CN107049303A - It is a kind of that the adaptive Fetal ECG signal extraction system and method led with many chests is led based on many abdomens - Google Patents
It is a kind of that the adaptive Fetal ECG signal extraction system and method led with many chests is led based on many abdomens Download PDFInfo
- Publication number
- CN107049303A CN107049303A CN201710148736.7A CN201710148736A CN107049303A CN 107049303 A CN107049303 A CN 107049303A CN 201710148736 A CN201710148736 A CN 201710148736A CN 107049303 A CN107049303 A CN 107049303A
- Authority
- CN
- China
- Prior art keywords
- mrow
- linear
- many
- adaptive
- fetal ecg
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/344—Foetal cardiography
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/43—Detecting, measuring or recording for evaluating the reproductive systems
- A61B5/4306—Detecting, measuring or recording for evaluating the reproductive systems for evaluating the female reproductive systems, e.g. gynaecological evaluations
- A61B5/4343—Pregnancy and labour monitoring, e.g. for labour onset detection
- A61B5/4362—Assessing foetal parameters
-
- 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
- A61B5/7235—Details of waveform analysis
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
Abstract
It is a kind of that the adaptive Fetal ECG signal extraction system and method led with many chests is led based on many abdomens, belong to Fetal ECG signal detection technique field.Many abdomens, which lead electrocardiosignal and obtain the Src Chan of Adaptive Noise Canceller by linear colligator, always to be inputted；Many chests, which lead electrocardiosignal and obtain the reference channel of Adaptive Noise Canceller by Linear and nonlinear filters, always to be exported；The coefficient of linear colligator in Src Chan is adjusted using adaptive RLS algorithm；Adjust the weights of Linear and nonlinear filters in each reference channel respectively using ADAPTIVE LMS ALGORITHM；The Src Chan always input of Adaptive Noise Canceller is subtracted each other obtained error with reference channel always output to export, the estimation of Fetal ECG signal is used as.This method not only overcomes many abdomen pathway optimization problems, also solve that parent electrocardio signal is non-linear between chest area to abdomen area to cause the problem of Fetal ECG signal extraction precision is low, Fetal ECG signal can be extracted in real time and is diagnosed for clinical medicine.
Description
Technical field
The invention belongs to Fetal ECG signal detection technique field, and in particular to it is a kind of based on many abdomens lead with many chests lead from
Adapt to Fetal ECG signal extraction system and method.
Background technology
With the lifting of China's economic level, people progressively strengthen the understanding of prenatal and postnatal care, to the fetus in perinatal period
The birth level and quality of pregnant woman is lifted with pregnant woman's real-time monitoring, positive shadow will be produced to the further improvement of China human mortality quality
Ring.Meanwhile, the advanced Medical Instruments and medical technology reached its maturity will be that the pregnant woman of perinatal period and fetal monitoring provide technology and protected
Barrier, this development trend has attracted the extensive concern of people.Therefore, people must use effective fetal monitoring measure, in real time
Monitor development of fetus and the growth situation of the gestational period and term.
Heart of fetus first produces physiology electrical activity before bounce, and the physiology electrical activity of heart of fetus is through carcass tissue and mother
Body viscera tissue is conducted to the body surface of parent, using Fetal ECG signal detection instrument the Fetal ECG passed in maternal abdominal
Activation record is got off, and forms so-called FECG (Fetal Electrocardiogram, FECG).Its produce principle with
Adult electrocardiogram is similar, by detecting the physiology electrical stimulation signal of heart of fetus bioelectrical activity, and is traced on electrocardiograph paper
Go out the synthetic vector waveform of cardiac electrophysiology activity.FECG can distinguish the fine change of fetus heartbeat each time,
Than traditional doppler ultrasound fetal heart monitoring, it can more reflect cardiomotility overall picture.FECG is heart of fetus physiological activity
Objective indicator, it reflects growth and health status of the fetus in the pregnancy period, can help diagnosing gestational early and childbirth
Whether the fetus of phase has the symptoms such as congenital heart disease or Following Hypoxia in Uterus.
The extraction quality of current Fetal ECG signal is mainly by restriction following aspects：(1) belly lead electrocardio
Signal to noise ratio in signal between Fetal ECG signal component and chest electro-cardiologic signal components is very low；(2) it is mixed from belly lead
Close unknown non-linear of presence between mother's electrocardiosignal that electrocardiosignal and chest are led；(3) external electrode quantity and position will
The information of different Fetal ECG signal waveforms is brought, and then influences the reconstruct of Fetal ECG signal waveform.
Despite the presence of above mentioned problem, many Fetal ECG signal extraction skills based on Adaptive Noise Canceler existing at present
Art, they are inputted using belly lead (abdomen is led) electrocardiosignal as the Src Chan of Adaptive Noise Canceler, using chest
Lead (chest is led) electrocardiosignal in portion's is inputted as corresponding reference channel.1975, Widrow of Stanford University et al. was used
Adaptive Noise Canceller structure based on linear FIR (finite impulse response, finite impulse response (FIR)) wave filter
To extract Fetal ECG signal, employ single Src Chan and multiple reference channels (B.Widrow, J.R.Glover,
J.M.McCool.Adaptive Noise Cancelling:Principles and Applications[J]
.Proceedings of the IEEE,1975,63(12):1692–1716.).This method can quickly track the chest leads heart
Still included in electric signal and its change in belly electrocardiosignal composition, but the error output of the Adaptive Noise Canceller
Remaining chest electrocardiosignal composition, experiment indicates the linear FIR filter in the structure, exists in estimation parent electrocardio signal
Chest area is delivered to non-linear aspect, poor-performing between abdomen area.Later, Camps-Valls et al., which was proposed, is based on
The nonlinear adaptive adaptive noise canceller of FIR neutral nets, utilizes the structure of single Src Chan and single reference channel
(G.Camps-Valls,M.Martinez-Sober,E.Soria-Olivas,et al.FoetalEcg Recovery Using
Dynamic Neural Networks[J].Artificial Intelligence in Medicine,2004,31(3):
197–209.).This method can tackle above-mentioned nonlinear problem well, but the initialization of neutral net and learning algorithm need compared with
High time complexity.2008, a kind of Adaptive Noise Canceller based on Volterra nonlinear filters was suggested, and is adopted
With multiple Src Chans and single reference channel, for complete same Fetal ECG signal extraction task (M.Shadaydeh,
Y.Xiao,R.K.Ward.Extraction of Fetal Ecg Using Adaptive Volterra Filters[C]
.European Signal Processing Conference.2008:1–5.).What this method can approximately exist well
The problem of linear processes, but its performance do not improve significantly with the increase of kernel function exponent number, while bringing heavy
Computation burden.Behar et al. proposes non-based on ESN (echo state neural network, Echo State Networks)
Linear adaption adaptive noise canceller (J.Behar, A.Johnson, G.D Clifford, et al.A Comparison of
Single Channel Fetal Ecg Extraction Methods[J].Annals of Biomedical
Engineering,2014,42(6):Fetal ECG method for extracting signal 1340-1353.), using single Src Chan and single ginseng
Examine channel design.This method is limited by the initialization matter of weights, and seeks suitable neural network node number needs one
Individual prolonged search procedure.
Above method extracts fetus using the Adaptive Noise Canceller based on single Src Chan or single reference channel
Electrocardiosignal.When only with single Src Chan, the shape information of some Fetal ECG signals may be lost；Only with list
When individual reference channel and linear FIR filter, may cause can not well approximate parent electrocardio signal in chest area to abdomen
It is non-linear between portion region, and then influence to extract the accuracy of result.
Therefore, on the basis of traditional Adaptive Noise Canceller based on single Src Chan or single reference channel, need
A kind of more reliable efficient adaptive Fetal ECG method for extracting signal is provided.
The content of the invention
It is an object of the invention to overcome the shortcoming and deficiency of existing method there is provided a kind of based on many abdomens to lead what is led with many chests
Adaptive Fetal ECG signal extraction system and method.This method can be used for Fetal ECG signal clinical detection, can not only overcome
Many abdomen pathway optimization problems, moreover it is possible to solve parent electrocardio signal and exist from chest area to abdomen area non-linear, and then lead
The problem of Fetal ECG signal extraction precision is low is caused, Fetal ECG signal can be extracted in real time and is diagnosed for clinical medicine.
The present invention adopts the technical scheme that to solve above-mentioned technical problem：
The invention provides a kind of adaptive Fetal ECG method for extracting signal led and led with many chests based on many abdomens, it is described from
Adaptation Fetal ECG method for extracting signal leads electrocardiosignal (1), many chests including many abdomens and leads electrocardiosignal (2), linear colligator
(3), the error output of Linear and nonlinear filters (4), adaptive LMS/RLS algorithm (5) and Adaptive Noise Canceller
(6)。
Many abdomens lead electrocardiosignal (1) by linear colligator (3) obtain Adaptive Noise Canceller Src Chan it is always defeated
Enter；Many chests lead electrocardiosignal (2) by Linear and nonlinear filters (4) obtain Adaptive Noise Canceller reference channel it is total
Output；
Utilize the coefficient of linear colligator (3) in the RLS algorithm adjustment Src Chan of adaptive LMS/RLS algorithm (5)；Profit
Adjust Linear and nonlinear filters (4) in each reference channel respectively with the LMS algorithm of adaptive LMS/RLS algorithm (5)
Weights；
The Src Chan of Adaptive Noise Canceller is always inputted and subtracts each other adaptively making an uproar of obtaining with reference channel always output
The error output (6) of acoustic suppressor, is used as the estimation of Fetal ECG signal.
It is described it is a kind of the adaptive Fetal ECG method for extracting signal led with many chests is led based on many abdomens, many abdomens lead electrocardio letter
It is from mother's belly lead and the collection of mother's chest leads and after pretreatment respectively that number (1) and many chests, which lead electrocardiosignal (2),
Obtain.
In the Src Chan of Adaptive Noise Canceller, many abdomens, which lead electrocardiosignal (1), includes p passage, can table
It is shown asP coefficient of the linear colligator (3) is represented byMany abdomens lead electrocardiosignal (1) process
Always input is represented by the Src Chan for the Adaptive Noise Canceller that linear colligator (3) is obtained
Wherein, p coefficient of linear colligator (3) must be met
In the reference channel of Adaptive Noise Canceller, many chests, which lead electrocardiosignal (2), includes q passage, can table
It is shown asThe Linear and nonlinear filters (4) also include q passage, wherein, i-th passage it is linear and non-
The chest of linear filter i-th of passage of correspondence leads electrocardiosignal；Many chests lead electrocardiosignal (2) and pass through Linear and nonlinear filters
(4) the total output of reference channel for obtaining Adaptive Noise Canceller is represented by md(n)。
It is described it is a kind of the adaptive Fetal ECG method for extracting signal led with many chests is led based on many abdomens, using adaptive
The coefficient of linear colligator (3) in the RLS algorithm adjustment Src Chan of LMS/RLS algorithms (5)；Calculated using adaptive LMS/RLS
The LMS algorithm of method (5) adjusts the weights of Linear and nonlinear filters (4) in each reference channel respectively.
In the Src Chan of Adaptive Noise Canceller, the RLS algorithm using adaptive LMS/RLS algorithm (5) is real-time
The coefficient of linear colligator (3) in ground adjustment Src ChanAnd the change that many abdomens lead electrocardiosignal is quickly tracked, from
The selection optimization of many abdomen pathways is adaptively carried out, its process is
Wherein, p abdomen leads electrocardiosignal and is expressed as vector form d (n)=[f1(n),f2(n),…,fp(n)]T；Linear junction
P coefficient vector form of clutch is C (n)=[c1(n),c2(n),…,cp(n)]T；P coefficient estimate of linear colligator
Vector form isE=[1,1 ..., 1]TVectorial for p × 1, P (n) is p × p gain
Matrix；[·]TRepresent the transposition of vector or matrix；E (n) exports for the error of Adaptive Noise Canceller.
In the reference channel of Adaptive Noise Canceller, e (n) energy minimum principle is exported based on error, using adaptive
The LMS algorithm of LMS/RLS algorithms (5) is answered to adjust the weights of Linear and nonlinear filters (4) in each reference channel respectively.
On the premise of guarantee system is stable, the renewal step-length of LMS algorithm can influence the convergence rate of filter weights.I-th with reference to logical
Road it is non-linear between chest area to abdomen area, can be by reference to corresponding i-th of linear processes in passage
The weights of wave filter are provided；So as to which the weights of all q Linear and nonlinear filters in reference channel can be achieved all q
The nonlinear estimation between chest area to abdomen area of reference channel.
It is described it is a kind of the adaptive Fetal ECG method for extracting signal led with many chests is led based on many abdomens, adaptive noise
The Src Chan of Canceller always inputs fm(n) m is always exported with reference channeld(n) mistake of obtained Adaptive Noise Canceller is subtracted each other
Difference output (6), i.e. e (n)=fm(n)-md(n), as the estimation of Fetal ECG signal.
The present invention has advantages below and useful achievement：
1st, the inventive method leads electrocardiosignal using many abdomens, abundant Fetal ECG signal waveform information is contained, after being
Continuous Fetal ECG waveform analysis provides foundation；
2nd, the inventive method uses the linear colligator in Src Chan, many abdomens is led with electrocardiosignal progress channel selecting excellent
Change, with it is adaptive the characteristics of, many abdomens of real-time tracing lead the change of electrocardiosignal；
3rd, the inventive method uses multiple Linear and nonlinear filters in reference channel, is adaptively adjusted wave filter
Weights, can estimate that parent electrocardio signal is delivered in chest area more well non-linear between abdomen area；
4th, the inventive method design is succinct, it is easy to accomplish, and computation complexity is low, and Fetal ECG signal can be extracted in real time
For clinical medicine diagnosis.
What the present invention was provided leads the adaptive Fetal ECG method for extracting signal led with many chests based on many abdomens, is led using many abdomens
Electrocardiosignal and many chests lead electrocardiosignal respectively through linear processes in linear colligator in Src Chan and reference channel
Wave filter, utilizes adaptive RLS (recursive least square, recursive least-squares) algorithm and adaptive LMS
(least mean square, lowest mean square) algorithm adjusts the coefficient and reference channel of linear colligator in Src Chan respectively
The weights of middle Linear and nonlinear filters.
The present invention leads the Src Chan that electrocardiosignal obtains Adaptive Noise Canceller by linear colligator using many abdomens
Total input；Many chests lead electrocardiosignal by Linear and nonlinear filters obtain Adaptive Noise Canceller reference channel it is always defeated
Go out；The coefficient of linear colligator in Src Chan is adjusted using adaptive RLS algorithm；Adjusted respectively using ADAPTIVE LMS ALGORITHM
The weights of Linear and nonlinear filters in each reference channel；The Src Chan of Adaptive Noise Canceller is always inputted and ginseng
Examine passage always output and subtract each other obtained error output, be used as the estimation of Fetal ECG signal.
Carried with the Fetal ECG signal of the existing Adaptive Noise Canceller based on single Src Chan or single reference channel
Method is taken to compare, the present invention not only overcomes many abdomen pathway optimization problems, while also solving parent electrocardio signal from chest
Region to abdomen area exist it is non-linear cause the problem of Fetal ECG signal extraction precision is low, be it is a kind of it is more reliable more effectively
Adaptive Fetal ECG method for extracting signal.The present invention can extract Fetal ECG signal in real time and be diagnosed for clinical medicine.
Brief description of the drawings
Fig. 1 is the adaptive Fetal ECG signal extraction structural representation of the inventive method；
Fig. 2 is the structural representation of p-th of linear colligator in i-th of FLANN in the inventive method；
Fig. 3 is the structural representation of i-th of GFLANN in the inventive method；
Fig. 4 gives the sample data that length in database is at 1000 points；
Fig. 5 gives the Fetal ECG signal obtained when different number of abdomen leads electrocardiosignal；
Fig. 6 gives the Fetal ECG signal obtained when reference channel uses the form of different wave filters.
Embodiment
With reference to embodiment and accompanying drawing, the present invention is described in further detail, but embodiments of the present invention are not
Only not limited to this.
As shown in figure 1, the invention provides a kind of adaptive Fetal ECG signal extraction led and led with many chests based on many abdomens
Method, the adaptive Fetal ECG method for extracting signal leads electrocardiosignal 1, many chests including many abdomens and leads electrocardiosignal 2, linear junction
Clutch 3, Linear and nonlinear filters 4, the error output 6 of adaptive LMS/RLS algorithm 5 and Adaptive Noise Canceller.Institute
Linear FIR filter and GFLANN (generalized functional can be used respectively by stating Linear and nonlinear filters 4
Link artificial neural network, generalized function link artificial neural network) the specific reality of nonlinear filter
It is existing.
Many abdomens are led the Src Chan that electrocardiosignal 1 obtains Adaptive Noise Canceller by linear colligator 3 and always inputted；It is many
Chest leads the ginseng that electrocardiosignal 2 obtains Adaptive Noise Canceller by linear FIR filter and GFLANN nonlinear filter devices
Passage is examined always to export.
It is described it is a kind of the adaptive Fetal ECG method for extracting signal led with many chests is led based on many abdomens, many abdomens lead electrocardio letter
Number 1 and many chests lead electrocardiosignal 2 and led respectively from mother's abdomen and gather and obtained after pretreatment with mother's chest leads.
In the Src Chan of Adaptive Noise Canceller, many abdomens, which lead electrocardiosignal 1, includes p passage, can represent
ForP coefficient of the linear colligator 3 is represented byMany abdomens lead electrocardiosignal 1 and pass through linear junction
Always input is represented by the Src Chan for the Adaptive Noise Canceller that clutch (3) is obtained
Wherein, p coefficient of linear colligator 3 must be met
In the reference channel of Adaptive Noise Canceller, many chests, which lead electrocardiosignal 2, includes q passage, can represent
ForLinear FIR filter and the GFLANN nonlinear filter also includes q passage, wherein, i-th of passage
The chest of linear FIR filter i-th of passage corresponding with GFLANN nonlinear filters lead electrocardiosignal；Many chests lead electrocardiosignal
2 reference channels for obtaining Adaptive Noise Canceller by linear FIR filter and GFLANN nonlinear filters are always exported can
It is expressed as md(n)。
As shown in figure 1, adaptive Fetal ECG signal extraction structural representation.I-th of linear FIR filter is output as
In formula, L1- 1 is the exponent number of linear FIR filter, wi,l(n) it is l-th of filtering of i-th of linear FIR filter
Weights.
As shown in Fig. 2 give i-th FLANN (functional link artificial neural network,
Function link artificial neural network) in p-th of linear colligator structural representation.mi(n) led for the chest of i-th of reference channel
Electrocardiosignal；z-1Represent first-order lag computing；For p-th linear colligator in i-th of FLANN
Filter weights；L2For the time delay of FLANN nonlinear extensions, value may differ from L1.It is L to length2Input vector
[mi(n),mi(n-1),…,mi(n-L2+1)]TCarry out the non-linear triangle extension of pth rank, the vector after being extended accordingly
For
[sin[pπmi(n)],cos[pπmi(n)],sin[pπmi(n-1)],cos[pπmi(n-1)],…,
sin[pπmi(n-L2+1)],cos[pπmi(n-L2+1)]]T
In formula, []TRepresent vector or matrix transposition computing.
So, corresponding i-th of FLANN is output as
In formula, PmaxRepresent non-linear triangle extension exponent number p maximum.
As shown in figure 3, giving i-th of GFLANN structural representation, it includes i-th of FLANN and cross term two
Part.Nonlinear extensions exponent number is pgAnd time delay is Nd(1≤Nd≤L3- 1) cross term can pass through following 2NdIndividual vector,
Can compactly it be expressed as
In formula,(1≤k≤2Nd) represent pth in i-th of reference channelgK-th of input of individual linear colligator
Vector；Nonlinear extensions exponent number pgMaximum Pgmax。
Correspondingly, i-th of GFLANN is output as
In formula, the nonlinear object relevant with cross term in i-th of GFLANN of Section 2 correspondence of equation right-hand side expression；WithFor the filter weights relevant with cross term；PgmaxRepresent p in cross termgMaximum, L3To intersect
Time delay in, its value likely differs from L1And L2。
Finally, the reference channel of Adaptive Noise Canceller is always output as
It is described it is a kind of the adaptive Fetal ECG method for extracting signal led with many chests is led based on many abdomens, using adaptive
The coefficient of linear colligator 3 in the RLS algorithm adjustment Src Chan of LMS/RLS algorithms 5；Utilize adaptive LMS/RLS algorithm 5
LMS algorithm adjusts the weights of linear FIR filter and GFLANN nonlinear filters in each reference channel respectively.
In the Src Chan of Adaptive Noise Canceller, using adaptive LMS/RLS algorithm 5 RLS algorithm in real time
Adjust the coefficient of linear colligator 3 in Src ChanAnd the change that many abdomens lead electrocardiosignal is quickly tracked, adaptively
Ground carries out the selection optimization of many abdomen pathways, and its process is as follows
Wherein, p abdomen leads electrocardiosignal and is expressed as vector form d (n)=[f1(n),f2(n),…,fp(n)]T；Linear junction
P coefficient vector form of clutch is C (n)=[c1(n),c2(n),…,cp(n)]T；P coefficient estimate of linear colligator
Vector form isE=[1,1 ..., 1]TVectorial for p × 1, P (n) is p × p gain
Matrix；E (n) exports for the error of Adaptive Noise Canceller.
In the reference channel of Adaptive Noise Canceller, e (n) energy minimum principle is exported based on error, using adaptive
The LMS algorithm of LMS/RLS algorithms 5 is answered to adjust linear FIR filter and GFLANN nonlinear filterings in each reference channel respectively
The weights of device.On the premise of ensureing that system is stable, the renewal step-length of LMS algorithm can influence the convergence rate of filter weights.
I-th reference channel it is non-linear between chest area to abdomen area, can be by reference to corresponding i-th of line in passage
The weights of property FIR filter and GFLANN nonlinear filters are provided；So as to all q linear FIR filters in reference channel
With the weights of GFLANN nonlinear filters, the non-between chest area to abdomen area of all q reference channels can be achieved
Linear estimation.
The more new formula of i-th of reference channel linear FIR filter weights is
wi,l(n+1)=wi,l(n)+μfir,ie(n)mi(n-l)
In formula, μfir,iFor the renewal step-length of i-th of linear FIR filter.Similarly, line all in i-th of GFLANN
The right value update mode of property colligator is
αi,p,l(n+1)=αi,p,l(n)+μf,ie(n)sin[pπmi(n-l)]
βi,p,l(n+1)=βi,p,l(n)+μf,ie(n)cos[pπmi(n-l)]
In formula, μf,iAnd μg,iLinear colligator right value update step-length and i-th of GFLANN in i-th of FLANN are corresponded to respectively
Cross term in linear colligator right value update step-length.Unified renewal step-length can be used, μ is designated asfir、μf、μg, that is, meet μfir
=μfir,i、μf=μf,i、μg=μg,i, to all i=1,2 ..., q.
It is described it is a kind of the adaptive Fetal ECG method for extracting signal led with many chests is led based on many abdomens, adaptive noise
The Src Chan of Canceller always inputs fm(n) m is always exported with reference channeld(n) mistake of obtained Adaptive Noise Canceller is subtracted each other
Difference output 6, i.e. e (n)=fm(n)-md(n), as the estimation of Fetal ECG signal.
Above-described embodiment is the present invention preferably embodiment, but embodiments of the present invention are not by above-described embodiment
Limitation, other any Spirit Essences without departing from the present invention and the change made under principle, modification, replacement, combine, simplification,
Equivalent substitute mode is should be, is included within protection scope of the present invention.
The present invention combines actual open DaISy databases, and further the technique effect of the checking present invention is as follows：
The database includes obtaining from pregnant woman's external electrode, and sample rate is 250Hz, and time span is 10 seconds.Choose three
The chest of passage leads electrocardiosignal (q=3) and the abdomen of three passages leads electrocardiosignal (p=3), as shown in figure 4, giving database
Middle length is 1000 points of sample data.The adaptive Fetal ECG signal extraction structure provided according to Fig. 1, is filtered by adjusting
The customer parameter such as length and step-length so that whole adaptive Fetal ECG signal extraction reaches good performance, and then obtains phase
To high-quality Fetal ECG signal.The customer parameter value decibel of selection is μfir=0.02, μf=0.00005, μg=
0.002、L1=75, L2=35, L3=35, Pmax=30, Pgmax=2 and Nd=3.
Adaptive Fetal ECG signal extraction effect of the electrocardiosignal to the present invention is led in order to contrast different number of abdomen, can
Electrocardiosignal is led using equal number of chest, q=3 is chosen, correspondingly changes the number that abdomen leads electrocardiosignal, the checking present invention's
The performance of Fetal ECG signal extraction system and method.Obtained as shown in figure 5, giving when different number of abdomen leads electrocardiosignal
The Fetal ECG signal obtained：Correspond to abdomen respectively from top to bottom and lead electrocardiosignal f1(n), the abdomen of single passage leads electrocardiosignal (p=
1) Fetal ECG signal, the abdomen of two passages after being extracted when extracted when leading electrocardiosignal (p=2) after Fetal ECG signal,
Fetal ECG signal when the abdomen of three passages leads electrocardiosignal (p=3) after extraction.Part is marked respectively using filled arrows
The position of Fetal ECG signal R ripples and parent electrocardio signal R ripples.The three kinds of different abdomens provided in Fig. 5 are led in the case of electrocardiosignal
The mean square error of Fetal ECG signal waveform is respectively 5.31 × 10-3(p=1), 2.92 × 10-3(p=2), 2.45 × 10-3(p
=3).The chest that marked three different time points using solid oval leads the denoising result of electrocardiosignal.Understand, as abdomen leads the heart
The number increase of electric signal, Fetal ECG signal extraction system of the invention will obtain relatively sharp Fetal ECG signal.
In order to contrast influence of the different Linear and nonlinear filters to Fetal ECG signal extraction result, using three
The chest of passage leads electrocardiosignal (q=3) and the abdomen of three passages leads electrocardiosignal (p=3).In this example, reference channel is used
The form of linear FIR filter and GFLANN nonlinear filters, the reference channel with conventional extraction system is only with linear FIR
The form of wave filter is contrasted.As shown in fig. 6, giving the tire obtained when reference channel uses the form of different wave filters
Youngster's electrocardiosignal：Correspond to abdomen respectively from top to bottom and lead electrocardiosignal f1(n), reference channel only with linear FIR filter when carry
Tire when Fetal ECG signal, reference channel after taking use linear FIR filter and GFLANN nonlinear filters after extraction
Youngster's electrocardiosignal.It marked the position of part Fetal ECG signal R ripples and parent electrocardio signal R ripples respectively using filled arrows.
The two kinds of reference channels provided in Fig. 6 use the Fetal ECG signal mean square error difference 4.33 obtained during different filter forms
×10-3With 2.45 × 10-3.Similarly, the chest that marked three different time points using solid oval leads the denoising of electrocardiosignal
Effect.Understand, while relatively being passed using the Fetal ECG signal extraction system of linear FIR filter and GFLANN nonlinear filters
That unites has better performance only with the Fetal ECG signal extraction system of linear FIR filter.
Claims (3)
1. a kind of lead the adaptive Fetal ECG signal extraction system led with many chests based on many abdomens, it is characterised in that described adaptive
Answer Fetal ECG signal extraction system to include linear colligator (3), Linear and nonlinear filters (4), adaptive LMS/RLS to calculate
Method (5) and Adaptive Noise Canceller (6)；
Many abdomens, which lead electrocardiosignal (1) and obtain the Src Chan of Adaptive Noise Canceller by linear colligator (3), always to be inputted；It is many
Chest, which leads electrocardiosignal (2) and obtains the reference channel of Adaptive Noise Canceller by Linear and nonlinear filters (4), always to be exported；
Many abdomens lead electrocardiosignal (1) and many chests to lead electrocardiosignal (2) led from mother's belly lead and mother's chest respectively
What connection was gathered and obtained after pretreatment；
Utilize the coefficient of linear colligator (3) in the RLS algorithm adjustment Src Chan of adaptive LMS/RLS algorithm (5)；Using certainly
The LMS algorithm for adapting to LMS/RLS algorithms (5) adjusts the weights of Linear and nonlinear filters (4) in each reference channel respectively；
The Src Chan of Adaptive Noise Canceller (6) is always inputted always to export with reference channel and subtracts each other obtained adaptive noise
The error output e (n) of Canceller, is used as the estimation of Fetal ECG signal.
2. a kind of lead the adaptive Fetal ECG method for extracting signal led with many chests based on many abdomens, it is characterised in that described adaptive
Fetal ECG method for extracting signal is answered based on system described in claim 1 to realize, its process is：
In the Src Chan of Adaptive Noise Canceller, many abdomens, which lead electrocardiosignal (1), includes p passage, is expressed asP coefficient of the linear colligator (3) is represented byMany abdomens lead electrocardiosignal (1) by linear
Always input is represented by the Src Chan for the Adaptive Noise Canceller that colligator (3) is obtained
<mrow>
<msub>
<mi>f</mi>
<mi>m</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>p</mi>
</munderover>
<msub>
<mi>c</mi>
<mi>j</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<msub>
<mi>f</mi>
<mi>j</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
</mrow>
Wherein, p coefficient of linear colligator (3) must be met
In the reference channel of Adaptive Noise Canceller, many chests, which lead electrocardiosignal (2), includes q passage, is represented byThe Linear and nonlinear filters (4) also include q passage, wherein, the linear processes of i-th of passage
The chest of wave filter i-th of passage of correspondence leads electrocardiosignal；Many chests lead electrocardiosignal (2) by Linear and nonlinear filters (4)
The total output of reference channel for obtaining Adaptive Noise Canceller is represented by md(n)；
Utilize the coefficient of linear colligator (3) in the RLS algorithm adjustment Src Chan of adaptive LMS/RLS algorithm (5)；Using certainly
The LMS algorithm for adapting to LMS/RLS algorithms (5) adjusts the weights of Linear and nonlinear filters (4) in each reference channel respectively,
In the Src Chan of Adaptive Noise Canceller, adjusted in real time using the RLS algorithm of adaptive LMS/RLS algorithm (5)
The coefficient of linear colligator (3) in whole Src ChanAnd the change that many abdomens lead electrocardiosignal is quickly tracked, adaptively
Ground carries out the selection optimization of many abdomen pathways, and its process is
<mrow>
<mover>
<mi>C</mi>
<mo>^</mo>
</mover>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mover>
<mi>C</mi>
<mo>^</mo>
</mover>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>+</mo>
<mfrac>
<mrow>
<mi>P</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mi>d</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mn>1</mn>
<mo>+</mo>
<msup>
<mi>d</mi>
<mi>T</mi>
</msup>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mi>P</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mi>d</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mi>e</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>P</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>P</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mfrac>
<mrow>
<mi>P</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mi>d</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<msup>
<mi>d</mi>
<mi>T</mi>
</msup>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mi>P</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mn>1</mn>
<mo>+</mo>
<msup>
<mi>d</mi>
<mi>T</mi>
</msup>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mi>P</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mi>d</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
</mrow>
<mrow>
<mi>C</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mover>
<mi>C</mi>
<mo>^</mo>
</mover>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mi>P</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mfrac>
<mrow>
<msup>
<mi>E</mi>
<mi>T</mi>
</msup>
<mover>
<mi>C</mi>
<mo>^</mo>
</mover>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mrow>
<msup>
<mi>E</mi>
<mi>T</mi>
</msup>
<mi>P</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mi>E</mi>
</mrow>
</mfrac>
<mi>E</mi>
</mrow>
Wherein, p abdomen leads electrocardiosignal and is expressed as vector form d (n)=[f1(n),f2(n),…,fp(n)]T；Linear colligator
P coefficient vector form be C (n)=[c1(n),c2(n),…,cp(n)]T；The p coefficient estimate vector of linear colligator
Form isE=[1,1 ..., 1]TVectorial for p × 1, P (n) is p × p gain matrix；
[·]TRepresent the transposition of vector or matrix；E (n) exports for the error of Adaptive Noise Canceller；
The Src Chan of the Adaptive Noise Canceller always inputs fm(n) m is always exported with reference channeld(n) subtract each other obtain from
Adapt to the error output of adaptive noise canceller (6), i.e. e (n)=fm(n)-md(n), as the estimation of Fetal ECG signal.
3. it is according to claim 2 it is a kind of the adaptive Fetal ECG method for extracting signal led with many chests is led based on many abdomens,
Characterized in that,
In the reference channel of Adaptive Noise Canceller, e (n) energy minimum principle is exported based on error, using adaptive
The LMS algorithm of LMS/RLS algorithms (5) adjusts the weights of Linear and nonlinear filters (4) in each reference channel respectively；
I-th reference channel it is non-linear between chest area to abdomen area, can be by reference in passage corresponding i-th
The weights of individual Linear and nonlinear filters are provided；So as to the power of all q Linear and nonlinear filters in reference channel
Value, can be achieved the nonlinear estimation between chest area to abdomen area of all q reference channels.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710148736.7A CN107049303A (en) | 2017-03-10 | 2017-03-10 | It is a kind of that the adaptive Fetal ECG signal extraction system and method led with many chests is led based on many abdomens |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710148736.7A CN107049303A (en) | 2017-03-10 | 2017-03-10 | It is a kind of that the adaptive Fetal ECG signal extraction system and method led with many chests is led based on many abdomens |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107049303A true CN107049303A (en) | 2017-08-18 |
Family
ID=59623111
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710148736.7A Pending CN107049303A (en) | 2017-03-10 | 2017-03-10 | It is a kind of that the adaptive Fetal ECG signal extraction system and method led with many chests is led based on many abdomens |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107049303A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108113665A (en) * | 2017-12-14 | 2018-06-05 | 河北大学 | A kind of automatic noise-reduction method of electrocardiosignal |
| CN110720905A (en) * | 2019-09-20 | 2020-01-24 | 中南大学 | Fetal electrocardiosignal interference suppression system based on array processing |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5042499A (en) * | 1988-09-30 | 1991-08-27 | Frank Thomas H | Noninvasive electrocardiographic method of real time signal processing for obtaining and displaying instantaneous fetal heart rate and fetal heart rate beat-to-beat variability |
| WO2000054650A2 (en) * | 1999-03-15 | 2000-09-21 | The Johns Hopkins University | Apparatus and method for non-invasive, passive fetal heart monitoring |
| CN1520272A (en) * | 2001-05-29 | 2004-08-11 | ���ֳ� | System for detection and analysis of material uterine, material and fetal cardiac and fetal brain activity |
| CN101513345A (en) * | 2008-12-26 | 2009-08-26 | 华南理工大学 | Quick blind source separating fetal electrocardioscanner and detection method |
| CN102090888A (en) * | 2011-03-30 | 2011-06-15 | 南京大学 | Method for separating and extracting fetal electrocardiogram by adopting multi-lead signal fusion mode |
| CN102613970A (en) * | 2012-02-13 | 2012-08-01 | 南京大学 | Method for expanding channel number of collected multichannel fetal electrocardiosignals |
| CN103610460A (en) * | 2013-12-11 | 2014-03-05 | 哈尔滨工业大学 | Fetus electrocardiosignal extracting method based on self-adaptation FLANN filter |
| CN104027105A (en) * | 2014-04-23 | 2014-09-10 | 河南科技大学 | Novel maternal and fetal electrocardiogram separation method |
-
2017
- 2017-03-10 CN CN201710148736.7A patent/CN107049303A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5042499A (en) * | 1988-09-30 | 1991-08-27 | Frank Thomas H | Noninvasive electrocardiographic method of real time signal processing for obtaining and displaying instantaneous fetal heart rate and fetal heart rate beat-to-beat variability |
| WO2000054650A2 (en) * | 1999-03-15 | 2000-09-21 | The Johns Hopkins University | Apparatus and method for non-invasive, passive fetal heart monitoring |
| CN1520272A (en) * | 2001-05-29 | 2004-08-11 | ���ֳ� | System for detection and analysis of material uterine, material and fetal cardiac and fetal brain activity |
| CN101513345A (en) * | 2008-12-26 | 2009-08-26 | 华南理工大学 | Quick blind source separating fetal electrocardioscanner and detection method |
| CN102090888A (en) * | 2011-03-30 | 2011-06-15 | 南京大学 | Method for separating and extracting fetal electrocardiogram by adopting multi-lead signal fusion mode |
| CN102613970A (en) * | 2012-02-13 | 2012-08-01 | 南京大学 | Method for expanding channel number of collected multichannel fetal electrocardiosignals |
| CN103610460A (en) * | 2013-12-11 | 2014-03-05 | 哈尔滨工业大学 | Fetus electrocardiosignal extracting method based on self-adaptation FLANN filter |
| CN104027105A (en) * | 2014-04-23 | 2014-09-10 | 河南科技大学 | Novel maternal and fetal electrocardiogram separation method |
Non-Patent Citations (2)
| Title |
|---|
| MA, YAPING，ET AL: "A multichannel nonlinear adaptive noise canceller based on generalized FLANN for fetal ECG extraction", 《MEASUREMENT SCIENCE AND TECHNOLOGY》 * |
| MAHA SHADAYDEH，ET AL: "Extraction of fetal ECG using adaptive Volterra filters", 《6TH EUROPEAN SINGAL PROCESSING CONFERENC(EUSIPCO 2008)》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108113665A (en) * | 2017-12-14 | 2018-06-05 | 河北大学 | A kind of automatic noise-reduction method of electrocardiosignal |
| CN110720905A (en) * | 2019-09-20 | 2020-01-24 | 中南大学 | Fetal electrocardiosignal interference suppression system based on array processing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103610460B (en) | A kind of Fetal ECG method for extracting signal based on self adaptation FLANN wave filter | |
| CN107811626A (en) | A kind of arrhythmia classification method based on one-dimensional convolutional neural networks and S-transformation | |
| CN102160787B (en) | Time-frequency-transformation-based blind extraction method of fetal electrocardiography | |
| CN107822622A (en) | Electrocardiographic diagnosis method and system based on depth convolutional neural networks | |
| Kumar et al. | Performance analysis of different filters for power line interface reduction in ECG signal | |
| Jaros et al. | Novel hybrid extraction systems for fetal heart rate variability monitoring based on non-invasive fetal electrocardiogram | |
| CN106889981B (en) | A kind of intelligent terminal for being used to extract fetal heart frequency | |
| CN109431492A (en) | ECG lead signals based on neural network algorithm simulate method for reconstructing | |
| CN107928654A (en) | A kind of pulse wave signal blood pressure detecting method based on neutral net | |
| Sinha et al. | Backpropagation artificial neural network classifier to detect changes in heart sound due to mitral valve regurgitation | |
| CN109567792A (en) | A kind of single channel abdomen record fetal electrocardiogram extracting method | |
| CN113384277B (en) | Electrocardiogram data classification method and classification system | |
| CN109893118A (en) | A kind of electrocardiosignal classification diagnosis method based on deep learning | |
| CN107049303A (en) | It is a kind of that the adaptive Fetal ECG signal extraction system and method led with many chests is led based on many abdomens | |
| Mohammed Kaleem et al. | A survey on FECG extraction using neural network and adaptive filter | |
| CN107480637A (en) | Heart failure based on heart sound feature method by stages | |
| CN110974213A (en) | Electrocardiosignal identification method based on deep stack network | |
| Ruffo et al. | Non-invasive foetal monitoring with combined ECG-PCG system | |
| Poungponsri et al. | Electrocardiogram (ECG) signal modeling and noise reduction using wavelet neural networks | |
| CN106691437B (en) | A kind of fetal heart frequency extracting method based on parent electrocardio signal | |
| CN103876731A (en) | Extraction device and method for fetus electrocardiosignals | |
| CN105662391B (en) | A kind of feature extraction and classifying system and method for stomach magnetic signal | |
| Rakotomamonjy et al. | Automated neural network detection of wavelet preprocessed electrocardiogram late potentials | |
| Anisha et al. | Survey on fetal ECG extraction | |
| Rajesh et al. | Comprehensive study on fetal ECG extraction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170818 |