Method and device for inhibiting tooth metal artifacts based on optical pumping magnetometer-magnetocardiogram data

Abstract

The invention discloses a tooth metal artifact suppression method and device based on optical pumping magnetometer-magnetocardiogram data, and relates to the technical field of magnetocardiogram artifact signal preprocessing, and the method comprises the steps: carrying out pre-whitening processing on a data matrix of the optical pumping magnetometer-magnetocardiogram data to generate a pre-whitening matrix; generating a delay correlation matrix of the data matrix according to the pre-whitening matrix and a pre-generated delay sequence; performing joint diagonalization iterative approximation processing on the delay correlation matrix to generate a decomposition matrix of the data matrix; and inhibiting tooth metal artifacts in the optical pumping magnetometer-magnetocardiogram data according to the decomposition matrix and a predetermined time-frequency domain threshold. According to the method, the quality of MCG data is improved, the limitation of using MCG by a patient with teeth provided with metal materials is reduced, and the measurement range of MCG clinical application is expanded.

Description

Technical Field

[0001] This application belongs to the technical field of bio-signal processing, especially the technical field of preprocessing of magnetocardiogram artifact signals, and specifically relates to a method and device for suppressing tooth metal artifacts in optically pumped magnetometer-magnetocardiogram data. Background Art

[0002] Magnetocardiography (MCG) detection is a non-invasive and non-contact method for monitoring cardiac electrophysiology. Compared with electrocardiogram (ECG), it has higher spatio-temporal resolution and can be used to complement and verify ECG measurements. The diagnosis of heart diseases has always been the focus of clinical research. Magnetocardiography detection devices based on superconducting quantum interference devices (SQUID-MCG) have been continuously developed and expanded in medical clinical diagnosis, such as coronary artery diseases, cardiac ischemia, arrhythmia, and fetal heart rate pre-detection.

[0003] However, in recent years, due to the disadvantages of high maintenance cost, immobility, and relatively long distance from the measurement target of SQUID devices, optically pumped magnetometers (OPMs) have emerged in the field of vision of researchers with their advantages of low cost, wearable, and close-range measurement, which is expected to achieve more clinical application progress in the future.

[0004] However, OPM-MCG is also vulnerable to the influence of the environment and various artifacts. For background environmental noise, since magnetocardiogram signals are relatively large (in the pT range), simple band-pass filtering, notch filtering, empirical mode decomposition, or FastICA and other methods can be used for preprocessing to obtain better MCG signals. However, for the strong magnetic metal interference caused by implantable devices (such as tooth fixation devices made of metal materials, metal dentures, and even implantable cardioverter defibrillators (ICDs)) in many heart patients, traditional methods cannot effectively suppress metal artifacts, which also limits the effective clinical application of MCG measurement and diagnosis in this part of patients. Therefore, the suppression of metal artifacts in magnetocardiogram signals is the research focus of the clinical promotion of OPM-MCG.

[0005] Currently, the research on removing magnetocardiogram metal artifacts is very limited. The related methods currently disclosed are as follows:

[0006] (1) Independent component analysis has been used to remove strong magnetic interference generated by the steel wire for suturing the sternum after heart surgery;

[0007] (2) Temporal Decorrelation source Separation (TDSEP) was once used to extract cardiac signals from measurements interfered by ICD. However, in most cases, only clean QRS complexes could be obtained, and the remaining characteristic waves were still submerged in noise.

[0008] (3) Based on the beat-by-beat denoising method and the principal component analysis denoising method, for the strong magnetic interference of implantable permanent cardiac pacemakers and stainless steel wires, less distortion can be obtained compared with denoising methods such as band-pass filtering, wavelet, and ensemble empirical mode decomposition. Moreover, the beat-by-beat dynamic characteristics of the cardiac time series are retained, but it depends on the selection of the optimal QRS channel.

[0009] There are very few methods for suppressing dental metal materials in magnetocardiogram measurements, and no relevant literature records have been found. For this part of the research, the following possible solutions have been recorded in the literature related to magnetoencephalogram, including:

[0010] (1) The tSSS algorithm has been verified to be able to suppress strong magnetic interference from orthodontic materials. However, this algorithm has not been optimized on OPM devices at present. Due to its strong dependence on the geometric calibration and parameter settings of the sensor array, its application on the magnetocardiogram planar array is limited at present.

[0011] (2) Blind source separation algorithms, including the Fast Independent Component Analysis (FastICA) based on higher-order statistics, whose premise of successful convergence is the non-Gaussianity assumption of independent components; and the Information Maximization (Infomax) algorithm, which can ensure the accurate separation of signal sources and interference sources only when the statistical independence of each component of the measurement signal is the highest. Therefore, for the separation of low-frequency, sub-Gaussian, and respiration-modulated dental metal artifacts, these two algorithms cannot achieve satisfactory results. Summary of the Invention

[0012] An object of the present invention is to provide a method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data, so as to solve the technical problem that the existing methods for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data have poor separation effects on low-frequency, sub-Gaussian, and respiration-modulated dental metal artifacts.

[0013] Another object of the present invention is to provide a device for suppressing tooth metal artifacts in optically pumped magnetometer-cardiac magnetic data. Still another object of the present invention is to provide an electronic device, which includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the above-mentioned method for suppressing tooth metal artifacts in optically pumped magnetometer-cardiac magnetic data are implemented. Still another object of the present invention is to provide a readable medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned method for suppressing tooth metal artifacts in optically pumped magnetometer-cardiac magnetic data are implemented.

[0014] To solve the technical problems in the background art of the present application, the present invention provides the following technical solutions:

[0015] In a first aspect, the present invention provides a method for suppressing tooth metal artifacts in optically pumped magnetometer-cardiac magnetic data, including:

[0016] Performing pre-whitening processing on the data matrix of the optically pumped magnetometer-cardiac magnetic data to generate a pre-whitening matrix;

[0017] Generating a delay correlation matrix of the data matrix according to the pre-whitening matrix and a pre-generated delay sequence;

[0018] Performing joint diagonalization iterative approximation processing on the delay correlation matrix to generate a decomposition matrix of the data matrix;

[0019] Suppressing tooth metal artifacts in the optically pumped magnetometer-cardiac magnetic data according to the decomposition matrix and a pre-determined time-frequency domain threshold.

[0020] In some embodiments of the present invention, performing joint diagonalization iterative approximation processing on the delay correlation matrix to generate a decomposition matrix of the data matrix includes:

[0021] Determining the non-diagonal elements of all covariance matrices in the delay correlation matrix;

[0022] Generating a plurality of two-dimensional matrices according to the non-diagonal elements;

[0023] Vectorizing the plurality of two-dimensional matrices and combining the vectorized results of the plurality of two-dimensional matrices to generate a vector matrix;

[0024] Determining the eigenvector with the largest eigenvalue in the vector matrix;

[0025] Performing joint diagonalization iterative approximation processing on the delay correlation matrix through the eigenvector to generate a decomposition matrix of the data matrix.

[0026] In some embodiments of the present invention, performing a joint diagonalization iterative approximation process on the delay correlation matrix through the eigenvector includes:

[0027] Generating a rotation matrix according to the vector elements of the eigenvector;

[0028] Performing a joint diagonalization iterative approximation process on the delay correlation matrix through the rotation matrix.

[0029] In some embodiments of the present invention, pre-whitening the data matrix of the optically pumped magnetometer-magnetocardiogram data to generate a pre-whitening matrix includes:

[0030] Generating a zero-time-delay correlation matrix of the data matrix;

[0031] Performing singular value decomposition on the zero-time-delay correlation matrix to generate the pre-whitening matrix.

[0032] In some embodiments of the present invention, the signal components in the decomposition matrix include: metal artifact interference components, QRS wave source components, TP wave source components, and unknown interference components.

[0033] In some embodiments of the present invention, suppressing dental metal artifacts in the optically pumped magnetometer-magnetocardiogram data according to the decomposition matrix and a pre-determined time-frequency domain threshold includes:

[0034] Decomposing multiple components in the optically pumped magnetometer-magnetocardiogram data through the decomposition matrix;

[0035] Determining the total time-frequency domain value of the multiple components;

[0036] According to the pre-determined time-frequency domain thresholds of the metal artifact interference components, QRS wave source components, TP wave source components, and unknown interference components respectively, determining the respective positions of the metal artifact interference components, QRS wave source components, TP wave source components, and unknown interference components in the total time-frequency domain value to suppress the dental metal artifacts.

[0037] In a second aspect, the present invention provides a dental metal artifact suppression device for optically pumped magnetometer-magnetocardiogram data, and the device includes:

[0038] A pre-whitening matrix generation module, configured to pre-whiten the data matrix of the optically pumped magnetometer-magnetocardiogram data to generate a pre-whitening matrix;

[0039] A delay correlation matrix generation module, configured to generate a delay correlation matrix of the data matrix according to the pre-whitening matrix and a pre-generated time delay sequence;

[0040] A decomposition matrix generation module, configured to perform joint diagonalization iterative approximation processing on the delay correlation matrix to generate a decomposition matrix of the data matrix;

[0041] A dental metal artifact suppression module, configured to suppress dental metal artifacts in the optically pumped magnetometer-magnetocardiogram data according to the decomposition matrix and a predetermined time-frequency domain threshold.

[0042] In some embodiments of the present invention, the decomposition matrix generation module includes:

[0043] A diagonal element determination unit, configured to determine the non-diagonal elements of all covariance matrices in the delay correlation matrix;

[0044] A two-dimensional matrix generation unit, configured to generate a plurality of two-dimensional matrices according to the non-diagonal elements;

[0045] A vector matrix generation unit, configured to vectorize the plurality of two-dimensional matrices and combine the vectorized results of the plurality of two-dimensional matrices to generate a vector matrix;

[0046] An eigenvector determination unit, configured to determine an eigenvector with the largest eigenvalue in the vector matrix;

[0047] A decomposition matrix generation unit, configured to perform joint diagonalization iterative approximation processing on the delay correlation matrix through the eigenvector to generate a decomposition matrix of the data matrix.

[0048] In some embodiments of the present invention, the decomposition matrix generation unit includes:

[0049] A rotation matrix generation unit, configured to generate a rotation matrix according to the vector elements of the eigenvector;

[0050] A diagonalization iterative processing unit, configured to perform joint diagonalization iterative approximation processing on the delay correlation matrix through the rotation matrix.

[0051] In some embodiments of the present invention, the pre-whitening matrix generation module includes:

[0052] A zero-time-delay correlation matrix generation unit, configured to generate a zero-time-delay correlation matrix of the data matrix;

[0053] A pre-whitening matrix generation unit, configured to perform singular value decomposition on the zero-time-delay correlation matrix to generate the pre-whitening matrix.

[0054] In some embodiments of the present invention, the signal components in the decomposition matrix include: metal artifact interference components, QRS wave source components, TP wave source components, and unknown interference components.

[0055] In some embodiments of the present invention, the dental metal artifact suppression module includes:

[0056] A magnetocardiogram data decomposition unit for decomposing multiple components in the optically pumped magnetometer - magnetocardiogram data through the decomposition matrix;

[0057] A total time - frequency domain value determination unit for determining the total time - frequency domain value of the multiple components;

[0058] A dental metal artifact suppression unit for determining the positions of the metal artifact interference component, QRS wave source component, TP wave source component, and unknown interference component in the total time - frequency domain value according to the pre - determined time - frequency domain thresholds of the metal artifact interference component, QRS wave source component, TP wave source component, and unknown interference component respectively, so as to suppress the dental metal artifacts.

[0059] In a third aspect, the present invention provides a computer program product, including a computer program / instructions, which when executed by a processor, implements the steps of a method for suppressing dental metal artifacts in optically pumped magnetometer - magnetocardiogram data.

[0060] In a fourth aspect, the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the steps of a method for suppressing dental metal artifacts in optically pumped magnetometer - magnetocardiogram data.

[0061] In a fifth aspect, the present invention provides a computer - readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of a method for suppressing dental metal artifacts in optically pumped magnetometer - magnetocardiogram data.

[0062] As can be seen from the above description, the embodiments of the present invention provide a method and device for suppressing dental metal artifacts in optically pumped magnetometer - magnetocardiogram data. The corresponding method includes: First, perform pre - whitening processing on the data matrix of the optically pumped magnetometer - magnetocardiogram data to generate a pre - whitening matrix; then, generate a delay - correlation matrix of the data matrix according to the pre - whitening matrix and the pre - generated delay sequence; perform joint diagonalization iterative approximation processing on the delay - correlation matrix to generate a decomposition matrix of the data matrix; suppress the dental metal artifacts in the optically pumped magnetometer - magnetocardiogram data according to the decomposition matrix and the pre - determined time - frequency domain thresholds.

[0063] A method and device for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram (OPM-MCG) data provided by the present invention, through sufficient simulation experiments and OPM-MCG experiments of actual acquisition, using various evaluation indexes, verify that compared with Fast Independent Component Analysis (FastICA) and Information Maximization (Infomax), it has superior metal artifact suppression effect and can reconstruct MCG signals without distortion. In summary, the present invention not only improves the quality of MCG data, but also reduces the limitations of using MCG for patients with dental metal materials, and expands the measurement range of MCG clinical applications. Description of the Drawings

[0064] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0065] Figure 1 Schematic flowchart of a method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram (OPM-MCG) data in an embodiment of the present invention;

[0066] Figure 2 Schematic flowchart of step 300 of a method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram (OPM-MCG) data in an embodiment of the present invention;

[0067] Figure 3 Schematic flowchart of step 305 of a method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram (OPM-MCG) data in an embodiment of the present invention;

[0068] Figure 4 Schematic flowchart of step 100 of a method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram (OPM-MCG) data in an embodiment of the present invention;

[0069] Figure 5 Schematic flowchart of step 100 of a method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram (OPM-MCG) data in an embodiment of the present invention;

[0070] Figure 6 Schematic flowchart of a method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram (OPM-MCG) data in the specific embodiment of the present invention;

[0071] Figure 7Mind map of a method for suppressing dental metal artifacts in optically pumped magnetometer-cardiac magnetic data in a specific embodiment of the present invention;

[0072] Figure 8 Block diagram of an apparatus for suppressing dental metal artifacts in optically pumped magnetometer-cardiac magnetic data in an embodiment of the present invention;

[0073] Figure 9 Block diagram of the decomposition matrix generation module 30 in an embodiment of the present invention;

[0074] Figure 10 Block diagram of the decomposition matrix generation unit 30e in an embodiment of the present invention;

[0075] Figure 11 Block diagram of the pre-whitening matrix generation module 10 in an embodiment of the present invention;

[0076] Figure 12 Block diagram of the dental metal artifact suppression module 40 in an embodiment of the present invention;

[0077] Figure 13 Structural schematic diagram of an electronic device in an embodiment of the present invention. Specific embodiments

[0078] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0079] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0080] It should be noted that the terms "including" and "having" in the description, claims and above-mentioned drawings of this application, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices. Without conflict, the embodiments in this application and the features in the embodiments can be combined with each other. The following will describe this application in detail with reference to the drawings and in combination with the embodiments.

[0081] Magnetocardiogram plays an increasingly important role in the clinical diagnosis of heart diseases. However, magnetocardiogram signals are vulnerable to interference from the environment and various metal artifacts. Among them, many patients have metal materials in their teeth, and these strong magnetic interferences greatly distort the magnetocardiogram signals, which limits the effective application of MCG for this part of patients. Based on this, embodiments of the present invention provide a specific implementation manner of a method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data. See Figure 1 , and the method specifically includes the following contents:

[0082] Step 100: Perform pre-whitening processing on the data matrix of the optically pumped magnetometer-magnetocardiogram data to generate a pre-whitening matrix;

[0083] Step 200: Generate a delay correlation matrix of the data matrix according to the pre-whitening matrix and a pre-generated time delay sequence;

[0084] Step 300: Perform joint diagonalization iterative approximation processing on the delay correlation matrix to generate a decomposition matrix of the data matrix;

[0085] Step 400: Suppress the dental metal artifacts in the optically pumped magnetometer-magnetocardiogram data according to the decomposition matrix and a pre-determined time-frequency domain threshold.

[0086] As can be seen from the above description, embodiments of the present invention provide a method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data, which can effectively separate the ultra-low frequency artifact components caused by metal materials on teeth, and then realize the automatic screening of metal artifacts and other unknown interference source components according to the time-domain and frequency-domain characteristics of metal artifact components, QRS wave components, TP wave components and other unknown interference source components.

[0087] For step 100, first, use OPMs to collect MCG data of coronary heart disease patients with medical metal materials on their teeth in a shielded environment. Divide the data of different subjects into a training set and a test set. Additionally, collect MCG data of healthy people without metal materials on their teeth for simulation experiment verification. Then, use a separation method based on second-order time-domain features to pre-whiten the OPM-MCG multi-channel measurement data containing magnetocardiogram signals and various artifact signals by calculating the whitening matrix Q.

[0088] For step 200, sequentially select the delay values in the delay sequence and calculate them with the pre-whitening matrix to generate a delay correlation matrix.

[0089] For step 300, according to the principle of minimizing the sum of the squares of the non-diagonal elements of all time-delay covariance matrices in the delay correlation matrix (using a matrix group), iteratively calculate the optimal post-whitening separation matrix. Combining the whitening matrix and the separation matrix, the measured signals can be separated into components, obtaining a signal group containing metal artifacts, magnetocardiogram signals, and other unknown interference signals.

[0090] For step 400, perform component separation on the subject training set data. Combine expert experience to pre-define various mixed signal components and analyze their time-domain and frequency-domain characteristics. Construct relevant indicators based on their differences and summarize the empirical trends to guide the thresholds for subsequent identification of multiple types of components. Use the frequency-domain energy ratio and time-domain information quantity characteristics, and according to the trends analyzed and summarized from the training set, finally achieve automatic screening and extraction of metal artifact interference components and other unknown source interference components.

[0091] In some embodiments of the present invention, refer to Figure 2 , step 300 includes:

[0092] Step 301: Determine the non-diagonal elements of all covariance matrices in the delay correlation matrix;

[0093] Step 302: Generate multiple two-dimensional matrices according to the non-diagonal elements;

[0094] In steps 301 to 302, define the post-whitening decomposition matrix as V. Assume that the delay correlation matrix (R s (τ)) of the source signals satisfies the relationship: V T R z (τ)V = R s (τ). Initialize the matrix V = I, and perform joint diagonalization iterative approximation on the post-whitening delay correlation matrix R z (τ i )(R z (τ i ) is the delay correlation matrix of the whitening matrix Z(t)) to obtain V, and find the matrix group R zAll non - diagonal elements of the covariance matrices in (τ) the positions of i and j at , and extract the matrix group R z Four elements corresponding to the positions (i, i), (i, j), (j, j), and (j, i) in all covariance matrices in (τ) form K 2×2 two - dimensional matrices A.

[0095] Step 303: Vectorize the multiple two - dimensional matrices, and combine the vectorized results of the multiple two - dimensional matrices to generate a vector matrix;

[0096] Step 304: Determine the eigenvector with the largest eigenvalue in the vector matrix;

[0097] In steps 303 to 305, the two - dimensional matrix A is separately [A 11 -A 22 , A 12 +A 21 T vectorized, where A 11 , A 22 are the two diagonal elements of the two - dimensional matrix A, A 12 and A 21 are non - diagonal elements. All k two - dimensional matrices are vectorized and combined to obtain a vector matrix G of dimension K×2. Perform singular value decomposition on G T G to obtain the eigenvector v = [v1, v2] of the largest eigenvalue T .

[0098] Step 305: Perform joint diagonalization iterative approximation processing on the delay - correlation matrix through the eigenvector to generate a decomposition matrix of the data matrix.

[0099] In some embodiments of the present invention, referring to Figure 3 , step 305 includes:

[0100] Step 3051: Generate a rotation matrix according to the vector elements of the eigenvector;

[0101] Step 3052: Perform joint diagonalization iterative approximation processing on the delay - correlation matrix through the rotation matrix.

[0102] In steps 3051 to 3052, calculate the rotation matrix P according to the v vector elements step The calculation formula is where cosθ represents cosine calculation, sinθ is sine calculation, and θ is the rotation angle, Use the rotation matrix P step to perform iteration on the covariance matrix group R z (τ) ​Calculate and iterate V

[0103] Loop the above steps until the covariance matrix group R z (τ) satisfies the condition to obtain the optimal whitened decomposition matrix V. The final decomposition matrix is W = V T Q.

[0104] In some embodiments of the present invention, refer to Figure 4 , step 100 includes:

[0105] Step 101: Generate the zero-time-delay correlation matrix of the data matrix;

[0106] Step 102: Perform singular value decomposition on the zero-time-delay correlation matrix to generate the pre-whitening matrix.

[0107] In step 101 and step 102, perform pre-whitening processing on the data matrix x(t) measured by OPM-MCG, calculate the zero-time-delay correlation matrix R xx (0) = E[x(t)x T (t)], perform singular value decomposition on R xx (0) as R xx (0) = UΛU T , where Λ is a diagonal matrix, the main diagonal elements are the singular values of R xx (0), and the rest of the elements are 0. Take as the pre-whitening matrix, and the processed signal is z(t) = qx(t).

[0108] In some embodiments of the present invention, the signal components in the decomposition matrix include: metal artifact interference components, QRS wave source components, TP wave source components, and unknown interference components.

[0109] Specifically, pre-define the decomposed signal components as metal artifact interference components, QRS wave source components, TP wave source components, and other unknown interference components.

[0110] In some embodiments of the present invention, refer to Figure 5 , step 400 includes:

[0111] Step 401: Decompose multiple components from the optically pumped magnetometer-magnetocardiogram data through the decomposition matrix;

[0112] Step 402: Determine the total time-frequency domain value of the multiple components;

[0113] Step 403: Determine the positions of the metal artifact interference component, QRS wave source component, TP wave source component, and unknown interference component in the total time-frequency domain value according to the pre-determined time-frequency domain thresholds of the metal artifact interference component, QRS wave source component, TP wave source component, and unknown interference component respectively, so as to suppress the dental metal artifact.

[0114] For the time-frequency domain threshold, perform component separation on the data of the test training set, pre-define various mixed signal components in combination with expert experience, and conduct time-domain and frequency-domain feature analysis. Construct relevant indicators based on their differences, and summarize the empirical trends to guide the thresholds for subsequent identification of multiple types of components.

[0115] Analyzing the frequency-domain distributions of different components, it can be seen that the metal artifact has a relatively high proportion of absolute power below 4 Hz in the low frequency, and the TP wave has a relatively high proportion of absolute power in the range of 1 - 8 Hz. Set the corresponding time-frequency domain indicators and indicators to extract the metal artifact component and the TP wave source component. For the component with a larger APR1 index, it tends to be judged as the metal artifact component, and for the component with a larger APR2 index, it tends to be judged as the TP wave component.

[0116] At the same time, it is noted that the frequency-domain distribution of other unknown source interference components is similar to that of the TP wave component. The TP wave component cannot be effectively separated from the APR2 index alone, and the mutual information index in the time domain needs to be used. Considering that the TP wave component generally has strong regularity, the sample entropy SE is selected as the further separation index. The larger the value, the higher the disorder and complexity of the time series of the separated component, and the greater the possibility of being a noise interference source. On the contrary, it tends to be judged as the TP wave.

[0117] Using an evaluation method that combines the frequency-domain energy ratio and the time-domain information content, set multiple thresholds according to the summarized empirical trends to automatically screen and extract the metal artifact interference component and other unknown source interference components. First, calculate the frequency-domain indicators N (t)} for the N separated source components {s1(t), s2(t), …, s and

[0118] Sort the APR1 results in descending order, and select the top 30% of the components to form component list A. List A mainly contains metal artifact components, TP wave components, and other unknown interference source components.

[0119] According to the summarized characteristic trends, define the components in list A that satisfy APR1 > 95% as the metal artifact interference component I a , and mark the components in list A that satisfy APR2 > 70% as the TP wave components.

[0120] After excluding the selected metal artifact interference and the components corresponding to the TP wave in List A, name the set of the remaining components as List A res , which contains some TP components and interference components from other unknown sources.

[0121] For each separated component s i (t), calculate the sample entropy SE, sort the resulting values in descending order, and select the top 30% of the components to form List B.

[0122] According to the inducted characteristic trends, mark the intersection components of the components in List B and the components in List A res as potential interference components I1, and mark the top 10% of the components in List B as other interference components I2. Define the union of the above I1 and I2 as other unknown interference source components I b = I1 ∪ I2. The finally screened corresponding components of metal artifacts and other unknown interference sources are I a ∪I b .

[0123] As can be seen from the above description, the embodiment of the present invention provides a method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data. Use OPMs to collect MCG data of multiple coronary heart disease patients with medical metal materials on their teeth in a shielding environment, divide the data into a training set and a test set, and collect MCG data of healthy people without metal materials on their teeth for subsequent simulation experiment verification.

[0124] After pre-whitening the measured data matrix, use the second-order time-domain characteristics and K time-delay sequences to calculate the time-delay covariance matrix of the data, and obtain a group of time-delay covariance matrices R z (τ). Using the principle of minimizing the sum of the squares of the non-diagonal elements of all time-delay covariance matrices in the matrix group R z (τ), iteratively calculate the optimal post-whitening separation matrix V. Combining the whitening matrix and the separation matrix can separate the components of the measured signal to obtain a signal group containing metal artifacts, magnetocardiogram signals, and other unknown interference signals.

[0125] To further illustrate the solution, the present invention also provides a specific implementation of a method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data. See Figure 6 and Figure 7 , which specifically includes the following content.

[0126] The technical problem to be solved by the present invention is as follows: Aiming at the problem that the multi-channel OPM-MCG measurement data is easily interfered by metal artifacts generated by implantable metal sources such as dental braces, vagus nerve stimulators, and deep brain stimulators, a dental metal artifact suppression method based on OPM-MCG is proposed to improve the quality of MCG data, reduce the limitations of MCG use for patients with teeth containing metal materials, and expand the measurement range of MCG clinical applications.

[0127] The technical solution adopted by the present invention is as follows: Using a separation method based on second-order time-domain features, for the OPM-MCG multi-channel measurement data containing magnetocardiogram signals and various artifact signals, first calculate the whitening matrix Q for pre-whitening, and then calculate the time-delay covariance matrix group. Use joint diagonalization to iteratively obtain the optimal separation matrix V. Finally, the measurement signal x(t) can be separated through W = V^TQ to obtain a signal group separated from the magnetocardiogram source signal, metal artifacts, and other undefined noise interferences. The process is as shown in Figure 7 the dashed box in part (a) of the figure. Specifically, the specific implementation manner of a method for suppressing dental metal artifacts in a magnetometer-magnetocardiogram data provided by the present invention includes the following steps:

[0128] S1: Use OPMs to collect MCG data of coronary heart disease patients with medical metal materials on their teeth in a shielding environment, divide the data of different subjects into a training set and a test set, and collect MCG data of healthy people without metal materials on their teeth for simulation experiment verification.

[0129] Specifically, 32-channel OPMs were used to collect MCG data of 30 coronary heart disease patients with metal materials on their teeth in a semi-open shielding barrel. Although the metal content of their teeth is different, they all belong to medical metal materials. In addition, MCG data of 4 healthy subjects without any metal materials on their teeth were collected as clean data for building a simulation experiment.

[0130] The selected subjects are in the age range of 25 to 50 years old. During the measurement process, the subjects are required to maintain normal breathing and swallowing, and at the same time, try to keep their bodies still. The collection time is about 3 minutes, and both the collection and stimulation devices are outside the shielding barrel. Among the MCG data of 30 patients, 20 sets of data are set as the training data set, and the other 10 sets are set as the test data set. The training data set is used for subsequent parameter analysis and threshold debugging, and the test data is used for algorithm testing and verification.

[0131] The OPM-MCG experimental equipment consists of a semi-open shielding barrel, an OPMs array, a non-magnetic bed, and data acquisition equipment. The residual magnetic field at the center of the semi-open shielding barrel is less than 50 nT. The OPMs probe uses the QZFMGen-2 magnetometer developed by QuSpin Company, with a sensitivity of <15 fT / Hz1 / 2 (typically 7-10 fT / Hz1 / 2) in the 3-100 Hz frequency band, a dynamic range of ±5 nT, and measurement modes of only Z-axis / only Y-axis / double Z and Y-axis (simultaneously). The sensor size is 12.4×16.6×24.4 mm, and the total power is 5W. The data acquisition and control chassis realizes the functions of sensor data acquisition and control, and finally presents the data in the computer acquisition system.

[0132] The overall panel of the OPM sensor array is constructed by 3D printing. During actual measurement, the sensor is inserted into the slot and measured in the direction perpendicular to the Z-axis of the sensor array. The array is located directly above the subject's heart, about 1-2 cm vertically from the chest. During measurement, the subject is pushed into the barrel, and after magnetic compensation of the sensor, the acquisition starts.

[0133] S2: Using the separation method based on second-order time-domain features, for the OPM-MCG multi-channel measurement data containing magnetocardiogram signals and various artifact signals, first calculate the whitening matrix Q for pre-whitening, and then calculate the time-delay covariance matrix group. Use joint diagonalization to iteratively obtain the optimal separation matrix V. Finally, the measurement signal x(t) can be separated through W = V T Q to obtain a signal group separated from the magnetocardiogram source signal, metal artifacts, and other undefined noise interferences.

[0134] Perform pre-whitening processing on the data matrix x(t) of OPM-MCG measurement, and calculate the zero-time-delay correlation matrix R xx (0) = E[x(t)x T (t)], perform singular value decomposition on R xx (0) = UΛU xx (0), where Λ is a diagonal matrix, the main diagonal elements are the singular values of R T (0), and the rest of the elements are 0. Take xx as the pre-whitening matrix, and the processed signal is z(t) = Qx(t). as the pre-whitening matrix, and the processed signal is z(t) = Qx(t).

[0135] Set the time delay sequence τ = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 14, 16, 18, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 120, 140, 160, 180, 200, 220, 240, 260, 280, 300, 350, 400, 450, 500, 550, 600, 650, 700, 750, 800, 850, 900, 950, 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000] ms. Select the time delay value τ in turn and calculate the delay correlation matrix R z (τ i ) = E[z(t)z T (t + τ i )].

[0136] Define the whitened decomposition matrix as V. Assume that the source signal correlation matrix satisfies the relationship: V T R z (τ)V = R s (τ). Initialize the matrix V = I. For R z (τ i ) perform joint diagonalization iterative approximation to obtain V. Find the positions of i and j at all non-diagonal elements of the covariance matrices in the matrix group R z (τ), and extract the elements at the four positions (i, i), (i, j), (j, j), (j, i) in all covariance matrices in the matrix group R to form K 2×2 two-dimensional matrices A of order two-dimensional matrix A. z (τ).

[0137] Vectorize the two-dimensional matrix A separately as [A 11 - A 22 , A 12 + A 21 , where A T , A 11 are the two diagonal elements of the two-dimensional matrix A, and A 22 , A 12 , A 21 are the non-diagonal elements. Vectorize and combine the k two-dimensional matrices to obtain a vector matrix G of dimension K×2. Perform singular value decomposition on G T G to obtain the eigenvector v = [v1, v2] of the largest eigenvalue T .

[0138] Calculate the rotation matrix P according to the elements of the v vector step The calculation formula is where cosθ represents cosine calculation, sinθ represents sine calculation, and θ is the rotation angle. Use the rotation matrix P step to iterate over the covariance matrix group R z (τ) Calculate and iterate V

[0139] Loop the above steps until the covariance matrix group R z (τ) satisfies the condition to obtain the optimal whitened decomposition matrix V. The final decomposition matrix is W = V T Q, and the estimated source components after separation are

[0140] S3: Perform component separation on the subject training set data, pre-define various mixed signal components in combination with expert experience, perform time-domain and frequency-domain feature analysis, construct relevant indicators based on their differences, and summarize the empirical trends to guide the thresholds for identifying multiple types of components in the follow-up.

[0141] Here, the decomposed signal components are pre-defined as metal artifact interference components, QRS wave source components, TP wave source components, and other unknown interference components.

[0142] Analyzing the frequency-domain distributions of different components shows that the metal artifacts have a relatively high proportion of absolute power below 4 Hz in the low frequency, and the TP wave has a relatively high proportion of absolute power in the range of 1 - 8 Hz. Set the corresponding time-frequency domain indicators and the indicator to extract the metal artifact components and TP wave source components. For the components with a larger indicator APR1, they tend to be judged as metal artifact components, and for the components with a larger indicator APR2, they tend to be judged as TP wave components.

[0143] At the same time, it is noted that the frequency-domain distributions of other unknown source interference components are similar to those of the TP wave components. Relying solely on the indicator APR2 cannot effectively separate the TP wave components, and the mutual information indicator in the time domain needs to be used. Considering that the TP wave components generally have strong regularity, the sample entropy SE is selected as the further separation indicator. The larger the value, the higher the disorder and complexity of the time series of the separated components, and the greater the possibility of being a noise interference source. Conversely, it tends to be judged as a TP wave.

[0144] S4: Use an evaluation method that combines the frequency-domain energy ratio and the time-domain information content. According to the summarized empirical trends, set multiple thresholds to automatically screen and extract the metal artifact interference components and other unknown source interference components.

[0145] See Figure 7 as shown in the dashed box in part (b) ofN (t) Calculate frequency domain indicators and Sort the APR1 results in descending order, and select the top 30% of the components to form component list A. List A mainly contains metal artifact components, TP wave components, and other unknown interference source components. According to the inductive characteristic trend, the components in list A that satisfy APR1>95% are defined as metal artifact interference components I a , and the components in list A that satisfy APR2>70% are marked as TP wave components. After excluding the selected metal artifact interference and the components corresponding to the TP wave in list A, the set of the remaining components is named list A res , which contains some TP components and other unknown source interference components. For each separated component s i (t) Calculate the sample entropy SE, sort the result values in descending order, and select the top 30% of the components to form list B. According to the inductive characteristic trend, the intersection components of the components in list B and the components in list A res are marked as potential interference components I1, and the top 10% of the components in list B are marked as other interference components I2. The union of the above I1 and I2 is defined as other unknown interference source components I b =I1∪I2. The finally selected components corresponding to metal artifacts and other unknown interference sources are I a ∪I b .

[0146] S5: Extract the part of the channels with the strongest energy in the artifact frequency band from the patient's MCG data as the channels of the main metal artifact interference. Use the linear regression method to project the artifact components onto the signals of all sensor channels, and superimpose them on the healthy and metal artifact-free subject data to obtain simulated metal artifact-contaminated MCG data for simulation verification.

[0147] Collect the MCG data of healthy subjects without metal materials on their teeth, and perform band-pass filtering from 1 - 43.5 Hz and notch filtering at 50 Hz to remove power frequency on the metal-free OPM-MCG acquisition data to obtain clean magnetocardiogram simulation data. Collect the MCG data of patients with metal materials on their teeth, perform band-pass filtering from 0.2 - 4 Hz on it, calculate the root mean square value for each channel, and select the top 10% of the channels with the root mean square value as the channels of the main metal artifact interference.

[0148] Use the linear regression method to calculate the propagation matrix of the artifact channels according to the formula , where C XO is the cross-correlation matrix between the main artifact channel matrix O and the MCG measurement data matrix X contaminated by the metal substances on the teeth, is the inverse of the autocorrelation matrix of the main artifact channel matrix O. The selected artifact components are projected into all sensor channel signals using the propagation matrix G to simulate metal artifact interference data.

[0149] By superimposing the clean MCG data with the metal artifact interference data, the final simulated metal artifact contaminated MCG data can be obtained.

[0150] Furthermore, the simulation experiment mainly analyzes the effect of the metal artifact suppression method on the data of healthy heart subjects. According to the model formula S = C + HY, two sets of simulation data are generated for two experimental subjects. Where S represents the simulation data interfered by metal artifacts, C is the normal data without metal artifact interference in the resting state, Y is the reasonably selected metal artifact interference component signal, and H is the artifact propagation matrix calculated based on the linear regression method. This formula shows the simulation data generation process. The artifact components extracted from the experimental data channels severely interfered by artifacts are used to obtain the propagation matrix of the artifact components in the sensor measurement channels through the linear regression method. The metal artifact components are projected into all channels to simulate the real conduction process of the strong magnetic interference artifacts brought by metal braces in the sensors. Finally, the MCG simulation data containing metal artifacts is obtained by superimposing with the MCG data of a healthy heart without metal artifact interference for subsequent evaluation of the algorithm effect.

[0151] For the evaluation of the performance of the metal artifact separation and extraction algorithm, multiple time-frequency domain indicators are used for each data channel, including the signal-to-noise ratio (SNR) of the average MCG data, the root mean square error (RMSE) of the average MCG data, and the RMSE of the MCG data. Since the time of one heartbeat is about 750 - 900 ms, after detecting the R wave peak of the heartbeat, 250 ms before the R wave peak and 500 ms after the R wave peak are taken as one heartbeat epoch for MCG averaging.

[0152] The formula for calculating the SNR of the average MCG data is:

[0153]

[0154] where, σ QRS is the standard deviation of the Q - R - S band, and σ ST is the standard deviation of the S - T band.

[0155] The formula for calculating RMSE1 of the average MCG data is:

[0156]

[0157] where, X avereconst (t) is the artifact suppression reconstruction signal averaged over the j - th MCG channel, and X avesim(t) is the clean signal averaged over the j-th MCG channel of the simulation. M is the sampling point of the MCG average data, and M = 751 ms in the present invention.

[0158] The calculation formula of RMSE2 for MCG average data is:

[0159]

[0160] Among them, X reconst (t) is the artifact-suppressed reconstructed signal of the j-th MCG channel, and X sim (t) is the clean signal of the j-th MCG channel of the simulation. M = 60000 ms in the present invention.

[0161] The larger the SNR, the smaller RMSE1 and RMSE2, the better the artifact suppression effect of the algorithm, and the higher the data quality.

[0162] The proposed method, the FastICA algorithm, and Infomax are respectively used to perform component separation on 4 groups of simulated artifact-interfered MCG data. FastICA and Infomax use the ICA function, the number of components is set to the number of data channels, and other parameters are set to the default values.

[0163] The above evaluation indexes are used to evaluate the processed data of 4 subjects. The results are shown in Table 1. It can be seen that the SNR of the data after suppressing artifacts using the proposed method is closest to the SNR of the simulated clean MCG data. In addition, the RMSE1 and RMSE2 values obtained by this method are also the smallest, which is better than the traditional FastICA algorithm, indicating that the MCG signal reconstructed by the proposed method is closest to the simulated clean MCG signal.

[0164] Table 1 Evaluation of data quality after suppressing artifacts in simulated data using different methods

[0165]

[0166]

[0167] Four cases are randomly selected from the test data set of 10 coronary heart disease patients for artifact suppression algorithm testing. The parameter settings of the proposed SOBI algorithm based on the improved delay matrix, the FastICA algorithm, and the Infomax algorithm are the same as those in the simulation. Using SNR as the evaluation index, the results are shown in Table 2. It can be seen that the signal-to-noise ratio obtained by the proposed method of the present invention is the highest, which is consistent with the simulation experiment results.

[0168] Table 2 Evaluation of data quality after suppressing artifacts in real data using different methods

[0169]

[0170] From the indicators in the two groups of data, namely the metal artifact interference simulation data and the real data, it can be seen that the newly proposed method achieves the best improvement in indicators compared with other methods, effectively removing the metal artifact components. It can be seen that the present invention can effectively separate and identify the metal artifact components and other unknown source interference components in the magnetocardiogram signal. After removal, the interference of metal artifacts on the magnetoencephalogram image signal is greatly reduced, the signal quality is improved, and it is convenient for the subsequent clinical application and popularization of the OPM-MCG magnetocardiogram device.

[0171] From the above description, the specific implementation of the present invention provides a method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data. Using OPMs, MCG data of multiple coronary heart disease patients with medical metal materials on their teeth are collected in a shielded environment. The data is divided into a training set and a test set. In addition, MCG data of healthy people without metal materials on their teeth are collected for subsequent simulation experiment verification.

[0172] After pre-whitening the measurement data matrix, the time-delay covariance matrix of the data is calculated using the second-order time-domain characteristics and K time-delay sequences, obtaining a group of time-delay covariance matrices R z (τ). Using the principle of minimizing the sum of the squares of the non-diagonal elements of all time-delay covariance matrices in the matrix group R z (τ), the optimal post-whitening separation matrix V is iteratively calculated. Combining the whitening matrix and the separation matrix, the measurement signal can be component-separated to obtain a signal group containing metal artifacts, magnetocardiogram signals, and other unknown interference signals.

[0173] Perform component separation on the training set data of the subjects. Combine expert experience to pre-define and analyze the time-domain and frequency-domain characteristics of various mixed signal components. Construct relevant indicators based on their differences, and summarize the empirical trends to guide the thresholds for identifying multiple types of components in the future (this example will provide specific threshold standards for reference indicators).

[0174] Using the frequency-domain energy ratio and time-domain information content characteristics, and setting multiple thresholds according to the trends analyzed and summarized from the training set, finally realize the automatic screening and extraction of metal artifact interference components and other unknown source interference components.

[0175] The advantages of the present invention compared with the prior art are as follows:

[0176] (1) Currently, there are very few methods for suppressing metal materials on teeth in magnetocardiogram measurement, and no relevant literature records have been found. The present invention can effectively suppress metal artifact components and reconstruct the MCG signal with almost no distortion, expanding the scanning population of MCG and further promoting the clinical application of OPM-MCG;

[0177] (2)Both the FastICA and Infomax separation algorithms have different defects. FastICA is difficult to effectively separate the metal artifacts with high Gaussianity; Infomax uses information entropy as the optimization objective, and there are residual metal artifact components in the separated signals. The present invention combines the second-order time-domain covariance characteristics of the signals, selects the time-delay sequence suitable for metal artifacts to calculate the covariance matrix group, iteratively optimizes the separation matrix from the covariance matrix group, and effectively separates the metal artifacts;

[0178] (3)The present invention does not need to rely on a reference channel, and only considers the characteristic differences of different source signals themselves for separation, and has strong extensibility. As the signal characteristics of different artifact components are understood, more indicators and thresholds can be continuously refined and embedded for classification;

[0179] (4)Metal artifact separation and identification is a complex and difficult signal processing field. Although various methods have obvious disadvantages, they still have certain effects, indicating that the current separation and identification of metal artifacts cannot be the optimal solution with a single method. The best solution is a series of complementary methods. Therefore, the present invention provides a new method for calculating the separation matrix using the second-order time-domain characteristics of metal artifacts, combining the low-frequency power ratio in the frequency domain and the time-domain information content for artifact identification, constructing indicators and setting threshold experience based on the measured data itself, separating metal artifacts and other unknown interferences from the magnetocardiogram signals, and automatically identifying them. Experiments prove that the combined new method is more effective for metal artifacts than other methods.

[0180] Based on the same inventive concept, the embodiment of the present application also provides a device for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data, which can be used to implement the method described in the above embodiment, such as the following embodiment. Since the principle of solving problems by the device for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data is similar to that of the method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data, the implementation of the device for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data can refer to the implementation of the method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data, and the repeated parts will not be described again. As used below, the term "unit" or "module" can be a combination of software and / or hardware that can implement a predetermined function. Although the systems described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0181] The embodiment of the present invention provides a specific implementation manner of a device for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data that can implement the method for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data. Refer to Figure 8 , a device for suppressing dental metal artifacts in optically pumped magnetometer-magnetocardiogram data specifically includes the following contents:

[0182] A pre-whitening matrix generation module 10 is configured to perform pre-whitening processing on the data matrix of the optically pumped magnetometer-cardiac magnetic data to generate a pre-whitening matrix;

[0183] A delay correlation matrix generation module 20 is configured to generate a delay correlation matrix of the data matrix according to the pre-whitening matrix and a pre-generated time delay sequence;

[0184] A decomposition matrix generation module 30 is configured to perform joint diagonalization iterative approximation processing on the delay correlation matrix to generate a decomposition matrix of the data matrix;

[0185] A dental metal artifact suppression module 40 is configured to suppress dental metal artifacts in the optically pumped magnetometer-cardiac magnetic data according to the decomposition matrix and a pre-determined time-frequency domain threshold.

[0186] In some embodiments of the present invention, referring to Figure 9 , the decomposition matrix generation module 30 includes:

[0187] A diagonal element determination unit 30a is configured to determine non-diagonal elements of all covariance matrices in the delay correlation matrix;

[0188] A two-dimensional matrix generation unit 30b is configured to generate a plurality of two-dimensional matrices according to the non-diagonal elements;

[0189] A vector matrix generation unit 30c is configured to vectorize the plurality of two-dimensional matrices and combine the vectorization results of the plurality of two-dimensional matrices to generate a vector matrix;

[0190] An eigenvector determination unit 30d is configured to determine an eigenvector with the largest eigenvalue in the vector matrix;

[0191] A decomposition matrix generation unit 30e is configured to perform joint diagonalization iterative approximation processing on the delay correlation matrix through the eigenvector to generate a decomposition matrix of the data matrix.

[0192] In some embodiments of the present invention, referring to Figure 10 , the decomposition matrix generation unit 30e includes:

[0193] A rotation matrix generation unit 30e1 is configured to generate a rotation matrix according to vector elements of the eigenvector;

[0194] A diagonalization iterative processing unit 30e2 is configured to perform joint diagonalization iterative approximation processing on the delay correlation matrix through the rotation matrix.

[0195] In some embodiments of the present invention, referring to Figure 11 , the pre-whitening matrix generation module 10 includes:

[0196] A zero-time-delay correlation matrix generation unit 10a for generating a zero-time-delay correlation matrix of the data matrix;

[0197] A pre-whitening matrix generation unit 10b for performing singular value decomposition on the zero-time-delay correlation matrix to generate the pre-whitening matrix.

[0198] In some embodiments of the present invention, the signal components in the decomposition matrix include: a metal artifact interference component, a QRS wave source component, a TP wave source component, and an unknown interference component.

[0199] In some embodiments of the present invention, referring to Figure 12 , the dental metal artifact suppression module 40 includes:

[0200] A magnetocardiogram data decomposition unit 40a for decomposing multiple components in the optically pumped magnetometer-magnetocardiogram data through the decomposition matrix;

[0201] A total time-frequency domain value determination unit 40b for determining the total time-frequency domain value of the multiple components;

[0202] A dental metal artifact suppression unit 40c for determining the positions of the metal artifact interference component, the QRS wave source component, the TP wave source component, and the unknown interference component in the total time-frequency domain value according to the pre-determined time-frequency domain thresholds of the metal artifact interference component, the QRS wave source component, the TP wave source component, and the unknown interference component respectively, so as to suppress the dental metal artifacts.

[0203] As can be seen from the above description, an embodiment of the present invention provides a dental metal artifact suppression device for optically pumped magnetometer-magnetocardiogram data, including: a plurality of classification result generation modules for classifying change operations in a system change plan according to a pre-generated change plan classification model to generate a plurality of classification results; a plurality of word slot result generation modules for selecting corresponding word slots from a pre-generated word slot set according to the current classification result and filling the current classification result into the selected word slots to generate a plurality of word slot results; a plurality of keyword extraction modules for extracting a plurality of keywords corresponding to the codes of system change operations; and a change operation review module for reviewing whether the system change operation conforms to the system change plan according to the vector spaces corresponding to the plurality of word slot results and the vector spaces of the plurality of keywords.

[0204] An embodiment of the present application also provides a specific implementation manner of an electronic device capable of implementing all steps in the dental metal artifact suppression method for optically pumped magnetometer-magnetocardiogram data in the above embodiments. Referring to Figure 13 , the electronic device specifically includes the following contents:

[0205] A processor 1201, a memory 1202, a communication interface 1203, and a bus 1204;

[0206] Among them, the processor 1201, the memory 1202, and the communication interface 1203 communicate with each other through the bus 1204; the communication interface 1203 is used to implement information transmission between related devices such as server-side devices and client-side devices;

[0207] The processor 1201 is used to call the computer program in the memory 1202. When the processor executes the computer program, all steps in the method for suppressing dental metal artifacts in the optically pumped magnetometer-cardiac magnetic data in the above embodiments are implemented. For example, when the processor executes the computer program, the following steps are implemented:

[0208] Step 100: Perform pre-whitening processing on the data matrix of the optically pumped magnetometer-cardiac magnetic data to generate a pre-whitening matrix;

[0209] Step 200: Generate a delay correlation matrix of the data matrix according to the pre-whitening matrix and a pre-generated delay sequence;

[0210] Step 300: Perform joint diagonalization iterative approximation processing on the delay correlation matrix to generate a decomposition matrix of the data matrix;

[0211] Step 400: Suppress dental metal artifacts in the optically pumped magnetometer-cardiac magnetic data according to the decomposition matrix and a pre-determined time-frequency domain threshold.

[0212] An embodiment of the present application also provides a computer-readable storage medium capable of implementing all steps in the method for suppressing dental metal artifacts in the optically pumped magnetometer-cardiac magnetic data in the above embodiments. A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, all steps in the method for suppressing dental metal artifacts in the optically pumped magnetometer-cardiac magnetic data in the above embodiments are implemented. For example, when the processor executes the computer program, the following steps are implemented:

[0213] Step 100: Perform pre-whitening processing on the data matrix of the optically pumped magnetometer-cardiac magnetic data to generate a pre-whitening matrix;

[0214] Step 200: Generate a delay correlation matrix of the data matrix according to the pre-whitening matrix and a pre-generated delay sequence;

[0215] Step 300: Perform joint diagonalization iterative approximation processing on the delay correlation matrix to generate a decomposition matrix of the data matrix;

[0216] Step 400: Suppress the dental metal artifacts in the optically pumped magnetometer-cardiac magnetic data according to the decomposition matrix and a predetermined time-frequency domain threshold.

[0217] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the hardware + program type embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiments.

[0218] The specific embodiments of this specification are described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the figures do not necessarily require the specific order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0219] Although this application provides method operation steps such as in the embodiments or flowcharts, based on routine or non-creative labor, there can be more or fewer operation steps. The step order listed in the embodiments is only one way among many step execution orders and does not represent the only execution order. When the actual device or user terminal product is executed, it can be executed in the order shown in the embodiments or the figures or in parallel (such as in an environment with parallel processors or multi-threaded processing).

[0220] For the convenience of description, when describing the above device, it is divided into various modules according to functions for separate description. Of course, when implementing the embodiments of this specification, the functions of each module can be implemented in the same or multiple software and / or hardware, or the modules implementing the same function can be realized by a combination of multiple sub-modules or sub-units, etc. The device embodiments described above are only illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical or other forms.

[0221] Those skilled in the art also know that, in addition to implementing the controller in the form of pure computer-readable program code, the method steps can be logically programmed to enable the controller to implement the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. Therefore, such a controller can be regarded as a hardware component, and the devices included therein for implementing various functions can also be regarded as the structures within the hardware component. Or even, the devices for implementing various functions can be regarded as either software modules for implementing the method or structures within the hardware component.

[0222] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0223] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory such as read-only memory (ROM) or flash RAM). The memory is an example of computer-readable media.

[0224] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiment. In the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the embodiments of this specification. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0225] The above is only the embodiment of the embodiments of this specification and is not used to limit the embodiments of this specification. For those skilled in the art, various changes and modifications can be made to the embodiments of this specification. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the embodiments of this specification shall be included within the scope of the claims of the embodiments of this specification.

Claims

1. A method for suppressing dental metal artifacts in optically pumped magnetometer-cardiac magnetic data, characterized in that Including: Pre-whiten the data matrix of the optically pumped magnetometer-cardiac magnetic data to generate a pre-whitening matrix; Generate the delay correlation matrix of the data matrix according to the pre-whitening matrix and a pre-generated time delay sequence; Perform joint diagonalization iterative approximation processing on the delay correlation matrix to generate a decomposition matrix of the data matrix; Suppress tooth metal artifacts in the optically pumped magnetometer-cardiac magnetic data according to the decomposition matrix and a pre-determined time-frequency domain threshold.

2. The method for suppressing dental metal artifacts according to claim 1, wherein Performing joint diagonalization iterative approximation processing on the delay correlation matrix to generate a decomposition matrix of the data matrix includes: Determine the non-diagonal elements of all covariance matrices in the delay correlation matrix; Generate multiple two-dimensional matrices according to the non-diagonal elements; Vectorize the multiple two-dimensional matrices and combine the vectorization results of the multiple two-dimensional matrices to generate a vector matrix; Determine the eigenvector with the largest eigenvalue in the vector matrix; Perform joint diagonalization iterative approximation processing on the delay correlation matrix through the eigenvector to generate a decomposition matrix of the data matrix.

3. The method for suppressing dental metal artifacts according to claim 2, wherein, Performing joint diagonalization iterative approximation processing on the delay correlation matrix through the eigenvector includes: Generate a rotation matrix according to the vector elements of the eigenvector; Perform joint diagonalization iterative approximation processing on the delay correlation matrix through the rotation matrix.

4. The method for suppressing dental metal artifacts according to claim 1, wherein The pre-whitening the data matrix of the optically pumped magnetometer-cardiac magnetic data to generate a pre-whitening matrix includes: Generate a zero-time-delay correlation matrix of the data matrix; Perform singular value decomposition on the zero-time-delay correlation matrix to generate the pre-whitening matrix.

5. The method for suppressing dental metal artifacts according to claim 1, wherein The signal components in the decomposition matrix include: metal artifact interference components, QRS wave source components, TP wave source components, and unknown interference components.

6. The method for suppressing dental metal artifacts according to claim 5, wherein Suppressing tooth metal artifacts in the optically pumped magnetometer-cardiac magnetic data according to the decomposition matrix and a pre-determined time-frequency domain threshold includes: Decompose multiple components in the optically pumped magnetometer-cardiac magnetic data through the decomposition matrix; Determine the total time-frequency domain value of the multiple components; According to the pre-determined time-frequency domain thresholds of the metal artifact interference components, QRS wave source components, TP wave source components, and unknown interference components respectively, determine the positions of the metal artifact interference components, QRS wave source components, TP wave source components, and unknown interference components in the total time-frequency domain value to suppress the tooth metal artifacts.

7. An apparatus for suppressing dental metal artifacts in optically pumped magnetometer-cardiac magnetic data, characterized in that Including: A pre-whitening matrix generation module for pre-whitening the data matrix of the optically pumped magnetometer-cardiac magnetic data to generate a pre-whitening matrix; A delay correlation matrix generation module for generating the delay correlation matrix of the data matrix according to the pre-whitening matrix and a pre-generated time delay sequence; A decomposition matrix generation module for performing joint diagonalization iterative approximation processing on the delay correlation matrix to generate a decomposition matrix of the data matrix; A tooth metal artifact suppression module for suppressing tooth metal artifacts in the optically pumped magnetometer-cardiac magnetic data according to the decomposition matrix and a pre-determined time-frequency domain threshold.

8. The dental metal artifact suppression device according to claim 7, characterized in that, The decomposition matrix generation module includes: A diagonal element determination unit for determining non-diagonal elements of all covariance matrices in the delay correlation matrix; A two-dimensional matrix generation unit for generating a plurality of two-dimensional matrices based on the non-diagonal elements; A vector matrix generation unit for vectorizing the plurality of two-dimensional matrices and combining the vectorized results of the plurality of two-dimensional matrices to generate a vector matrix; An eigenvector determination unit for determining an eigenvector having the largest eigenvalue in the vector matrix; A decomposition matrix generation unit for performing a joint diagonalization iterative approximation process on the delay correlation matrix through the eigenvector to generate a decomposition matrix of the data matrix.

9. The dental metal artifact suppression device according to claim 8, wherein The decomposition matrix generation unit includes: A rotation matrix generation unit for generating a rotation matrix based on vector elements of the eigenvector; A diagonalization iterative processing unit for performing a joint diagonalization iterative approximation process on the delay correlation matrix through the rotation matrix.

10. The dental metal artifact suppression device according to claim 7, wherein The pre-whitening matrix generation module includes: A zero-time-delay correlation matrix generation unit for generating a zero-time-delay correlation matrix of the data matrix; A pre-whitening matrix generation unit for performing singular value decomposition on the zero-time-delay correlation matrix to generate the pre-whitening matrix.

11. The dental metal artifact suppression device according to claim 7, wherein Signal components in the decomposition matrix include: metal artifact interference components, QRS wave source components, TP wave source components, and unknown interference components.

12. The dental metal artifact suppression device according to claim 11, wherein, The dental metal artifact suppression module includes: A magnetocardiogram data decomposition unit for decomposing a plurality of components in the optically pumped magnetometer-magnetocardiogram data through the decomposition matrix; A total time-frequency domain value determination unit for determining the total time-frequency domain value of the plurality of components; A dental metal artifact suppression unit for determining the positions of the metal artifact interference components, QRS wave source components, TP wave source components, and unknown interference components in the total time-frequency domain value according to pre-determined time-frequency domain thresholds of the metal artifact interference components, QRS wave source components, TP wave source components, and unknown interference components respectively, so as to suppress the dental metal artifacts.

13. A computer program product, comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by a processor, the steps of the method for suppressing dental metal artifacts of the optically pumped magnetometer-magnetocardiogram data according to any one of claims 1 to 6 are implemented.

14. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, the steps of the method for suppressing dental metal artifacts of the optically pumped magnetometer-magnetocardiogram data according to any one of claims 1 to 6 are implemented.

15. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, the steps of the method for suppressing dental metal artifacts of the optically pumped magnetometer-magnetocardiogram data according to any one of claims 1 to 6 are implemented.

Images