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410334899.0A CN104127194B (en) 2014-07-14 2014-07-14 A kind of evaluating system of the depression based on heart rate variance analyzing method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410334899.0A CN104127194B (en) 2014-07-14 2014-07-14 A kind of evaluating system of the depression based on heart rate variance analyzing method

Publications (2)

Publication Number Publication Date
CN104127194A true CN104127194A (en) 2014-11-05
CN104127194B CN104127194B (en) 2016-05-04

Family

ID=51800236

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410334899.0A CN104127194B (en) 2014-07-14 2014-07-14 A kind of evaluating system of the depression based on heart rate variance analyzing method

Country Status (1)

Country Link
CN (1) CN104127194B (en)

Cited By (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104523263A (en) * 2014-12-23 2015-04-22 华南理工大学 Mobile internet based pregnant and lying-in woman health surveillance system
CN104983434A (en) * 2015-06-17 2015-10-21 重庆邮电大学 Multi-parameter mental stress assessment method based on analytical hierarchy method and device
CN105022929A (en) * 2015-08-07 2015-11-04 北京环度智慧智能技术研究所有限公司 Cognition accuracy analysis method for personality trait value test
CN105147248A (en) * 2015-07-30 2015-12-16 华南理工大学 Physiological information-based depressive disorder evaluation system and evaluation method thereof
CN105938520A (en) * 2016-04-15 2016-09-14 深圳市德力凯医疗设备股份有限公司 Data processing method and system for foaming test check
CN105997019A (en) * 2016-05-09 2016-10-12 鲍崇智 A body sensor network-based multi-dimensional heartbeat information synchronous collection method and system
CN106037705A (en) * 2016-05-23 2016-10-26 韩剑 Data detection method and system based on wearable equipment
CN106073765A (en) * 2016-05-27 2016-11-09 成都信汇聚源科技有限公司 Die suddenly in a kind of miniature holter devices the implementation method of risk profile
CN106175742A (en) * 2016-07-19 2016-12-07 北京心量科技有限公司 A kind of heart sign acquisition methods and device
CN106326644A (en) * 2016-08-16 2017-01-11 沈阳东软熙康医疗系统有限公司 Method and device for calculating HRV (heart rate variability) parameters and fatigue indexes
CN106419937A (en) * 2016-09-12 2017-02-22 南京邮电大学 Mental stress analysis system based on heart sound HRV theory
CN107088061A (en) * 2017-05-04 2017-08-25 太原理工大学 A kind of HRV on-line analysis systems and its method based on Shewhart control figures
CN108366763A (en) * 2015-06-15 2018-08-03 Medibio有限公司 Method and system for assessing the state of mind
CN108577804A (en) * 2018-02-02 2018-09-28 西北工业大学 A kind of BCG signal analysis methods and system towards hypertensive patient's monitoring
CN108652613A (en) * 2017-03-30 2018-10-16 深圳市理邦精密仪器股份有限公司 The method and device that signal time-frequency figure generates
CN109077714A (en) * 2018-07-05 2018-12-25 广州视源电子科技股份有限公司 Signal recognition method, device, equipment and storage medium
CN109394197A (en) * 2018-11-27 2019-03-01 中山大学 A kind of heart rate variability measurement method, device and equipment based on time frequency analysis
WO2019100561A1 (en) * 2017-11-27 2019-05-31 乐普(北京)医疗器械股份有限公司 Artificial intelligence-based interference recognition method for electrocardiogram

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101337103A (en) * 2008-08-28 2009-01-07 北京阳光易德科技发展有限公司 Pressure and emotion adjustment method, system and a physiology signal acquisition device
CN101642369A (en) * 2008-08-04 2010-02-10 南京大学 Autonomic nervous function biological feedback method and system
WO2012156427A1 (en) * 2011-05-18 2012-11-22 Alain Gilles Muzet System and method for determining sleep and sleep stages of a person

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101642369A (en) * 2008-08-04 2010-02-10 南京大学 Autonomic nervous function biological feedback method and system
CN101337103A (en) * 2008-08-28 2009-01-07 北京阳光易德科技发展有限公司 Pressure and emotion adjustment method, system and a physiology signal acquisition device
WO2012156427A1 (en) * 2011-05-18 2012-11-22 Alain Gilles Muzet System and method for determining sleep and sleep stages of a person

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
CARMILLA M.M.LICHT ET.AL: "Association between anxiety disorders and Heart Rate Variability in the Netherlands Study of Depression and Anxiety(NESDA)", 《PSYCHOSOMATIC MEDICINE》 *
JOANNA LEWEK ET.AL: "Clinical and electrocardiographic covariates of deceleration capacity in patients with ST-segment elevation myocardial infarction", 《CARDIOLOGY JOURNAL》 *
别怀玺 等: "心率变异分析在精神疾病领域中的应用", 《医学综述》 *

Cited By (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104523263A (en) * 2014-12-23 2015-04-22 华南理工大学 Mobile internet based pregnant and lying-in woman health surveillance system
CN108366763A (en) * 2015-06-15 2018-08-03 Medibio有限公司 Method and system for assessing the state of mind
US10912508B2 (en) 2015-06-15 2021-02-09 Medibio Limited Method and system for assessing mental state
CN104983434A (en) * 2015-06-17 2015-10-21 重庆邮电大学 Multi-parameter mental stress assessment method based on analytical hierarchy method and device
CN104983434B (en) * 2015-06-17 2018-05-01 重庆邮电大学 Multi-parameter psychological pressure appraisal procedure and device based on analytic hierarchy process (AHP)
WO2017016086A1 (en) * 2015-07-30 2017-02-02 华南理工大学 Depression evaluating system and method based on physiological information
CN105147248A (en) * 2015-07-30 2015-12-16 华南理工大学 Physiological information-based depressive disorder evaluation system and evaluation method thereof
CN105022929A (en) * 2015-08-07 2015-11-04 北京环度智慧智能技术研究所有限公司 Cognition accuracy analysis method for personality trait value test
CN105938520A (en) * 2016-04-15 2016-09-14 深圳市德力凯医疗设备股份有限公司 Data processing method and system for foaming test check
CN105938520B (en) * 2016-04-15 2018-07-31 深圳市德力凯医疗设备股份有限公司 A kind of data processing method and system of frothing test inspection
CN105997019A (en) * 2016-05-09 2016-10-12 鲍崇智 A body sensor network-based multi-dimensional heartbeat information synchronous collection method and system
CN105997019B (en) * 2016-05-09 2019-02-19 鲍崇智 Multidimensional heartbeat message synchronous collection method and system based on body-sensing net
CN106037705A (en) * 2016-05-23 2016-10-26 韩剑 Data detection method and system based on wearable equipment
CN106073765A (en) * 2016-05-27 2016-11-09 成都信汇聚源科技有限公司 Die suddenly in a kind of miniature holter devices the implementation method of risk profile
CN106175742A (en) * 2016-07-19 2016-12-07 北京心量科技有限公司 A kind of heart sign acquisition methods and device
CN106326644A (en) * 2016-08-16 2017-01-11 沈阳东软熙康医疗系统有限公司 Method and device for calculating HRV (heart rate variability) parameters and fatigue indexes
CN106419937A (en) * 2016-09-12 2017-02-22 南京邮电大学 Mental stress analysis system based on heart sound HRV theory
CN108652613B (en) * 2017-03-30 2020-11-03 深圳市理邦精密仪器股份有限公司 Method and device for generating signal time-frequency diagram
CN108652613A (en) * 2017-03-30 2018-10-16 深圳市理邦精密仪器股份有限公司 The method and device that signal time-frequency figure generates
CN107088061A (en) * 2017-05-04 2017-08-25 太原理工大学 A kind of HRV on-line analysis systems and its method based on Shewhart control figures
WO2019100561A1 (en) * 2017-11-27 2019-05-31 乐普(北京)医疗器械股份有限公司 Artificial intelligence-based interference recognition method for electrocardiogram
CN108577804A (en) * 2018-02-02 2018-09-28 西北工业大学 A kind of BCG signal analysis methods and system towards hypertensive patient's monitoring
CN109077714A (en) * 2018-07-05 2018-12-25 广州视源电子科技股份有限公司 Signal recognition method, device, equipment and storage medium
CN109394197A (en) * 2018-11-27 2019-03-01 中山大学 A kind of heart rate variability measurement method, device and equipment based on time frequency analysis

Also Published As

Publication number Publication date
CN104127194B (en) 2016-05-04

Similar Documents

Publication Publication Date Title
Neilson Fetal electrocardiogram (ECG) for fetal monitoring during labour
Clifford et al. Non-invasive fetal ECG analysis
von Rosenberg et al. Resolving ambiguities in the LF/HF ratio: LF-HF scatter plots for the categorization of mental and physical stress from HRV
Xiao et al. Sleep stages classification based on heart rate variability and random forest
CN107951485B (en) Ambulatory ECG analysis method and apparatus based on artificial intelligence self study
CN107495962B (en) Sleep automatic staging method for single-lead electroencephalogram
CN104523266B (en) A kind of electrocardiosignal automatic classification method
Martis et al. Application of higher order statistics for atrial arrhythmia classification
Silva et al. Noninvasive fetal ECG: the PhysioNet/computing in cardiology challenge 2013
Valenza et al. Point-process nonlinear models with laguerre and volterra expansions: Instantaneous assessment of heartbeat dynamics
WO2017016086A1 (en) Depression evaluating system and method based on physiological information
CN106108889B (en) Electrocardiogram classification method based on deep learning algorithm
Acharya et al. Heart rate analysis in normal subjects of various age groups
Yang Multiscale recurrence quantification analysis of spatial cardiac vectorcardiogram signals
Cahill et al. Association and prediction of neonatal acidemia
Bagby et al. Are there subtypes of alexithymia?
RU2678364C2 (en) Method and system for estimating brain concussion
CN103083013B (en) Electrocardio signal QRS complex wave detection method based on morphology and wavelet transform
EP1179319B1 (en) Apparatus to detect acute cardiac syndromes in specified groups of patients using ECG
US6580944B1 (en) Method and apparatus for diagnosing sleep breathing disorders while a patient in awake
Inbar et al. On surface EMG spectral characterization and its application to diagnostic classification
David et al. An estimate of fetal autonomic state by time-frequency analysis of fetal heart rate variability
Karvounis et al. An automated methodology for fetal heart rate extraction from the abdominal electrocardiogram
Behar et al. A practical guide to non-invasive foetal electrocardiogram extraction and analysis
Karpagachelvi et al. ECG feature extraction techniques-a survey approach

Legal Events

Date Code Title Description
PB01 Publication
C06 Publication
SE01 Entry into force of request for substantive examination
C10 Entry into substantive examination
GR01 Patent grant
C14 Grant of patent or utility model