Multi-channel magnetocardiogram automatic noise reduction superposition processing method and device

Abstract

The application provides a kind of multichannel magnetocardiogram signal automatic noise reduction superposition processing method and device, method includes: obtaining the initial detection data of magnetocardiogram signal;The initial detection data is independently component decomposition, obtains N independent components, wherein, at least one target magnetocardiogram signal component and multiple noise components are included in N independent components;The N independent components are screened, and the noise component screened out is removed;The screened independent component is divided into cardiac cycle, and the segmented period is superimposed;The application can realize the automatic identification of non-magnetocardiogram signal component in the independent component decomposed by ICA, and can automatically pick up the R peak in cardiac cycle, and the full channel data is automatically divided into cardiac cycle and superimposed, to effectively reduce the experience dependence of data processing, shorten processing time, further promote the clinical application of magnetocardiogram.

Description

Technical Field

[0001] This invention relates to the field of magnetic resonance signal processing technology, and in particular to a method and apparatus for automatic noise reduction and superposition processing of multi-channel magnetic resonance signals. Background Technology

[0002] Magnetocardiography (MCG) detects the extremely weak magnetic field generated by the activation current along myocardial fibers produced by ion flow across cell membranes during cardiac activity. With its superior sensitivity compared to electrocardiograms (ECG) in detecting ischemic myocardium under resting and stress conditions, MCG has been gradually applied in clinical trials. The amplitude of detectable MCG signals on the body surface is approximately on the order of tens of pT; in contrast, the Earth itself, as a giant magnet, has a magnetic field amplitude on the order of tens to hundreds of μT, about seven orders of magnitude greater than that of MCG. In urban or indoor environments, the magnetic field characteristics are even more complex. The movement of any ferromagnetic object, such as cars, subways, indoor elevators, and AC power, as well as the use of AC power, all cause magnetic field fluctuations in the environment. These magnetic noises and interferences pose a significant challenge to the detection and extraction of MCG signals.

[0003] Independent Component Analysis (ICA) is a statistical method used to separate multivariate signals into statistically independent sub-signals. Its basic idea is that the observed signal is a linear mixture of unknown underlying signals with non-Gaussian probability distributions and that these underlying signals are independent of each other. The goal of ICA is to find a linear transformation that makes the transformed signal as close as possible to the underlying signals, i.e., maximizing the independence of the signals.

[0004] ICA is a common blind source separation algorithm used to separate signals from different sources (such as magnetocardiogram signals and noise from different sources). However, determining whether the separated independent components are noise or artifacts often relies on empirical judgment, and the separated artifact components also need to be manually selected and removed. The complex data processing flow has brought difficulties to the clinical application of magnetocardiography. Summary of the Invention

[0005] This invention provides a method and apparatus for automatic noise reduction and superposition processing of multi-channel magnetic field signals, which solves the shortcomings of existing independent component analysis methods for separating magnetic field signals from noise, which rely on experience judgment and manual processing and limit clinical application, and realizes automatic identification of non-magnetic field signal components.

[0006] This invention provides an automatic noise reduction method for multi-channel magnetic resonance imaging (MRI) signals, comprising the following steps: Acquire initial detection data of the magnetic field signal; The initial detection data is decomposed into N independent components, which include at least one target magnetic signal component and multiple noise components. The N independent components are screened, and the screened noise components are removed; The selected independent components were segmented into cardiac cycles, and the segmented cycle segments were superimposed.

[0007] The automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging (MRI) signals provided by this invention includes the following steps for screening the N independent components: Based on a preset first threshold, N independent components are screened for the first time. The first threshold is set according to the proportion characteristics of the magnetic heart signal. Based on a preset second threshold, the independent components after the first screening are screened a second time. The second threshold is set according to the quasi-periodic characteristics of the magnetic heart signal, and the second threshold is less than the first threshold. Based on the preset third and fourth thresholds, the independent components after the second screening are screened a third time. The third and fourth thresholds are set according to the noise judgment margin, and the third threshold is less than the fourth threshold.

[0008] The automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging (MRI) signals provided by the present invention includes the step of performing a first screening of N independent components based on a preset first threshold, specifically comprising: Calculate the autocorrelation function for each of the N independent components; Integrating each of the autocorrelation functions over a preset delay interval, the formula is as follows: in, Let be the integral value of the autocorrelation function of the i-th independent component. , These are the start and end times of the preset delay interval, respectively. For the time series of the i-th independent component, Delay the time series of the i-th independent component Time series of moments It is the mean of the time series of the i-th independent component; It is the variance of the time series of the i-th independent component; Based on the proportion characteristics of the magnetic resonance signal, a first threshold is set for the integral value of each autocorrelation function. ; Independent components with integral values ​​less than or equal to the first threshold are removed, and after filtering, M independent components are obtained, where M is less than N; Filtering has minimum The independent components of the value are denoted as the k-th sequence, which serves as the template sequence for the second screening.

[0009] The automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging (MRI) signals provided by the present invention includes a step of performing a second screening of independent components after the first screening based on a preset second threshold, specifically comprising: For the independent components after the first screening, the cross-correlation function between the template sequence and each of the remaining independent components is calculated respectively; The maximum values ​​of the absolute values ​​of each cross-correlation function are obtained within a preset delay interval. : in, These are the start and end times of the preset delay interval, respectively. For the time series of the k-th independent component, Delay the time series of the i-th independent component Time series of moments It is the mean of the time series of the kth independent component; It is the mean of the time series of the i-th independent component; It is the variance of the time series of the k-th independent component; It is the variance of the time series of the i-th independent component; Based on the quasi-periodic characteristics of the magnetic heart signal, for each of the aforementioned maxima... Set a second threshold ; Remove maxima Independent components smaller than the second threshold are filtered to obtain Q independent components, where Q is less than M.

[0010] The automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging (MRI) signals provided by the present invention includes a step of performing a third screening on the independent components after the second screening based on preset third and fourth thresholds, specifically comprising: Based on the preset noise judgment margin, a third threshold is set. and the fourth threshold ,in, , ; For the independent components after the second screening, remove independent components that simultaneously meet the following two conditions: , .

[0011] The automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging (MRI) signals provided by this invention includes the following steps for cardiac cycle segmentation of the selected independent components: From the Q independent components obtained after the first screening, select... The sequence of the independent component with the smallest value is used as the template sequence; Based on the template sequence, the R-peak of the magnetic heart signal is detected to obtain the position coordinates of the R-peak; Based on the position coordinates of the R peak, the selected independent components are segmented into cardiac cycles within a preset time interval to obtain the segmented cycle segments.

[0012] The automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging (MRI) signals provided by this invention includes the following steps for superimposing segmented periodic data: Align and divide the periodic segments according to the position of the R peak; The aligned periodic segments are then superimposed.

[0013] The automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging (MRI) signals provided by the present invention further includes the following steps before performing independent component decomposition on the initial detection data: The initial detection data of the magnetic heart signal is bandpass filtered, and noise is removed from the bandpass-filtered magnetic heart signal using a single-point DFT filter.

[0014] The present invention also provides a multi-channel magnetic resonance imaging (MRI) signal automatic noise reduction and superposition processing device, comprising: The acquisition module is used to acquire the initial detection data of the magnetic field signal; The decomposition module is used to perform independent component decomposition on the initial detection data to obtain N independent components, wherein the N independent components include at least one target magnetic cardiac signal component and multiple noise components; A screening module is used to screen the N independent components and remove the screened noise components; The segmentation and overlay module is used to segment the cardiac cycle of the selected independent components and overlay the segmented cycle segments.

[0015] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging signals as described above.

[0016] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging signals as described above.

[0017] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging signals as described above.

[0018] The present invention provides a method and apparatus for automatic noise reduction and superposition processing of multi-channel magnetocardiogram (MCC) signals. The method involves acquiring initial detection data of MCC signals; performing independent component decomposition (ICA) on the initial detection data to obtain N independent components, wherein each of the N independent components includes at least one target MCC signal component and multiple noise components; filtering the N independent components and removing the filtered noise components; segmenting the selected independent components into cardiac cycles and superimposing the segmented cycle segments. This invention can automatically identify non-MCC signal components from the independent components obtained by ICA decomposition and automatically pick up the R-peak in the cardiac cycle, performing automated cardiac cycle segmentation and superposition on all-channel data. This effectively reduces the reliance on experience in data processing, shortens processing time, and further promotes the clinical application of magnetocardiography. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is one of the flowcharts of the automatic noise reduction and superposition processing method for multi-channel magnetic field signals provided by the present invention.

[0021] Figure 2 This is the data processing flow of the automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging signals provided by the present invention.

[0022] Figure 3 This is a schematic diagram of the initial multi-channel data provided by the present invention.

[0023] Figure 4 This is a schematic diagram of the artifact screening process provided by the present invention.

[0024] Figure 5 This is a schematic diagram of segmented point R-peak extraction provided by the present invention.

[0025] Figure 6 This is a schematic diagram of the multi-channel periodic segmented superposition result provided by the present invention.

[0026] Figure 7 This is a schematic diagram of the structure of the multi-channel magnetic heart signal automatic noise reduction and superposition processing device provided by the present invention.

[0027] Figure 8 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0029] The present invention will now be described in detail with reference to the accompanying drawings. The specific operation methods in the method embodiments can also be applied to the device embodiments or system embodiments. In the description of the present invention, unless otherwise stated, "at least one" includes one or more. "Multiple" refers to two or more. For example, at least one of A, B, and C includes: A existing alone, B existing alone, A and B existing simultaneously, A and C existing simultaneously, B and C existing simultaneously, and A, B, and C existing simultaneously. In the present invention, " / " means "or". For example, A / B can mean A or B. "And / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone.

[0030] The present invention will now be described in detail with reference to specific embodiments.

[0031] In some specific embodiments of the present invention, such as Figure 1 As shown, this solution provides an automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging (MRI) signals, including the following steps: Step 100: Obtain initial detection data of the magnetic field signal; Step 200: Perform independent component decomposition on the initial detection data to obtain N independent components, wherein the N independent components include at least one target magnetic cardiac signal component and multiple noise components; Step 300: Screen the N independent components and remove the screened noise components; Step 400: Perform cardiac cycle segmentation on the screened independent components and superimpose the segmented cycle segments.

[0032] It should be noted that existing noise component identification schemes often rely on experience-based judgment, cannot achieve automatic identification, and can only be manually selected and removed, which is cumbersome to operate. The complex data processing flow cannot meet the needs of clinical application of magnetocardiography.

[0033] Therefore, this invention achieves automatic identification of non-magnetic signal components in the independent components (ICs) decomposed from the initial detection data of magnetocardiogram signals by performing independent component decomposition, screening the separated independent components according to a preset threshold, and segmenting and superimposing the screened independent components by cardiac cycle. This effectively reduces the experience dependence of data processing, shortens processing time, and is expected to further promote the clinical application of magnetocardiography.

[0034] This invention provides an automated processing flow for multi-channel magnetic heart rate signals. Based on the independent component decomposition method, it proposes evaluation indicators for the decomposed independent components to automatically determine whether the component is an artifact, and automatically selects the R peak as the cardiac cycle segmentation point to achieve automatic noise reduction and superposition of magnetic heart rate data.

[0035] The following is an appendix Figure 2 The data processing flow of the automatic noise reduction and superposition processing method for multi-channel magnetocardiogram signals is described in detail.

[0036] Step 100: Obtain initial detection data of the magnetic field signal; In some possible embodiments of the present invention, such as Figure 2 As shown, before performing independent component decomposition on the initial detection data, the method further includes: The initial detection data of the magnetic heart signal is bandpass filtered, and noise is removed from the bandpass-filtered magnetic heart signal using a single-point DFT filter.

[0037] Specifically, this embodiment provides a preprocessing method for initial detection data before independent component decomposition. By performing bandpass filtering on the initial detection data, some noise is initially filtered out while retaining useful signals to the maximum extent. Furthermore, a single-point DFT filter is used to remove noise from the bandpass-filtered magnetocardiogram signal, thereby removing a large amount of power frequency magnetic interference contained in the environment.

[0038] It should be noted that the initial multi-channel data contained a large amount of noise, and each channel had different cardiac signal characteristics, such as... Figure 3 As shown.

[0039] In a possible implementation, a bandpass filter of 0.5-100 Hz is applied to initially remove some noise while preserving the useful signal to the maximum extent possible. Since the environment contains a large amount of power frequency magnetic interference with distinct frequency characteristics, an additional 50 Hz single-point DFT filter is used for removal.

[0040] Step 200: Perform independent component decomposition on the initial detection data to obtain N independent components, wherein the N independent components include at least one target magnetic cardiac signal component and multiple noise components; Furthermore, the filtered signal is subjected to data dimensionality reduction and compression using an independent component decomposition algorithm, decomposing it into N independent components (ICs) with a number much smaller than the original number of channels. Some ICs contain components of the target magnetic cardiomyocyte signal, while others are mainly noise components. The noise ICs need to be removed before the remaining ICs are recombined.

[0041] The identification of noise integrated circuits (ICs) is usually based on experience, making it difficult to automate the process of eliminating ICs. This invention performs a quantitative noise evaluation of ICs by calculating statistical parameters, and identifies components whose calculated parameters are higher than or equal to a given threshold as noise-dominant components. The specific calculation and judgment process is described below through concrete examples.

[0042] Step 300: Screen the N independent components and remove the screened noise components; In some possible embodiments of the present invention, such as Figure 4 As shown, the steps for screening the N independent components specifically include: Based on a preset first threshold, N independent components are screened for the first time. The first threshold is set according to the proportion characteristics of the magnetic heart signal. Based on a preset second threshold, the independent components after the first screening are screened a second time. The second threshold is set according to the quasi-periodic characteristics of the magnetic heart signal, and the second threshold is less than the first threshold. Based on the preset third and fourth thresholds, the independent components after the second screening are screened a third time. The third and fourth thresholds are set according to the noise judgment margin, and the third threshold is less than the fourth threshold.

[0043] Specifically, this embodiment provides an implementation method for screening the N independent components by performing three screenings on the initial detection data to remove noise components.

[0044] In some possible embodiments of the present invention, the step of performing a first screening of N independent components based on a preset first threshold specifically includes: Calculate the autocorrelation function for each of the N independent components; Integrating each of the autocorrelation functions over a preset delay interval, the formula is as follows: in, Let be the integral value of the autocorrelation function of the i-th independent component. , These are the start and end times of the preset delay interval, respectively. For the time series of the i-th independent component, Delay the time series of the i-th independent component Time series of moments It is the mean of the time series of the i-th independent component; It is the variance of the time series of the i-th independent component; Based on the proportion characteristics of the magnetic resonance signal, a first threshold is set for the integral value of each autocorrelation function. ; Independent components with integral values ​​less than or equal to the first threshold are removed, and after filtering, M independent components are obtained, where M is less than N; Filtering has minimum The independent components of the value are denoted as the k-th sequence, which serves as the template sequence for the second screening.

[0045] Specifically, this embodiment provides an implementation method for the first screening of N independent components, which achieves the first removal of artifacts by using an artifact evaluation method based on the interval integration of the autocorrelation function.

[0046] It is worth noting that the noise component mentioned in the embodiments of the present invention, namely the artifact component in the magnetic field signal, can be removed by screening.

[0047] In a possible embodiment, the autocorrelation function of each of the N independent components is calculated over a given time delay interval ( to The integration is performed within a certain range. The integrated IC of the dominant cardiac component typically shows a small value, while the integrated values ​​of some noise components are very large. Therefore, a relatively large threshold is set for this screening. Initial screening identified all those that did not meet the requirements. The component is the artifact component. Simultaneously, the channel with the smallest integral value is selected as the "template" sequence for subsequent calculations, denoted as the k-th independent component IC.

[0048] (1).

[0049] In some possible embodiments of the present invention, the step of performing a second screening of the independent components after the first screening based on a preset second threshold specifically includes: For the independent components after the first screening, the cross-correlation function between the template sequence and each of the remaining independent components is calculated respectively; The maximum values ​​of the absolute values ​​of each cross-correlation function are obtained within a preset delay interval. : (2) in, These are the start and end times of the preset delay interval, respectively. For the time series of the k-th independent component, Delay the time series of the i-th independent component Time series of moments It is the mean of the time series of the kth independent component; It is the mean of the time series of the i-th independent component; It is the variance of the time series of the k-th independent component; It is the variance of the time series of the i-th independent component; Based on the quasi-periodic characteristics of the magnetic heart signal, for each of the aforementioned maxima... Set a second threshold ; Remove maxima Independent components smaller than the second threshold are filtered to obtain Q independent components, where Q is less than M.

[0050] Specifically, this embodiment provides an implementation method for a second screening of N independent components, which achieves the second removal of artifacts by using the cross-correlation function interval extreme value feature extraction method.

[0051] In a possible embodiment, the cross-correlation function between the k-th IC and the remaining components is calculated, and the maximum value of the absolute value of each function is taken within a given delay interval. This maximum value between the i-th component and k is denoted as... (0< <1), due to the quasi-periodic nature of the heartbeat itself, the component with the cardiac cycle will show a greater presence in this screening process. Components that do not exhibit quasi-periodic characteristics are very small. The process sets a small threshold. Perform a filter to remove all items that do not meet the requirements. The component is identified as an artifact component.

[0052] .

[0053] In some possible embodiments of the present invention, the step of performing a third screening on the independent components after the second screening based on preset third and fourth thresholds specifically includes: Based on the preset noise judgment margin, a third threshold is set. and the fourth threshold ,in, , ; For the independent components after the second screening, remove independent components that simultaneously meet the following two conditions: (3), (4).

[0054] Specifically, this embodiment provides an implementation method for a third screening of N independent components. The noise components are removed for the third time by using an artifact discrimination method based on interval dual thresholds of autocorrelation and cross-correlation functions.

[0055] Specifically, all remaining ICs are further screened by setting thresholds. and And there are , All ICs that simultaneously meet the following two conditions are identified as artifact components.

[0056] This process can be used to adjust the margin for artifact detection, including but not limited to: like , In this case, the third screening does not add additional artifact IC. Under this condition, the potential signal components in the original data can be preserved to the maximum extent.

[0057] like In this case, it is a further convergence of the second screening, and under this condition, the third screening is a further elimination of ICs with non-quasi-periodic characteristics.

[0058] like In this case, it is a further refinement of the first screening, and the third screening under this condition is a further elimination of IC components with high noise power.

[0059] The ICs ultimately identified as artifact components are the union of the three screenings, meaning all IC components selected in the three screenings are discarded. The remaining IC components are then reconstructed to obtain the denoised multi-channel magnetocardiogram signal, thus completing the automated noise reduction process.

[0060] This invention uses the interval integral of the autocorrelation function to evaluate the independent components separated by ICA (Independent Correlation Analysis). Based on the difference in statistical characteristics between noise and magnetocardiogram (MCA) signals, it automatically performs a preliminary screening of independent components whose main component is noise. This invention then uses the interval maximum value of the cross-correlation function to screen non-MCA components, and automatically performs a second screening based on the quasi-periodicity of the MCA signal itself. Finally, it uses a combined interval integral of the autocorrelation function and the interval maximum value of the cross-correlation function to apply a threshold for a second judgment. The result of the third screening is obtained when both threshold conditions are met.

[0061] Step 400: Perform cardiac cycle segmentation on the screened independent components and superimpose the segmented cycle segments.

[0062] In this step, the noise-reduced signal needs to be segmented into cardiac cycles to achieve cross-cycle superposition of cardiac signals. This superposition process usually takes the R peak of the cardiac cycle as the segmentation point, and takes a period of time before and after it. This time needs to meet two conditions: 1. It should encompass the entire cardiac cycle (P-QRS-T) signal, and 2. It should ensure that the cardiac cycles do not overlap.

[0063] In some possible embodiments of the present invention, the step of segmenting the cardiac cycle of the screened independent components specifically includes: From the Q independent components obtained after the first screening, select... The sequence of the independent component with the smallest value is used as the template sequence; Based on the template sequence, the R-peak of the magnetic heart signal is detected to obtain the position coordinates of the R-peak; Based on the position coordinates of the R peak, the selected independent components are segmented into cardiac cycles within a preset time interval to obtain the segmented cycle segments.

[0064] Specifically, this embodiment provides an implementation method for cardiac cycle segmentation of screened independent components, which achieves automatic selection of R-peak channels and extraction of R-peaks by selecting template sequences.

[0065] Specifically, such as Figure 5 As shown, the present invention has a minimum value in the kth component after the aforementioned independent component decomposition. The template sequence of the value is used to detect the R peak. This selection method can avoid the problem that the R peak may not exist in the original channel, and eliminates the manual selection process of the R peak channel detection channel, so as to realize the automation of the entire processing process.

[0066] This invention selects the independent component with the highest signal-to-noise ratio as the R-peak extraction channel, identifies the R-peak position, and segments the original signal. This avoids the problem of indistinct R-peaks in some channels of the magnetic resonance imaging (MRI) waveform, requiring manual selection.

[0067] In some possible embodiments of the present invention, the step of superimposing the segmented periodic segments specifically includes: Align and divide the periodic segments according to the position of the R peak; The aligned periodic segments are then superimposed.

[0068] Specifically, R-peak detection methods include, but are not limited to: The extreme value detection of the sequence determines the point as an extreme value if the amplitude of the detected data point is greater than or less than or equal to that of two adjacent points. At the same time, a minimum detection interval for extreme values ​​is set to avoid detecting P, T and other peak signals within the same period. Wavelet decomposition extracts the frequency band where the R-peak energy spectrum is located, the corresponding wavelet layer is taken for recombination, and the absolute value or square is taken to extract the peak point.

[0069] Furthermore, after detecting the location of the R-peak, the denoised multi-channel data is segmented into cardiac cycles. The segmented cycles are then aligned with the R-peak positions and superimposed to obtain a multi-channel single-cycle butterfly diagram, as shown below. Figure 6 As shown. This completes the entire automated data noise reduction and overlay process.

[0070] The automatic noise reduction and superposition processing method for multi-channel magnetic field signals provided by this invention can realize the automatic noise reduction and superposition of multi-channel magnetic field signals to obtain a single cardiac cycle signal with a high signal-to-noise ratio. Through three screenings of artifact components, the output can be adjusted by threshold to obtain a signal with a higher signal-to-noise ratio or to retain more detailed information. By using diversified features to determine artifacts, the false artifact rate can be reduced and useful signals can be retained to the greatest extent.

[0071] The following describes the multi-channel magnetocardiogram (MCC) signal automatic noise reduction and superposition processing device provided by the present invention. The multi-channel MCC signal automatic noise reduction and superposition processing device described below and the multi-channel MCC signal automatic noise reduction and superposition processing method described above can be referred to in correspondence.

[0072] In some specific embodiments of the present invention, such as Figure 7 As shown, this solution provides a multi-channel magnetic resonance imaging (MRI) signal automatic noise reduction and superposition processing device, comprising: The acquisition module 71 is used to acquire the initial detection data of the magnetic field signal; The decomposition module 72 is used to perform independent component decomposition on the initial detection data to obtain N independent components, wherein the N independent components include at least one target magnetic signal component and multiple noise components; The screening module 73 is used to screen the N independent components and remove the screened noise components; The segmentation and overlay module 74 is used to segment the cardiac cycle of the selected independent components and overlay the segmented cycle segments.

[0073] The multi-channel magnetocardiogram (MCC) signal automatic noise reduction and superposition processing method and apparatus provided in this invention achieve automated noise reduction and superposition processing of MCC signals through an automated processing flow and noise reduction method. It can automatically identify non-MCC signal components in the independent components decomposed by ICA and automatically pick up the R-peak in the cardiac cycle, performing automated cardiac cycle segmentation and superposition on the full-channel data. This effectively reduces the reliance on experience in data processing, shortens processing time, and is expected to further advance the clinical application of magnetocardiography.

[0074] Figure 8 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 8As shown, the electronic device may include a processor 810, a communications interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communications interface 820, and the memory 830 communicate with each other via the communication bus 840. The processor 810 can call logic instructions in the memory 830 to execute a multi-channel magnetocardiogram (MCC) signal automatic noise reduction and superposition processing method. This method includes: acquiring initial detection data of the MCC signal; performing independent component decomposition on the initial detection data to obtain N independent components, wherein the N independent components include at least one target MCC signal component and multiple noise components; filtering the N independent components and removing the filtered noise components; segmenting the filtered independent components into cardiac cycles and superimposing the segmented cycle segments.

[0075] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0076] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the multi-channel magnetic signal automatic noise reduction and superposition processing method provided by the above methods. The method includes: acquiring initial detection data of magnetic signals; performing independent component decomposition on the initial detection data to obtain N independent components, wherein the N independent components include at least one target magnetic signal component and multiple noise components; screening the N independent components and removing the screened noise components; performing cardiac cycle segmentation on the screened independent components and superimposing the segmented cycle segments.

[0077] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the multi-channel magnetic signal automatic noise reduction and superposition processing method provided by the above methods. The method includes: acquiring initial detection data of magnetic signals; performing independent component decomposition on the initial detection data to obtain N independent components, wherein the N independent components include at least one target magnetic signal component and multiple noise components; screening the N independent components and removing the screened noise components; performing cardiac cycle segmentation on the screened independent components and superimposing the segmented cycle segments.

[0078] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0079] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0080] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for automatic noise reduction and superposition processing of multi-channel magnetic resonance imaging (MRI) signals, characterized in that, include: Acquire initial detection data of the magnetic field signal; The initial detection data is decomposed into N independent components, which include at least one target magnetic signal component and multiple noise components. The N independent components are screened, and the screened noise components are removed; The selected independent components were segmented into cardiac cycles, and the segmented cycle segments were superimposed.

2. The automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging signals according to claim 1, characterized in that, The steps for screening the N independent components specifically include: Based on a preset first threshold, N independent components are screened for the first time. The first threshold is set according to the proportion characteristics of the magnetic heart signal. Based on a preset second threshold, the independent components after the first screening are screened a second time. The second threshold is set according to the quasi-periodic characteristics of the magnetic heart signal, and the second threshold is less than the first threshold. Based on the preset third and fourth thresholds, the independent components after the second screening are screened a third time. The third and fourth thresholds are set according to the noise judgment margin, and the third threshold is less than the fourth threshold.

3. The automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging signals according to claim 2, characterized in that, The first screening step, based on a preset first threshold, involves screening N independent components for the first time. Specifically, this includes: Calculate the autocorrelation function for each of the N independent components; Integrating each of the autocorrelation functions over a preset delay interval, the formula is as follows: ， in, Let be the integral value of the autocorrelation function of the i-th independent component. , These are the start and end times of the preset delay interval, respectively. For the time series of the i-th independent component, Delay the time series of the i-th independent component Time series of moments It is the mean of the time series of the i-th independent component; It is the variance of the time series of the i-th independent component; Based on the proportion characteristics of the magnetic resonance signal, a first threshold is set for the integral value of each autocorrelation function. ; Independent components with integral values ​​less than or equal to the first threshold are removed, and after filtering, M independent components are obtained, where M is less than N; Filtering has minimum The independent components of the value are denoted as the k-th sequence, which serves as the template sequence for the second screening.

4. The automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging signals according to claim 3, characterized in that, The step of performing a second screening of the independent components after the first screening, based on a preset second threshold, specifically includes: For the independent components after the first screening, the cross-correlation function between the template sequence and each of the remaining independent components is calculated respectively; The maximum values ​​of the absolute values ​​of each cross-correlation function are obtained within a preset delay interval. : ， in, These are the start and end times of the preset delay interval, respectively. For the time series of the k-th independent component, Delay the time series of the i-th independent component Time series of moments It is the mean of the time series of the kth independent component; It is the mean of the time series of the i-th independent component; It is the variance of the time series of the k-th independent component; It is the variance of the time series of the i-th independent component; Based on the quasi-periodic characteristics of the magnetic heart signal, for each of the aforementioned maxima... Set a second threshold ; Remove maxima Independent components smaller than the second threshold are filtered to obtain Q independent components, where Q is less than M.

5. The automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging signals according to claim 4, characterized in that, The steps for performing a third screening of the independent components after the second screening, based on preset third and fourth thresholds, specifically include: Based on the preset noise judgment margin, a third threshold is set. and the fourth threshold ,in, , ; For the independent components after the second screening, remove independent components that simultaneously meet the following two conditions: ， 。 6. The method for automatic noise reduction and superposition processing of multi-channel magnetic resonance imaging signals according to any one of claims 3-5, characterized in that, The steps for cardiac cycle segmentation of the screened independent components specifically include: From the Q independent components obtained after the first screening, select... The sequence of the independent component with the smallest value is used as the template sequence; Based on the template sequence, the R-peak of the magnetic heart signal is detected to obtain the position coordinates of the R-peak; Based on the position coordinates of the R peak, the selected independent components are segmented into cardiac cycles within a preset time interval to obtain the segmented cycle segments.

7. The automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging signals according to claim 6, characterized in that, The steps for superimposing the segmented periodic segments specifically include: Align and divide the periodic segments according to the position of the R peak; The aligned periodic segments are then superimposed.

8. The method for automatic noise reduction and superposition processing of multi-channel magnetic resonance imaging signals according to claim 1, characterized in that, Before performing independent component decomposition on the initial detection data, the method further includes: The initial detection data of the magnetic heart signal is bandpass filtered, and noise is removed from the bandpass-filtered magnetic heart signal using a single-point DFT filter.

9. A multi-channel magnetic resonance imaging (MRI) signal automatic noise reduction and superposition processing device, characterized in that, include: The acquisition module is used to acquire the initial detection data of the magnetic field signal; The decomposition module is used to perform independent component decomposition on the initial detection data to obtain N independent components, wherein the N independent components include at least one target magnetic cardiac signal component and multiple noise components; A screening module is used to screen the N independent components and remove the screened noise components; The segmentation and overlay module is used to segment the cardiac cycle of the selected independent components and overlay the segmented cycle segments.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the automatic noise reduction and superposition processing method for multi-channel magnetic resonance imaging signals as described in any one of claims 1 to 8.

Images