CN113017640A - Magnetocardiogram signal background noise S transform domain removing method and system - Google Patents
Magnetocardiogram signal background noise S transform domain removing method and system Download PDFInfo
- Publication number
- CN113017640A CN113017640A CN202110202791.6A CN202110202791A CN113017640A CN 113017640 A CN113017640 A CN 113017640A CN 202110202791 A CN202110202791 A CN 202110202791A CN 113017640 A CN113017640 A CN 113017640A
- Authority
- CN
- China
- Prior art keywords
- magnetocardiogram
- frequency
- time
- noise
- dimensional time
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 49
- 239000011159 matrix material Substances 0.000 claims abstract description 60
- 238000011426 transformation method Methods 0.000 claims abstract description 7
- 230000001131 transforming effect Effects 0.000 claims abstract description 7
- 238000012545 processing Methods 0.000 claims abstract description 6
- 230000009466 transformation Effects 0.000 claims description 19
- 230000008569 process Effects 0.000 claims description 14
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 230000000694 effects Effects 0.000 abstract description 4
- 230000000873 masking effect Effects 0.000 abstract description 3
- 230000008030 elimination Effects 0.000 abstract 1
- 238000003379 elimination reaction Methods 0.000 abstract 1
- 238000001228 spectrum Methods 0.000 description 15
- 238000001514 detection method Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000028161 membrane depolarization Effects 0.000 description 4
- 230000002107 myocardial effect Effects 0.000 description 4
- 238000000342 Monte Carlo simulation Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000002565 electrocardiography Methods 0.000 description 3
- 208000019622 heart disease Diseases 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 230000000747 cardiac effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000004217 heart function Effects 0.000 description 2
- 208000020446 Cardiac disease Diseases 0.000 description 1
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000005311 autocorrelation function Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 210000004903 cardiac system Anatomy 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000005672 electromagnetic field Effects 0.000 description 1
- 230000007831 electrophysiology Effects 0.000 description 1
- 238000002001 electrophysiology Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 210000005246 left atrium Anatomy 0.000 description 1
- 210000005240 left ventricle Anatomy 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001575 pathological effect Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000004451 qualitative analysis Methods 0.000 description 1
- 230000002336 repolarization Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 210000005245 right atrium Anatomy 0.000 description 1
- 210000005241 right ventricle Anatomy 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- 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
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Pathology (AREA)
- Medical Informatics (AREA)
- Psychiatry (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Physiology (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a method and a system for removing a background noise S transform domain of a magnetocardiogram signal, which belong to the technical field of processing of magnetocardiogram signals and comprise the following steps: s1: collecting noise data points; s2: s transforming the noise; s3: drawing a maximum two-dimensional time-frequency curved surface; s4: determining a distinguishing critical interface; s5: comparing the relation between the magnetocardiogram signal two-dimensional time-frequency matrix and the distinguishing critical interface; s6: an inverse S transform is performed. The method adopts the S transformation method to eliminate the background noise in the time-frequency domain, and has the characteristics of simple realization, good background noise elimination effect, high running speed, no need of purchasing additional hardware facilities and the like; the method has strong universality, is suitable for removing background noise of magnetocardiogram signal data interfered by high signal-to-noise ratio under the condition of no shielding and no masking and accurately extracting magnetocardiogram signal parameters, and is worthy of popularization and application.
Description
Technical Field
The invention relates to the technical field of heart magnetic signal processing, in particular to a method and a system for removing a background noise S transform domain of a heart magnetic signal.
Background
Heart disease is one of the major diseases that endanger human health. Early detection of cardiac function is critical for the prevention and treatment of cardiac disease. Currently, Electrocardiography (ECG) is used as a method for detecting and diagnosing heart function to detect and analyze electrical activity of heart. The electrocardiogram generally consists of a P wave, a QRS wave group and an ST wave band, wherein the P wave represents a left atrium depolarization process and a right atrium depolarization process, the QRS wave group represents a left ventricle depolarization process and a right ventricle depolarization process, the S-T wave band reflects a myocardial cell repolarization state, and transient information implied by the QRS wave group is extremely important for identifying the electromagnetic state change of a heart system and can also provide a premonitory characteristic of certain pathological states of the heart system.
With the development of magnetic sensing devices, superconducting quantum interferometers as detectors for nondestructive electromagnetic detection have the advantages of high sensitivity, wide bandwidth, good linearity and the like, and are widely applied to magnetocardiogram signal measurement and material nondestructive detection. The waveform of a Magnetocardiogram (MCG) is the same as that of an electrocardiogram, i.e., it is composed of P-wave, QRS-wave and ST-wave, and when a certain part of a human body changes, the current and magnetic field in the body change, and the detection of the cardiac electrophysiology through the change of the cardiac magnetic field is the theoretical basis for the clinical use of MCG. MCG has its unique features compared to ECG, which is related to the transmembrane current intensity of myocardial cells, and depends mainly on the electrical conductivity of myocardial tissue and the body current intensity near the body surface; MCG relies on the magnetic field distribution, which results from bulk currents, particularly the vortex current source and bulk current portion with high current density in myocardial tissue. Actual measurements of the magnetic field at the chest indicate that MCG can provide more important details of the nature of the electromagnetic field source of the cardiac system, particularly eddy current source information, which cannot be measured by ECG. However, under the condition of an ordinary unmasked and unshielded detection room or laboratory, a great amount of background noise exists in magnetocardiogram signals acquired by a high-sensitivity high-temperature radio-frequency superconducting quantum interferometer, and the main components of the background noise are 50Hz power frequency interference signals and other external environment noise. The low frequency magnetic noise interference is about a few nano-Tesla (10)-9T) is about 100 times of human heart signals, and the interference of 50Hz commercial power and harmonics thereof reaches 10- 8T even 10-7T magnitude, 10 of magnetic signal of human heart3～104On the order of multiples, these signals strongly influence the efficient identification of magnetocardiogram signal features. How to effectively extract magnetocardiogram signals and related wave band characteristics from strong background noise becomes a key for magnetocardiogram signal extraction, and becomes the key to be solved in the process of heart disease diagnosisThe key technical problem is solved. Therefore, a method and a system for removing the S transform domain of the magnetocardiogram signal background noise are provided.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: how to effectively extract magnetocardiogram signals and related wave band characteristics from strong background noise provides a magnetocardiogram signal background noise S transform domain removing method.
The invention solves the technical problems through the following technical scheme, and the invention comprises the following steps:
s1: collecting noise data points
Acquiring N data points of background noise under the environment of an instrument in real time;
s2: s-transform noise
Converting the acquired N data point background noises into a two-dimensional time-frequency matrix by adopting an S conversion method;
s3: drawing maximum two-dimensional time-frequency curved surface
Continuously acquiring for multiple times, calculating a two-dimensional time-frequency matrix corresponding to each acquired signal to obtain multiple groups of two-dimensional time-frequency matrixes, calculating the maximum value of the time-frequency matrixes of each time-frequency point in the statistical process to obtain a background noise S-transform two-dimensional time-frequency matrix under the statistical condition, and drawing the matrix to obtain a maximum two-dimensional time-frequency curved surface;
s4: determining a discriminating critical interface
Taking a set confidence coefficient surface of the maximum value surface as a distinguishing critical interface of noise and magnetocardiogram signals;
s5: comparing the relation between the two-dimensional time-frequency matrix of magnetocardiogram signals and the critical interface
Transforming the acquired magnetocardiogram signals into a two-dimensional time-frequency matrix by adopting an S transformation method, and reserving the two-dimensional time-frequency matrix of the magnetocardiogram signals when the two-dimensional time-frequency matrix of the acquired magnetocardiogram signals is larger than a distinguishing critical interface; when the two-dimensional time-frequency matrix of the magnetocardiogram signal is smaller than or equal to the distinguishing critical interface, setting the two-dimensional time-frequency matrix of the magnetocardiogram signal to be 0, and modifying the two-dimensional time-frequency matrix of the magnetocardiogram signal;
s6: performing inverse S transform
And performing inverse S transformation on the modified magnetocardiogram signal two-dimensional time-frequency matrix to obtain the magnetocardiogram signal without background noise.
Further, in the step S1, the environment of the instrument is the environment of the instrument before the magnetocardiogram signal is measured.
Further, in the step S2, the S transform is a short-time fourier transform using a gaussian window function, and is expressed as:
wherein t is time, f is frequency, τ is time shift factor, i is an imaginary number, and x (t) represents real-time continuous noise.
Further, in the step S4, a confidence level of 95% is set.
Further, in the step S6, the inverse S transform is an inverse short-time fourier transform using a gaussian window function, which is expressed as:
wherein u and t are time, f is frequency, τ is time shift factor, i is imaginary number, and x (u) is magnetocardiogram signal with background noise removed.
The invention also provides a system for removing the background noise S transform domain of the magnetocardiogram signal, which adopts the background noise removing method to remove the background noise in the magnetocardiogram signal, and comprises the following steps:
the noise acquisition module is used for acquiring N data points of background noise in real time under the environment where the instrument is located;
the noise transformation module is used for transforming the acquired background noise of the N data points into a two-dimensional time-frequency matrix by adopting an S transformation method;
the drawing module is used for continuously collecting the signals for multiple times, calculating a two-dimensional time-frequency matrix corresponding to each collected signal to obtain a plurality of groups of two-dimensional time-frequency matrixes, calculating the maximum value of each time-frequency matrix of a plurality of groups of time-frequency points in the statistical process to obtain a background noise S-transform two-dimensional time-frequency matrix under the statistical condition, and drawing the matrix to obtain a maximum value two-dimensional time-frequency curved surface;
the critical interface determining module is used for taking a set confidence coefficient curved surface of the maximum curved surface as a distinguishing critical interface of the noise and the magnetocardiogram signal;
the comparison module is used for converting the acquired magnetocardiogram signals into a two-dimensional time-frequency matrix by adopting an S conversion method, and reserving the two-dimensional time-frequency matrix of the magnetocardiogram signals when the two-dimensional time-frequency matrix of the acquired magnetocardiogram signals is larger than a distinguishing critical interface; when the two-dimensional time-frequency matrix of the magnetocardiogram signal is smaller than the distinguishing critical interface, setting the two-dimensional time-frequency matrix of the magnetocardiogram signal to be 0, and modifying the two-dimensional time-frequency matrix of the magnetocardiogram signal;
the inverse transformation module is used for carrying out inverse S transformation on the modified magnetocardiogram signal two-dimensional time-frequency matrix to obtain the magnetocardiogram signal without background noise;
the central processing module is used for sending instructions to other modules to complete related actions;
the noise acquisition module, the noise transformation module, the drawing module, the critical interface determination module, the comparison module and the inverse transformation module are all electrically connected with the central processing module.
Compared with the prior art, the invention has the following advantages: according to the magnetocardiogram signal background noise S transform domain removing method, the background noise is eliminated in a time-frequency domain by adopting the S transform method, the whole process can be operated by software, and no artificial redundant operation is needed; the method is simple to realize, good in background noise removing effect and high in running speed, and extra hardware facilities do not need to be purchased; the method has strong universality, is suitable for removing background noise of magnetocardiogram signal data interfered by high signal-to-noise ratio under the condition of no shielding and no masking and accurately extracting magnetocardiogram signal parameters, and is worthy of popularization and application.
Drawings
FIG. 1 is a flow chart of a method for removing background noise of magnetocardiogram signals according to an embodiment of the present invention;
FIG. 2a is a diagram of time-frequency characteristics in the noise S domain according to an embodiment of the present invention;
fig. 2b is an average power spectrum (K1000) in an embodiment of the present invention;
FIG. 2c is a graph of the average power spectrum and 95% signal line in an embodiment of the present invention;
FIG. 2d is a graph of ratio versus frequency for an embodiment of the present invention;
FIG. 3 is a magnetocardiogram signal with background noise according to an embodiment of the present invention;
FIG. 4 is a three-dimensional time-frequency diagram of a magnetocardiogram signal S with background noise according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a critical plane for distinguishing noise from magnetocardiogram signals according to an embodiment of the present invention;
FIG. 6 shows the magnetocardiogram signal after background noise is removed according to an embodiment of the present invention.
Detailed Description
The following examples are given for the detailed implementation and specific operation of the present invention, but the scope of the present invention is not limited to the following examples.
The following experimental contents were made in this example:
1. statistical properties of noisy S-transform domain
Let mean 0 and variance σ2The autocorrelation function of white Gaussian noise n (t) is
E[n(t)n(u)]＝σ2δ(t-u) (1)
Where E denotes expectation, u is time, and δ (t) is an impulse function.
If the time-frequency matrix after S transformation of gaussian noise n (t) is S (τ, f), the average power spectrum of gaussian white noise is:
where denotes taking a complex conjugate.
When the noise is colored noise, the time domain iterative relationship of the colored noise x (t) obtained by using the first-order markov model is as follows:
x(t)＝wx(t-1)+n(t) (3)
wherein w is a hysteresis factor.
The average power spectrum under the corresponding colored noise fourier transform is:
wherein m is a discrete frequency point.
In order to research the power spectrum characteristic of colored noise in an S-transform time-frequency domain, firstly, qualitative analysis is carried out on the characteristic distribution of the colored noise in the S-transform two-dimensional time-frequency domain, the overall distribution rule of the colored noise is guessed, and then the rule is verified by utilizing a Monte Carlo method.
Because the noise follows Gaussian distribution, the time-frequency characteristic of the noise still follows the Gaussian distribution after the noise is transformed by adopting a rapid S transformation algorithm. The specific derivation process is as follows:
the first step, calculating Fourier transform of input Gaussian noise, wherein a real part and an imaginary part of a transform result are subjected to Gaussian distribution;
secondly, frequency shifting is carried out on the conversion result, and the real part and the imaginary part of the frequency shifting result still obey Gaussian distribution;
thirdly, calculating the product of the frequency shift result and a fixed value Gaussian window, wherein the product result does not influence the distribution of a real part and an imaginary part;
and fourthly, performing inverse Fourier transform on the product result, wherein the real part and the imaginary part of the transform result still obey Gaussian distribution.
Non-zero emission due to average power spectrum | S (τ, f)2The squares of the real part and the imaginary part of the time-frequency matrix after S transformation are respectively obeyed
Distribution, therefore, | S (τ, f) & gtnon phosphor2Can comply with a degree of freedom of 2%2And (4) distribution. The Monte Carlo method is adopted to verify the distribution rule.
The main steps of the verification by the Monte Carlo method are as follows:
1) at a given significance level α, finding a confidence interval 1- α;
2) construct mean μ and variance σ2White noise n (t) or colored noise x (t);
3) performing S transformation on n (t) or x (t) to obtain instantaneous workRate spectrum | S (τ, f) & gtnon & gtphosphor2；
4) Repeating the step 2) and the step 3) for K times, and calculating an average power spectrum
Calculating the curved surface (1-alpha) max { | S positioned in the confidence interval 1-alpha1(τ,f)|2,|S2(τ,f)|2,…,|SK(τ,f)|2}；
5) Calculating the ratio of the curved surface at the confidence interval 1-alpha to the average power spectrum curved surface
2. Background noise removing method based on noise statistical characteristics
Setting the noise-containing signal as: x (t) ═ u (t) + n (t), where u (t) is the signal source and n (t) is noise. In order to utilize the statistical characteristics of noise in the S-transform time-frequency domain, the following original hypothesis and assumed are firstly made:
H0: x (t) is noise, H1: x (t) is the signal source.
Wherein H0For the former hypothesis, H1Is assumed. When x (t) is colored noise, according to the statistical property of C (tau, f) in S transform time-frequency domain of colored noise, C (tau, f) should obey
Distribution, given a significance level α, within the 1- α confidence interval:
the rejection area is:
substituting the formula (5) into the formula (6) can obtain the original assumption condition within the 1-alpha confidence interval as follows:
the left side of the inequality is a time-frequency domain satisfying the original assumption of 1-alpha confidence interval, namely x (t) is a time-frequency region where noise is located:
similarly, x (t) satisfies that the time frequency region assumed as the signal source is:
according to the above experimental procedure, in the case that the significance level α is 0.05 and the confidence interval is 95%, the experimental statistics uses random white gaussian noise with a mean value of 0 and a variance of 0.1, the number of sampling points N is 128, and the number of repeated experimental statistics K is 1000. Fig. 2(a) shows an S-transform three-dimensional time-frequency height map corresponding to a single random noise, and it can be seen from fig. 2(a) that the amplitude of the noise in the S domain increases with the increase of the frequency, and this characteristic is consistent with the theoretical result shown in equation (2). As the number of statistics increases, the average power spectrum curve of the noise becomes gradually smooth, as shown in fig. 2 (b).
When τ is 32, the average power spectrum obtained by calculation and the power spectrum at the 95% confidence line are shown in fig. 2 (c). The ratio between the power spectrum at which 95% of the channels are located and the average power spectrum is shown in fig. 2 (d). As can be seen from fig. 2(d), the ratio of the two fluctuates around 3. The results of this experiment are
The upper limit 2.995 of the 95% confidence interval is very close. From the experimental statistics result, the statistic C (tau, f) of the noise in the S-transform time-frequency domain obeys
The distribution, the average power spectrum magnitude of the noise is proportional to the frequency.
The embodiment provides a magnetocardiogram signal background noise removing method, which comprises the following steps:
(1): acquiring N data points of background noise under the environment of an instrument in real time;
(2): transforming the acquired N data point background noises into a two-dimensional time-frequency matrix S [ N, M ] by adopting an S transformation method, wherein M is N/2, and an integer is taken;
(3): continuously collecting for 1000 times, calculating a two-dimensional time-frequency matrix corresponding to each collected signal, and obtaining 1000 groups of two-dimensional time-frequency matrices S1[ N, M ], S2[ N, M ], … and S1000[ N, M ];
(4): calculating the maximum value { max { S1[1, 1], S2[1, 1], …, S1000[1, 1] }, max { S1[2, 1], S2[2, 1], …, S1000[2, 1] }, …, max { S1[ N, M ], S2[ N, M ], … and S1000[ N, M ] } of each time frequency point time frequency matrix in 1000 times of statistical process, obtaining a noise S transformation two-dimensional time frequency matrix under the condition of statistics, and drawing a maximum value two-dimensional time frequency surface;
(5): taking the maximum value curved surface with 95% confidence coefficient as a distinguishing critical interface S [ Nmax, Mmax ] of the noise and the magnetocardiogram signal under the condition that the confidence coefficient is 95%;
(6): transforming the acquired magnetocardiogram signals into two-dimensional time-frequency matrixes S [ N1, M1] by adopting an S transformation method, and reserving S [ N1, M1] when the acquired magnetocardiogram signal two-dimensional time-frequency matrixes S [ N1, M1] are larger than a distinguishing critical interface S [ Nmax, Mmax ]; when S [ N1, M1] is less than or equal to S [ Nmax, Mmax ], S [ N1, M1] is set to 0, and S [ N1, M1] is modified;
(7): and performing inverse S transformation on the modified two-dimensional time-frequency matrix S [ N1, M1] to obtain the magnetocardiogram signal with background noise removed.
The magnetocardiogram signal background noise S transform domain removing design method is characterized in that in the step (1), the environment where the instrument is located is the same environment where magnetocardiogram signal measurement is located.
In this embodiment, the transform S in step (2) is a short-time fourier transform using a gaussian window function, and its expression is:
wherein t is time, f is frequency, τ is time shift factor, i is an imaginary number, and x (t) represents real-time continuous noise.
In this embodiment, the confidence level of 95% in step (5) means that the value of 950 times of each temporal frequency point falls below the distinguishing critical interface S [ Nmax, Mmax ] in 1000 times of statistical processes.
In this embodiment, the inverse transform S in step (7) is an inverse short-time fourier transform using a gaussian window function, and its expression is:
u and t are time, f is frequency, tau is time shift factor, i is imaginary number, and x (u) is magnetocardiogram signal after background noise is removed.
The method is suitable for magnetocardiogram signals measured by a high-temperature radio-frequency superconducting quantum interferometer system under any nonmagnetic shielding room condition. For convenience of explanation, the following description will be made by taking an example in which an ideal magnetocardiogram signal is interfered by a 10dB white gaussian noise background noise to remove background noise.
When in design, the method mainly comprises the following points: firstly, acquiring background noise of an instrument in real time under an environment without contacting a human body, and obtaining a two-dimensional time-frequency matrix after S transformation, wherein the two-dimensional time-frequency matrix is shown in a figure 2 (a); the acquisition is repeated 1000 times, a two-dimensional time-frequency matrix obtained after the background noise is subjected to S conversion for 1000 times can be obtained, the maximum value of each time-frequency point for 1000 times is calculated, and a curved surface shown in fig. 2(b) can be obtained. At 95% confidence, the maximum-value correspondence curve and 95% confidence line are obtained as shown in fig. 2(c), and the ratio of the maximum-value correspondence curve and 95% confidence line is calculated as shown in fig. 2 (d). When the magnetocardiogram signals are collected in real time under the condition of no magnetic shielding room, the magnetocardiogram signals are subjected to power line interference and environmental noise interference, the magnetocardiogram signals collected in real time are shown in a graph 3, a corresponding S-transform time-frequency graph is shown in a graph 4, according to a hypothesis test method, a three-dimensional time-frequency graph of the magnetocardiogram signals is compared with a 1000-time noise statistics confidence curved surface, and when the S-transform amplitude of the magnetocardiogram signals is smaller than a critical surface, the magnetocardiogram signals are regarded as background noise; otherwise, the signal is a magnetocardiogram signal, and the magnetocardiogram signal S smaller than the critical curved surface is converted into time-frequency matrix data 0; finally, the magnetocardiogram signal after the background noise is removed can be recovered by adopting the inverse S transform, as shown in FIG. 6. 99% of the background noise of the magnetocardiogram signal can be effectively removed.
The simulation results and analysis are integrated to show that the magnetocardiogram signal background noise removing method is feasible.
In summary, the method for removing the magnetocardiogram signal background noise S transform domain in the above embodiment adopts the S transform method to remove the background noise in the time-frequency domain, and the whole process can be performed by software without artificial redundant operation; the method is simple to realize, good in background noise removing effect and high in running speed, and extra hardware facilities do not need to be purchased; the method has strong universality, is suitable for removing background noise of magnetocardiogram signal data interfered by high signal-to-noise ratio under the condition of no shielding and no masking and accurately extracting magnetocardiogram signal parameters, and is worthy of popularization and application.
Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.
Claims (6)
1. A method for removing a magnetocardiogram signal background noise S transform domain is characterized by comprising the following steps:
s1: collecting noise data points
Acquiring N data points of background noise under the environment of an instrument in real time;
s2: s-transform noise
Converting the acquired N data point background noises into a two-dimensional time-frequency matrix by adopting an S conversion method;
s3: drawing maximum two-dimensional time-frequency curved surface
Continuously acquiring for multiple times, calculating a two-dimensional time-frequency matrix corresponding to each acquired signal to obtain multiple groups of two-dimensional time-frequency matrixes, calculating the maximum value of the time-frequency matrixes of each time-frequency point in the statistical process to obtain a background noise S-transform two-dimensional time-frequency matrix under the statistical condition, and drawing the matrix to obtain a maximum two-dimensional time-frequency curved surface;
s4: determining a discriminating critical interface
Taking a set confidence coefficient surface of the maximum value surface as a distinguishing critical interface of noise and magnetocardiogram signals;
s5: comparing the relation between the two-dimensional time-frequency matrix of magnetocardiogram signals and the critical interface
Transforming the acquired magnetocardiogram signals into a two-dimensional time-frequency matrix by adopting an S transformation method, and reserving the two-dimensional time-frequency matrix of the magnetocardiogram signals when the two-dimensional time-frequency matrix of the acquired magnetocardiogram signals is larger than a distinguishing critical interface; when the two-dimensional time-frequency matrix of the magnetocardiogram signal is smaller than or equal to the distinguishing critical interface, setting the two-dimensional time-frequency matrix of the magnetocardiogram signal to be 0, and modifying the two-dimensional time-frequency matrix of the magnetocardiogram signal;
s6: performing inverse S transform
And performing inverse S transformation on the modified magnetocardiogram signal two-dimensional time-frequency matrix to obtain the magnetocardiogram signal without background noise.
2. The method for removing the background noise S transform domain of the magnetocardiogram signal according to claim 1, wherein: in step S1, the environment of the instrument is the environment of the instrument before the magnetocardiogram signal is measured.
3. The method for removing the background noise S transform domain of the magnetocardiogram signal according to claim 1, wherein: in step S2, the transform S is a short-time fourier transform using a gaussian window function, and is expressed as:
wherein t is time, f is frequency, τ is time shift factor, i is an imaginary number, and x (t) represents real-time continuous noise.
4. The method for removing the background noise S transform domain of the magnetocardiogram signal according to claim 1, wherein: in step S4, a confidence level of 95% is set.
5. The method for removing the background noise S transform domain of the magnetocardiogram signal according to claim 1, wherein: in step S6, the inverse S transform is an inverse short-time fourier transform using a gaussian window function, and is expressed as:
wherein u and t are time, f is frequency, τ is time shift factor, i is imaginary number, and x (u) is magnetocardiogram signal with background noise removed.
6. A magnetocardiogram signal background noise S transform domain removing system, which removes the background noise in the magnetocardiogram signal by the background noise removing method according to any one of claims 1 to 5, comprising:
the noise acquisition module is used for acquiring N data points of background noise in real time under the environment where the instrument is located;
the noise transformation module is used for transforming the acquired background noise of the N data points into a two-dimensional time-frequency matrix by adopting an S transformation method;
the drawing module is used for continuously collecting the signals for multiple times, calculating a two-dimensional time-frequency matrix corresponding to each collected signal to obtain a plurality of groups of two-dimensional time-frequency matrixes, calculating the maximum value of each time-frequency matrix of a plurality of groups of time-frequency points in the statistical process to obtain a background noise S-transform two-dimensional time-frequency matrix under the statistical condition, and drawing the matrix to obtain a maximum value two-dimensional time-frequency curved surface;
the critical interface determining module is used for taking a set confidence coefficient curved surface of the maximum curved surface as a distinguishing critical interface of the noise and the magnetocardiogram signal;
the comparison module is used for converting the acquired magnetocardiogram signals into a two-dimensional time-frequency matrix by adopting an S conversion method, and reserving the two-dimensional time-frequency matrix of the magnetocardiogram signals when the two-dimensional time-frequency matrix of the acquired magnetocardiogram signals is larger than a distinguishing critical interface; when the two-dimensional time-frequency matrix of the magnetocardiogram signal is smaller than the distinguishing critical interface, setting the two-dimensional time-frequency matrix of the magnetocardiogram signal to be 0, and modifying the two-dimensional time-frequency matrix of the magnetocardiogram signal;
the inverse transformation module is used for carrying out inverse S transformation on the modified magnetocardiogram signal two-dimensional time-frequency matrix to obtain the magnetocardiogram signal without background noise;
the central processing module is used for sending instructions to other modules to complete related actions;
the noise acquisition module, the noise transformation module, the drawing module, the critical interface determination module, the comparison module and the inverse transformation module are all electrically connected with the central processing module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110202791.6A CN113017640B (en) | 2021-02-23 | 2021-02-23 | Method and system for removing S transform domain of background noise of magnetocardiogram signals |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110202791.6A CN113017640B (en) | 2021-02-23 | 2021-02-23 | Method and system for removing S transform domain of background noise of magnetocardiogram signals |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113017640A true CN113017640A (en) | 2021-06-25 |
| CN113017640B CN113017640B (en) | 2024-06-04 |
Family
ID=76460958
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110202791.6A Active CN113017640B (en) | 2021-02-23 | 2021-02-23 | Method and system for removing S transform domain of background noise of magnetocardiogram signals |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113017640B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102685053A (en) * | 2012-05-15 | 2012-09-19 | 北京航空航天大学 | Communication signal modulating and identifying method based on generalized S transformation |
| CN103190898A (en) * | 2013-04-23 | 2013-07-10 | 何怡刚 | Cardiac magnetic signal noise adaptive filtering and eliminating design method |
| CN108553086A (en) * | 2018-04-10 | 2018-09-21 | 深圳大学 | A kind of minimizing technology and system of function Near-infrared Brain imaging shallow-layer noise |
| CN109558845A (en) * | 2018-11-30 | 2019-04-02 | 合肥工业大学 | A kind of mcg-signals intelligence fitting optimization method based on ant group algorithm |
| CN109658944A (en) * | 2018-12-14 | 2019-04-19 | 中国电子科技集团公司第三研究所 | Helicopter acoustic signal Enhancement Method and device |
| CN109884481A (en) * | 2019-02-18 | 2019-06-14 | 上海金艺检测技术有限公司 | Shelf depreciation transient state voltage-to-ground scene noise evaluation method |
| CN111046791A (en) * | 2019-12-11 | 2020-04-21 | 河海大学 | Current signal filtering and denoising method based on generalized S transform containing variable factors |
-
2021
- 2021-02-23 CN CN202110202791.6A patent/CN113017640B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102685053A (en) * | 2012-05-15 | 2012-09-19 | 北京航空航天大学 | Communication signal modulating and identifying method based on generalized S transformation |
| CN103190898A (en) * | 2013-04-23 | 2013-07-10 | 何怡刚 | Cardiac magnetic signal noise adaptive filtering and eliminating design method |
| CN108553086A (en) * | 2018-04-10 | 2018-09-21 | 深圳大学 | A kind of minimizing technology and system of function Near-infrared Brain imaging shallow-layer noise |
| CN109558845A (en) * | 2018-11-30 | 2019-04-02 | 合肥工业大学 | A kind of mcg-signals intelligence fitting optimization method based on ant group algorithm |
| CN109658944A (en) * | 2018-12-14 | 2019-04-19 | 中国电子科技集团公司第三研究所 | Helicopter acoustic signal Enhancement Method and device |
| CN109884481A (en) * | 2019-02-18 | 2019-06-14 | 上海金艺检测技术有限公司 | Shelf depreciation transient state voltage-to-ground scene noise evaluation method |
| CN111046791A (en) * | 2019-12-11 | 2020-04-21 | 河海大学 | Current signal filtering and denoising method based on generalized S transform containing variable factors |
Non-Patent Citations (2)
| Title |
|---|
| SAMIT ARI.ET AL: ""ECG signal enhancement using S-Transform"", 《COMPUTERS IN BIOLOGY AND MEDICINE》, vol. 43, no. 6, pages 1 * |
| YUSHUN LIU.ET AL: ""An Ultrahigh Frequency Partial Discharge Signal De-Noising Method Based on a Generalized S-Transform and Module Time-Frequency Matrix"", 《SENSORS》, vol. 16, no. 9, pages 1 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113017640B (en) | 2024-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sangaiah et al. | An intelligent learning approach for improving ECG signal classification and arrhythmia analysis | |
| CN110338786B (en) | Epileptic discharge identification and classification method, system, device and medium | |
| CN109561841A (en) | For measuring the non-invasive methods and system of myocardial ischemia, narrow identification, positioning and the estimation of blood flow reserve score | |
| CN103190898B (en) | Cardiac magnetic signal noise adaptive filtering and eliminating design method | |
| CN109907752A (en) | A kind of cardiac diagnosis and monitoring method and system of the interference of removal motion artifacts and ecg characteristics detection | |
| Kumar et al. | Detection and classification of ECG noises using decomposition on mixed codebook for quality analysis | |
| Sun et al. | An improved morphological approach to background normalization of ECG signals | |
| CN108577834A (en) | A method of it is detected automatically for phase spike between epilepsy | |
| Vaneghi et al. | A comparative approach to ECG feature extraction methods | |
| CN106889981A (en) | A kind of intelligent terminal for extracting fetal heart frequency | |
| KR101510522B1 (en) | System for classification electrocardiogram signals | |
| Elbuni et al. | ECG parameter extraction algorithm using (DWTAE) algorithm | |
| Wu et al. | A novel method for beat-to-beat detection of ventricular late potentials | |
| Zhongliang | Fetal electrocardiogram extraction using blind source separation and empirical mode decomposition | |
| Malathi et al. | A high-performance low complex design and implementation of QRS detector using modified MaMeMi filter optimized with Mayfly optimization algorithm | |
| CN113017640B (en) | Method and system for removing S transform domain of background noise of magnetocardiogram signals | |
| Dliou et al. | Noised abnormal ECG signal analysis by combining EMD and Choi-Williams techniques | |
| Taralunga et al. | A new method for fetal electrocardiogram denoising using blind source separation and empirical mode decomposition | |
| CN114861706A (en) | Electrocardio identity recognition method based on quality evaluation and deep transfer learning | |
| Khandait et al. | ECG signal processing using classifier to analyses cardiovascular disease | |
| Dembrani et al. | Accurate detection of ECG signals in ECG monitoring systems by eliminating the motion artifacts and improving the signal quality using SSG filter with DBE | |
| Pretorius et al. | Feature extraction from ECG for classification by artificial neural networks | |
| Suresh et al. | System on chip (SOC) based cardiac monitoring system using Kalman filtering with fast fourier transform (FFT) signal analysis algorithm | |
| Mejhoudi et al. | Implementation and evaluation of ECG signal processing algorithms on embedded architectures | |
| Li et al. | Joint improved fast independent component analysis and singular value decomposition for fetal electrocardiogram 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |