A SERF-MCG subendocardial ischemia magnetic field feature decoding system and method based on myocardial fiber anisotropy

Abstract

The application discloses a SERF-MCG subendocardial ischemia magnetic field feature decoding system and method based on myocardial fiber anisotropy, which comprises the following steps: constructing a local myocardial anatomical coordinate system according to a heart geometric model, generating a myocardial fiber direction field and anisotropic conduction parameters; constructing an ischemia source in a subendocardial layer and reconstructing electrophysiological parameters to obtain ischemia-related source current density or damage current density; further extracting an electric vortex topological feature formed by the interaction of myocardial fiber anisotropy and subendocardial ischemia damage current, and combining the magnetic field forward calculation results at the SERF-MCG sensor array to extract magnetic field features such as magnetic field angle, spatial topology, QRS-T angle and angle abnormality degree, and outputting a subendocardial ischemia magnetic field feature decoding result.

Description

Technical Field

[0001] This invention belongs to the fields of biomagnetic measurement, magnetocardiography signal analysis, cardiac electrophysiological modeling, medical signal feature extraction and multiphysics forward computation technology, specifically involving a SERF-MCG subendocardial ischemic magnetic field feature decoding system and method based on myocardial fiber anisotropy. Background Technology

[0002] Subendocardial ischemia is one of the important electrophysiological bases in diseases such as non-ST-segment elevation myocardial infarction. Compared with transmural injury, subendocardial ischemia is often confined to the inner wall of the ventricle, and the electrophysiological changes are relatively insidious. The surface electrical signals may be atypical, making early identification and mechanistic interpretation difficult. Traditional electrocardiography can reflect the overall changes in cardiac electrical activity, but its measurement results are easily affected by factors such as thoracic tissue electrical conduction, the position of surface electrodes, and signal attenuation. Its ability to express information on local tangential currents, eddy currents, and spatial topological rotation is limited.

[0003] Magnetocardiography (MCG) is a non-invasive technique for recording the weak magnetic field signals generated by cardiac electrical activity. With the development of spin-exchange relaxation free magnetometers (SERF-MCG), the SERF-MCG system can acquire highly sensitive cardiac magnetocardial signals under multi-channel array conditions, providing a new measurement method for analyzing the spatial distribution of cardiac electrical activity. However, SERF-MCG measures the magnetic field manifestation of myocardial electrophysiological sources after transmission through the trunk conductors and space. Without a decoding model constrained by a clear physiological mechanism, relying solely on empirical waveforms or purely data-driven methods is insufficient to explain the weak magnetic anomalies caused by subendocardial ischemia.

[0004] Existing forward modeling methods for myocardial vascular occlusion (MCG) can typically calculate the surface magnetic field distribution based on cardiac geometry, trunk conductors, and sensor arrays. However, some methods still employ isotropic conduction assumptions or equivalent dipole approximations. While these simplified models facilitate computation, they neglect the constraint effect of the myocardial fiber helical structure on the local current propagation path, easily losing information on tangential current components, magnetic field topological rotation, and vector reconstruction features during the repolarization phase. In subendocardial ischemia, the injury current does not simply diffuse radially along the transmural direction; its propagation path is redirected by the anisotropic tensor of myocardial fibers, thus forming a source current anomaly with spatial curl and topological structure.

[0005] Therefore, a system and method are needed that can unify myocardial geometry, fiber anisotropy, electrophysiological propagation, subendocardial ischemic sources, source current vortex topology, and SERF-MCG magnetic field topology. This system and method should be able to explain the subendocardial ischemic magnetic field characteristics from the chain of "myocardial fiber anisotropy—ischemic injury current—electric vortex topology—SERF-MCG magnetic field angle anomaly," providing physical mechanism constraints for SERF-MCG signal analysis, magnetic field feature extraction, and auxiliary assessment. Summary of the Invention

[0006] To address the shortcomings in existing SERF-MCG ischemic feature analysis regarding the anisotropic constraints of myocardial fibers, the topological structure of injury current vortices, and the magnetic field vector reconstruction mechanism, this invention provides a SERF-MCG subendocardial ischemic magnetic field feature decoding system and method based on myocardial fiber anisotropy. The aim is to obtain the topological characterization of the source current vortex generated by the interaction between myocardial fiber anisotropy and subendocardial ischemic injury current, and to combine this with SERF-MCG magnetic field angular and spatial topological features to achieve mechanistic decoding and auxiliary assessment of subendocardial ischemia-related magnetic field features.

[0007] To achieve the above objectives, the present invention provides the following solution: A SERF-MCG subendocardial ischemic magnetic field feature decoding system based on myocardial fiber anisotropy, the system comprising: a model parameter acquisition module, a myocardial anatomical coordinate construction module, a myocardial fiber anisotropy construction module, an anisotropic electrophysiological propagation modeling module, a subendocardial ischemia source modeling module, an electric vortex feature extraction module, a SERF-MCG magnetic field forward calculation module, a magnetic field topology feature extraction module, and a decoding output module; The model parameter acquisition module is used to acquire at least one of the following: cardiac geometric parameters, myocardial boundary parameters, trunk conductor parameters, forward computational domain parameters, SERF-MCG sensor array parameters, electrophysiological parameters, and subendocardial ischemic region parameters. The myocardial anatomy coordinate construction module is used to construct a local myocardial anatomy coordinate system based on cardiac geometric parameters and myocardial boundary parameters; The myocardial fiber anisotropy construction module is used to generate a myocardial fiber direction field based on the local myocardial anatomical coordinate system, and to construct anisotropic conduction parameters based on the myocardial fiber direction field. The anisotropic electrophysiological propagation modeling module is used to construct a myocardial electrophysiological propagation model based on anisotropic conduction parameters, and obtain the spatiotemporal distribution of transmembrane potential and source current density in the myocardial computational region. The subendocardial ischemia source modeling module is used to determine the subendocardial ischemia region within the myocardial calculation region based on the parameters of the subendocardial ischemia region, and to pathologically reconstruct the electrophysiological parameters within the subendocardial ischemia region to obtain ischemia-related electrophysiological source terms. The eddy current feature extraction module is used to extract source current eddy current topology features or ischemic injury current eddy current topology features based on the source current density or the injury current density determined by the ischemia-related electrophysiological source term. The SERF-MCG magnetic field forward calculation module is used to calculate the magnetic field distribution at the SERF-MCG sensor array based on at least one of the source current density, damage current density, trunk conductor parameters, forward calculation domain parameters, and SERF-MCG sensor array parameters. The magnetic field topology feature extraction module is used to extract magnetic field topology features based on the magnetic field distribution; The decoding output module is used to output the decoding result of the subendocardial ischemic magnetic field feature based on at least one of the source current vortex topology feature, the ischemic injury current vortex topology feature, and the magnetic field topology feature.

[0008] This invention also provides a method for decoding the subendocardial ischemic magnetic field features of SERF-MCG based on myocardial fiber anisotropy. The method is implemented using the aforementioned system and includes: Step 1: Obtain at least one of the following parameters: cardiac geometry, myocardial boundary, trunk conductor, forward computation domain, SERF-MCG sensor array, electrophysiological parameters, and subendocardial ischemic region parameters. Step 2: Construct a local myocardial anatomical coordinate system based on cardiac geometric parameters and myocardial boundary parameters; Step 3: Generate the myocardial fiber orientation field based on the local myocardial anatomical coordinate system, and construct anisotropic conduction parameters based on the myocardial fiber orientation field; Step 4: Construct a myocardial electrophysiological propagation model based on anisotropic conduction parameters to obtain the spatiotemporal distribution of transmembrane potentials and source current density; Step 5: Determine the subendocardial ischemic region based on the parameters of the subendocardial ischemic region, and perform pathological reconstruction of the electrophysiological parameters in the subendocardial ischemic region to obtain ischemia-related electrophysiological source terms; Step 6: Extract source current vortex topology features or ischemic injury current vortex topology features based on the source current density or the injury current density determined by the ischemia-related electrophysiological source terms. Step 7: Calculate the magnetic field distribution at the SERF-MCG sensor array based on at least one of the source current density, damage current density, trunk conductor parameters, forward computation domain parameters, and SERF-MCG sensor array parameters. Step 8: Extract the topological features of the magnetic field based on the magnetic field distribution; Step 9: Based on at least one of the source current vortex topology features, ischemic injury current vortex topology features, and magnetic field topology features, output the decoding result of the subendocardial ischemic magnetic field features.

[0009] Preferably, the method for constructing a local myocardial anatomical coordinate system based on cardiac geometric parameters and myocardial boundary parameters includes: Based on the endocardial and epicardial boundaries, a continuous scalar field is solved within the myocardial computation region, pointing from the endocardium to the epicardium or from the epicardium to the endocardium, and the transmural direction is determined based on the continuous scalar field. The cardiac anatomical long axis is determined based on the apical reference point and the basal reference point, and the normalized coordinates of the long axis within the myocardial computational region are obtained using vector projection. A local orthogonal basis is constructed based on the wall-penetrating direction and the major axis direction. The local orthogonal basis includes the wall-penetrating basis vector, the major axis basis vector, and the circumferential basis vector.

[0010] Preferably, the method for generating a myocardial fiber orientation field based on a local myocardial anatomical coordinate system and constructing anisotropic conduction parameters based on the myocardial fiber orientation field includes: Determine the transmural coordinates of the myocardium based on the endocardial distance, epicardial distance, or a continuous scalar field. The distribution of myocardial fiber helix angles was determined based on myocardial transmural coordinates and normalized major axis coordinates. The local orthogonal basis is rotated around the transmural direction according to the helical angle of the myocardial fiber to obtain the myocardial fiber directional field; Anisotropic diffusion tensors or anisotropic conductivity tensors are constructed based on the myocardial fiber directional field, so that electrophysiological excitation has different propagation characteristics in the myocardial fiber direction and non-fiber direction.

[0011] Preferably, the method for determining the subendocardial ischemic region based on parameters of the subendocardial ischemic region and pathologically reconstructing the electrophysiological parameters within the subendocardial ischemic region to obtain ischemia-related electrophysiological parameters includes: Based on at least one of the ischemia center, ischemia radius, spatial constraint function and transmural threshold, determine the local ischemia candidate region within the myocardial computational region; Based on the transmural coordinates of the myocardium, the candidate region for local ischemia is located in the subendocardial layer, thus obtaining the subendocardial ischemic region; In the subendocardial ischemic region, at least one of the following parameters is pathologically reconstructed: resting potential, excitation threshold, action potential amplitude parameter, recovery parameter, diffusion parameter, and local conductivity. The spatiotemporal distribution of transmembrane potential, source current density, or injury current density are calculated based on the reconstructed electrophysiological parameters, i.e., ischemia-related electrophysiological source terms.

[0012] Preferably, the method for extracting source current vortex topology features or ischemic injury current vortex topology features based on source current density or injury current density determined by ischemia-related electrophysiological source terms includes: The source current density is calculated based on the anisotropic conduction parameters and the spatial gradient of the transmembrane potential. The source current vortex field can be obtained from the spatial curl of the source current density, or the ischemic injury current vortex field can be obtained from the spatial curl of the injury current density. Based on the vortex field of the source current or the vortex field of the ischemic injury current, at least one of the following computational characteristics is obtained: vortex field modulus distribution, directional distribution, spatial distribution, boundary annular distribution, local extremum location, and vortex topological continuity.

[0013] Preferably, the method for extracting magnetic field topological features based on magnetic field distribution includes: The magnetic field source region and magnetic field sink region are determined based on the magnetic field distribution at the SERF-MCG sensor array; The direction of the magnetic field vector or the magnetic field angle is determined based on the positional relationship between the magnetic field source region and the magnetic field sink region. The QRS magnetic field vector and T-wave magnetic field vector were extracted within the QRS composite wave time window and the T-wave time window, respectively. Calculate at least one of the following based on the QRS magnetic field vector and the T-wave magnetic field vector: QRS-T angle, T-wave magnetic field angle offset, magnetic field angle collapse degree, and magnetic field topological rotation characteristics.

[0014] Preferably, the magnetic field feature decoding results include at least one of the following: ischemic injury current vortex field characterization information, vortex field spatial distribution information, magnetic field topological rotation direction, QRS-T angle abnormality, T wave magnetic field angle deviation, endocardial ischemia-related magnetic field feature indication information, and SERF-MCG auxiliary assessment results.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: First, this invention combines local myocardial anatomical coordinates, transmural coordinates, normalized major axis coordinates, and fiber helix angle mapping, enabling the anisotropic conduction parameters of myocardial fibers to adapt to the actual tilt geometry of the heart. Second, this invention obtains ischemia-related source currents or injury currents through subendocardial ischemic region constraints and ischemic electrophysiological parameter reconstruction, avoiding the use of only empirical equivalent sources to describe ischemic magnetic field anomalies. Third, this invention introduces source current vortex topology features or ischemic injury current vortex topology features to characterize the redirection effect of myocardial fiber anisotropy on the propagation path of ischemic injury currents. Fourth... This invention combines the topological features of electric vortexes with the topological features of SERF-MCG magnetic fields and the QRS-T angle features, which is beneficial for explaining the repolarization magnetic field vector reconstruction and magnetic field angle abnormalities caused by subendocardial ischemia. Fifth, the output of this invention is the magnetic field feature decoding result and auxiliary evaluation result, which can provide calculable and traceable physical mechanism constraints for SERF-MCG signal interpretation, feature screening, algorithm training and clinical research. Among them, the quantitative evaluation of the degree of magnetic field abnormality is mainly achieved by SERF-MCG magnetic field features such as magnetic field angle, spatial topology and QRS-T angle. Attached Figure Description

[0016] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the SERF-MCG subendocardial ischemic magnetic field feature decoding method provided in this embodiment of the invention. Figure 2 This is a schematic diagram of the module structure of the SERF-MCG subendocardial ischemic magnetic field feature decoding system provided in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the anisotropic structure results and construction process of myocardial fibers provided in an embodiment of the present invention, wherein... Figure 3 (a) shows the anisotropy of myocardial fibers. Figure 3 (b) is the process for inputting the endocardial / epidermal boundary, constructing the anatomical long axis and transmural coordinates, generating the fiber helix angle distribution, and constructing the anisotropic fiber direction field / tensor; Figure 4 This is a schematic diagram illustrating the extraction of subendocardial ischemic sources and injury current vortex topological features according to an embodiment of the present invention. Figure 4 (a) shows the calculation results of the injury current eddy source under subendocardial ischemia conditions. Figure 4 (b) is the workflow for ischemic region parameters, electrophysiological reconstruction, source current / damage current calculation and vortex topology feature extraction. Figure 5 This is a schematic diagram illustrating the extraction of SERF-MCG magnetic field topological features and QRS-T magnetic field angle features according to an embodiment of the present invention. Figure 5 (a) shows the QRS-T magnetic field angle results. Figure 5 (b) is the process for locating the magnetic field source / sink region, extracting the magnetic field vector, and extracting the QRS-T angle features. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] Example 1 like Figure 2 As shown, this embodiment provides a SERF-MCG subendocardial ischemic magnetic field feature decoding system based on myocardial fiber anisotropy. The system includes a model parameter acquisition module, a myocardial anatomical coordinate construction module, a myocardial fiber anisotropy construction module, an anisotropic electrophysiological propagation modeling module, a subendocardial ischemia source modeling module, an electric vortex feature extraction module, a SERF-MCG magnetic field forward calculation module, a magnetic field topology feature extraction module, and a decoding output module.

[0021] The model parameter acquisition module is used to acquire cardiac geometric parameters, myocardial boundary parameters, trunk conductor parameters, forward computational domain parameters, SERF-MCG sensor array parameters, myocardial electrophysiological parameters, and subendocardial ischemic region parameters. The cardiac geometric parameters are obtained from at least one of cardiac magnetic resonance imaging (MRI), computed tomography (CT) images, ultrasound images, or a pre-established three-dimensional cardiac geometric model. The myocardial computational region is obtained through image segmentation, surface reconstruction, and volume mesh generation. The endocardial boundary, epicardial boundary, and basal boundary are extracted from the myocardial computational region, and apical reference points and basal reference points are determined at the ventricular apex and basal center, respectively. The apical reference point, basal reference point, and the cardiac anatomical long axis determined by both are all considered cardiac geometric parameters.

[0022] The trunk conductor region is divided into at least one corresponding conductor sub-region from myocardium, blood, lung, bone, and trunk soft tissue according to tissue labels, and a corresponding preset conductivity is assigned to each sub-region based on its tissue category. Let label(X) represent the body conductor region of the torso, and let label(X) represent the tissue label at position X. Then, the body conductor sub-region corresponding to the p-th type of tissue is represented as follows: For each bulk conductor sub-region, a local conductivity is assigned according to its tissue type: The forward computational domain is jointly determined by the cardiac computational region, the trunk conductor region, and the external sensor measurement region. Myocardial electrophysiological parameters include at least one of the following: excitation threshold, action potential amplitude parameter, recovery parameter, diffusion parameter, and local conductivity, corresponding to normal and ischemic tissues. Parameters of the subendocardial ischemic region include at least one of the following: ischemic center, ischemic radius or semi-axis, transmural threshold, transition zone width, and spatial smoothness function parameter.

[0023] SERF-MCG sensor array parameters are obtained from sensor array design documents, calibration documents, or the actual measurement device, including the number of sensor channels, array arrangement, channel local coordinates, measurement direction, and measurement plane position. Using the torso coordinate system as a unified spatial coordinate system, the local coordinates of the k-th sensor are transformed into the unified coordinates: (1) in, Let be the position vector of the k-th sensor in the unified coordinate system. Let it be its local position vector. Let be a rotation matrix. It is a translation vector. · This represents matrix-vector multiplication. The model parameter acquisition module transforms the aforementioned geometric, boundary, conductor, electrophysiological, ischemic region, and sensor array parameters into a unified spatial coordinate system and a unified unit system, forming the input parameters for subsequent modules.

[0024] The myocardial anatomy coordinate construction module is used to construct a local myocardial anatomy coordinate system based on the myocardial boundary and the long axis of cardiac anatomy. This coordinate system does not directly rely on global Cartesian coordinates, but instead describes the transmural direction, long axis direction, and circumferential direction of myocardial tissue based on the anatomical orientation of the heart itself, thereby reducing the errors caused by the heart's tilt posture on fiber mapping and current direction estimation.

[0025] The myocardial fiber anisotropy construction module is used to generate the myocardial fiber orientation field based on the local myocardial anatomical coordinate system, and to construct anisotropic conduction parameters based on the myocardial fiber orientation field, that is, to further construct the anisotropic diffusion tensor or conductivity tensor.

[0026] The anisotropic electrophysiological propagation modeling module is used to input the anisotropic diffusion tensor or conductivity tensor into the myocardial electrophysiological propagation model, solve the spatiotemporal distribution of transmembrane potential and recovery variables in the myocardial calculation region, and calculate the source current density based on the spatial gradient of the transmembrane potential and the anisotropic diffusion tensor or conductivity tensor.

[0027] The subendocardial ischemia source modeling module is used to determine the subendocardial ischemia region within the myocardial calculation region based on the parameters of the subendocardial ischemia region, and to pathologically reconstruct the electrophysiological parameters within the subendocardial ischemia region to obtain ischemia-related electrophysiological source terms.

[0028] The eddy current feature extraction module is used to extract source current eddy current topology features or ischemic injury current eddy current topology features based on the source current density or the injury current density determined by ischemia-related electrophysiological source terms.

[0029] The SERF-MCG forward magnetic field calculation module is used to calculate the magnetic field distribution at the SERF-MCG sensor array based on at least one of the following: source current density, damage current density, body conductor parameters, forward calculation domain parameters, and SERF-MCG sensor array parameters.

[0030] The magnetic field topology feature extraction module is used to extract magnetic field topology features based on the magnetic field distribution; the magnetic field topology features include at least one of the following: spatial distribution of magnetic field source region and magnetic field sink region, magnetic field vector direction, magnetic field angle, magnetic field topology rotation direction, QRS-T angle, T-wave magnetic field angle offset, and magnetic field angle anomaly degree.

[0031] The decoding output module receives at least one of the following: source current eddy current topology characteristics, ischemic injury current eddy current topology characteristics, magnetic field angle characteristics, magnetic field spatial topology characteristics, QRS-T angle characteristics, and degree of angle collapse.

[0032] Specifically, the input features are first normalized according to time window, sensor channel, and feature type; second, the ischemic injury current vortex topology features are paired with the magnetic field direction, source / sink centroid, and spatial extremum position within the corresponding time window; and third, the decoding results of the subendocardial ischemic magnetic field features are output according to preset rules, threshold judgment models, or trained classification models.

[0033] The decoding results include at least one of the following: presence of ischemic injury current vortices, vortex spatial location, vortex direction, degree of QRS-T angle anomaly, T-wave magnetic field angle shift, degree of magnetic field angle collapse, ischemia-related magnetic field characteristic information, and SERF-MCG auxiliary assessment results.

[0034] Specifically, the model parameter acquisition module acquires model parameters in the following manner: it acquires pre-constructed forward computation model parameters of the heart-trunk-sensor array, wherein the forward computation model parameters include at least one of the following: heart geometric parameters, endocardial boundary parameters, epicardial boundary parameters, basal surface boundary parameters, trunk geometric model parameters, trunk conductor parameters, myocardial electrophysiological parameters, ischemic region parameters, and SERF-MCG sensor array parameters; the SERF-MCG sensor array parameters include at least one of the following: number of sensor channels, sensor array arrangement, sensor channel position, and sensor measurement direction; the model parameter acquisition module forms unified input parameters for myocardial anatomical coordinate construction, myocardial fiber anisotropy construction, electrophysiological propagation modeling, ischemia source modeling, and magnetic field forward computation based on the forward computation model parameters.

[0035] Construction of a myocardial anatomical coordinate and fiber anisotropy model: like Figure 3 As shown, this embodiment illustrates the construction method of myocardial anatomical coordinates and fiber anisotropy model. Figure 3 (a) shows the anisotropy of myocardial fibers calculated from transmural coordinates and fiber helix angles, where different colors correspond to different fiber helix angle values; Figure 3(b) shows the processing flow of the myocardial fiber orientation field obtained by inputting the endocardial boundary, epicardial boundary and basal surface boundary, constructing the anatomical long axis and transmural coordinates, generating the fiber helical angle distribution and combining local basis vectors.

[0036] Specifically, Figure 3 (b) shows the process of first determining the transmural scalar field based on the endocardial and epicardial boundaries, then determining the anatomical long axis based on the apical and basal reference points, and constructing local transmural, long axis and circumferential orthogonal bases at each myocardial grid node; interpolating according to the preset helical angles on the endocardial and epicardial sides and the transmural coordinates of the nodes to obtain the node fiber helical angles, and combining them with the local orthogonal bases to generate the unit fiber direction vector. Figure 3 (a) The color only indicates the numerical distribution of the helix angle. Figure 3 (b) outputs a fiber orientation field determined by the helix angle and the local orthogonal basis. Subsequently, the myocardial anatomy coordinate construction module constructs the coordinates based on the endocardial boundary. epicardial border and base surface boundary Determine the myocardial computation region Ω, and solve for the dimensionless scalar field U(r) that continuously varies from the endocardium to the epicardium within Ω: (2) The following Dirichlet boundary conditions are applied to the endocardial and epicardial boundaries respectively: (3) Apply a zero-flux boundary condition to the base surface boundary: (4) Where r is the spatial location vector within the myocardial computation region, and U is the transmural scalar field. , and These represent the endocardial boundary, epicardial boundary, and basal boundary, respectively; n is the external normal vector of the basal boundary; and ∇ is the vector relative to the spatial position. The gradient operator is denoted by ∇², and ∇² is the Laplace operator. To obtain the distance field for transmural coordinate calculations, the distance functions to the endocardium and epicardium can also be solved separately. (r) and (r): (5) Where endo represents the endocardium and epi represents the epicardium; s is the boundary marker, and when s=endo, d s (r)=d endo (r), Γ s =Γ endo When s = epi, d s (r)=depi (r), Γ s =Γ epi .

[0037] Secondly, in the three-dimensional geometric model of the myocardium, the grid node located at the ventricular apex region and farthest from the basal surface along the candidate long axis, or its neighborhood geometric center, is determined as the apical reference point. Project the nodes on the base surface boundary onto the base surface and take their geometric centers to determine the base reference point. After completing the above localization, determine the major axis vector of cardiac anatomy and its squared length: (6) (7) For any point within the myocardial computation region The normalized coordinates of the major axis are obtained by projecting the point onto the anatomical major axis: (8) in, ( () represents the normalized major axis coordinate from the apex to the basement. The through-wall basis vectors are determined based on the gradient of the through-wall scalar field. (9) global reference axis =[0,0,1] T Project to Orthogonal local tangent planes yield the major axis basis vectors: (10) The circumferential basis vectors are represented as: (11) in, , and These represent the transverse direction, the local long axis direction, and the circumferential direction, respectively, and together they form a right-handed local orthogonal basis.

[0038] Furthermore, based on the endocardial distance ( ) and epicardial distance ( Calculate the coordinates of the through wall: (12) Where, β( )=1 indicates proximity to the endocardium, β( A value of 0 indicates proximity to the epicardium. This is based on the helix angle of the fibers on the endocardial side. Epicardial fiber helix angle and major axis normalized coordinates ( ), generating spatial fiber helix angle: (13) In one implementation, Take +60°, -60° is selected; this value can also be determined based on a preset fiber map or individualized fiber measurement results. The fiber direction is obtained by combining the local circumferential base vector and the major axis base vector according to the fiber helix angle: (14) in, ( () represents the unit fiber direction vector. In the primary direction, an anisotropic diffusion tensor is defined in the fiber's local reference frame: (15) in, and These are the diffusion coefficients along and perpendicular to the fiber direction, respectively. The local fiber reference frame is mapped to the global coordinate system, or an equivalent tensor expression is used to construct the spatial anisotropic diffusion tensor: (16) in, ( () represents the global anisotropic diffusion tensor. It is a second-order unit tensor. The baseline isotropic diffusion coefficient, The additional diffusion coefficient along the fiber direction. Represents the tensor product. ( This is used for subsequent electrophysiological propagation and source current density calculation.

[0039] Modeling of subendocardial ischemia sources and anisotropic electrophysiological propagation: like Figure 4 As shown, this embodiment illustrates the construction method of the subendocardial ischemia source modeling module and the anisotropic electrophysiological propagation model. The anisotropic electrophysiological propagation modeling module uses the spatial anisotropic diffusion tensor... ( Input the myocardial electrophysiological propagation model and solve the transmembrane potential V( ) within the myocardial computational region Ω. ,t) and recovery variable w( The spatiotemporal distribution of ,t). In one implementation, a single-domain reaction-diffusion model is used: (17) Where V is the dimensionless transmembrane potential and t is time. For the second-order anisotropic diffusion tensor, Let w be the ion current term and w be the recovered variable. When using the Aliev-Panfilov model, the ion current term and the recovered variable satisfy: (18) Where a is the excitation threshold, k is the scaling parameter for the action potential amplitude and rise rate, and e is the time scale parameter for the recovery process; a, k, and e can be set as normal tissue parameters or spatial parameter fields reconstructed based on ischemia weights, respectively. After completing temporal and spatial discretization, V( ,t) and w( ,t).

[0040] Calculate the myocardial imprinting source current density based on the spatial gradient of the transmembrane potential and the anisotropic diffusion tensor: (19) in, The vector of myocardial source current density. ( )·∇ ( ,t) denotes the tensor-vector multiplication of the diffusion tensor and the transmembrane potential gradient. ( ,t) is the transmembrane potential with physical dimensions.

[0041] To simulate the depolarization and repolarization processes of cardiomyocytes, the Aliev-Panfilov model, the FitzHugh-Nagumo model, or other response-diffusion models that can describe the propagation of myocardial action potentials can be used. By encoding the myocardial fiber orientation field into D, the model can simulate the anisotropic characteristics of myocardial excitation propagating rapidly along the fiber direction and relatively slowly in the transverse direction, and provide the transmembrane potential distribution for subsequent source current calculations.

[0042] The subendocardial ischemia source modeling module is based on the ischemia center ischemic radius Transmural threshold and myocardial transmural coordinates β ( Within the myocardial calculation region Ω, the subendocardial ischemic area is determined: (20) in, This is the subendocardial ischemic area; β( It is obtained from equation (12). The permeability threshold is between 0 and 1. In one embodiment, =0.6, which restricts the ischemic region to the inner myocardium near the endocardium. For ellipsoidal ischemic regions, the spherical distance constraint can be replaced with an ellipsoidal distance constraint normalized according to the three semi-axes.

[0043] To avoid abrupt changes in the boundary parameters of the ischemic zone, the parameters are determined based on the transmural coordinates and the preset transition zone width. Constructing smooth ischemia weights: (twenty one) in, It is a second-order continuous smooth step function. Width of the ischemic boundary transition zone; ischemic core region Approximately 1, normal myocardial area Approaching 0. For any electrophysiological parameter P that needs to be reconstructed, construct a spatially continuous parameter field: (twenty two) in, and These are parameters for normal tissue and parameters for ischemic tissue, respectively. Parameter P includes at least one of the following: excitation threshold a, action potential scaling parameter k, recovery parameter e, diffusion coefficient, resting potential, and local conductivity.

[0044] Based on the reconstructed electrophysiological parameters from equations (17) to (22), the model calculates the transmembrane potential in the ischemic region and its boundary. ( ,t), and its spatial gradient ∇ ( ,t) and normal operating potential source current density ( The resting potential difference caused by ischemia is denoted as t). ( The additional effects caused by changes in conduction parameters and local conductivity are incorporated into the effective anisotropic tensor. ( ).

[0045] The ischemia-related electrophysiological components include the spatiotemporal distribution of transmembrane potentials. ( ,t), source current density ( ,t), ischemic injury current density ( ,t) and total current density ( At least one of ,t); the ischemic injury current density ( ,t) is used for characterizing the eddy current of ischemic injury and calculating the forward magnetic field.

[0046] In one implementation, the total source current density is obtained by superimposing the normal action potential source current and the ischemic injury current: (twenty three) Injury current density is determined by ischemic resting potential difference and local effective anisotropy tensor: (twenty four) The eddy current feature extraction module calculates the spatial curl of the source current density or the injury current density to obtain the source current eddy field or the ischemic injury current eddy field: (25) in, The electric vortex field is characterized by at least one of the following: magnitude, direction, local extremum location, annular distribution of ischemic boundary, and spatial continuity. The source current vortex field or ischemic injury current vortex field is determined by the spatial curl of the corresponding current density field and is used to characterize the local rotational tendency, topological closure structure, and directional redirection effect caused by anisotropic fiber structure in the current density field.

[0047] (26) Furthermore, the vortex field modulus is calculated at each myocardial grid node m within the target time window. and unit direction vector : (27) in, Let be the vortex field vector at node m. To prevent extremely small positive numbers where the denominator is zero during direction normalization, a preset threshold τ_ is used. Determine the set M of significant vortex nodes, and calculate the vortex center using the vortex field modulus as the weight: (28) in, Let be the position vector of node m. Let Rδ be the neighborhood centered at the ischemic boundary with a width δ. Calculate the proportion of magnitudes of significant vortices within this neighborhood, η_ring; based on the largest connected component in set M... The proportion is used to calculate the vortex topological continuity ηconn: Further, the maximum vortex field modulus within set M is taken. and its node positions As local extrema and extremum locations. , , η_ring, η_conn, , At least one of the following constitutes a source current vortex topology feature or an ischemic injury current vortex topology feature: vortex topology continuity.

[0048] In obtaining the source current density and damage current density Subsequently, the SERF-MCG magnetic field forward calculation module calculates the magnetic field based on the myocardial source domain. , trunk conductor domain Conductivity of bulk conductors, permeability of free space The magnetic field distribution is calculated using sensor array parameters. Under the quasi-static electromagnetic approximation, the total magnetic field in the sensor measurement area is obtained by superimposing the primary magnetic field generated by the myocardial source current and the secondary magnetic field generated by the return current in the trunk conductors. (29) in, ( ,t) is the magnetic flux density vector. For myocardial source current density, The return current density of the body conductor. The source point location, This represents the location of the observation point. For the k-th sensor, the location is determined based on the sensor's measurement direction. and sensor position Obtain channel output (t)= · ( The forward calculation module outputs at least one of the following: target time magnetic field map, QRS composite wave magnetic field map, T-wave magnetic field map, and multi-channel time-series magnetic field signal.

[0049] In this embodiment, the change in injury current at the subendocardial ischemic boundary is characterized as a calculable vortex topology. Furthermore, this vortex topology is linked to the spatial distribution of the SERF-MCG magnetic field, so that the subsequent extraction of magnetic field topology features no longer depends solely on the original waveform amplitude, but can utilize the physical mechanism information determined by the anisotropy of myocardial fibers.

[0050] Magnetic field topology feature extraction and decoding output: like Figure 5 (a) and Figure 5 As shown in (b), this embodiment illustrates the extraction methods for SERF-MCG magnetic field topology features and QRS-T magnetic field angle features. Figure 5 (a) shows the magnetic field angle results corresponding to the QRS composite wave time window and the T wave time window, which are used to compare the relative relationship between the QRS magnetic field direction and the T wave magnetic field direction under different models or different objects; Figure 5(b) shows the process from SERF-MCG magnetic field map input, magnetic field source / sink region location, source / sink centroid calculation, magnetic field vector and angle calculation to QRS-T angle feature output.

[0051] The magnetic field topology feature extraction module partitions the magnetic field map into positive and negative polarity zones within preset QRS composite wave time windows and T-wave time windows. The QRS composite wave time window is used to extract magnetic field features in the QRS phase, and the T-wave time window is used to extract magnetic field features in the T-wave phase. For any time window W, sensor channels with magnetic field amplitudes greater than a preset positive threshold are identified as magnetic field source channels, and sensor channels with magnetic field amplitudes less than a preset negative threshold are identified as magnetic field sink channels. Let p represent the polarity category, and p can be either src or sink, where src represents the magnetic field source and sink represents the magnetic field sink; let K... p Let b represent the set of sensor channels belonging to polarity category p; k Let r represent the signed feature amplitude extracted by the k-th sensor channel within the current time window W; k Let c represent the position vector of the k-th sensor; p (W) represents the weighted centroid of the region corresponding to polarity category p within the time window W. The source and sink centroids of the magnetic field are calculated based on the amplitude weights of each channel: (30) Wherein, the summation symbol Σ represents the summation over set K. p Summing all sensor channels in |b k | represents the signed characteristic magnitude b k The absolute value of |b is used as the non-negative weight; k |·r k This indicates that the sensor position vector r is weighted using this weight. k ; denominator Σ|b k | represents the sum of all channel weights; p∈{src,sink} represents the expression (30) used to calculate the source centroid and sink centroid of the magnetic field, respectively. If the set of source channels or sink channels is empty, or the sum of the corresponding weights is zero, the time window will not output the corresponding centroid and will be marked as invalid or missing, and will not be replaced by a zero vector.

[0052] The principal direction of the magnetic field within this time window is determined by the projection vector from the sink centroid to the source centroid of the magnetic field, and the magnetic field angle is expressed as: (31) Where, θ W represents the principal direction angle of the magnetic field within the time window W; atan2 represents the arctangent function with quadrant discrimination, whose first input is the ordinate component of the direction vector and the second input is the abscissa component of the direction vector. c in equation (31) src,y (W) and csink,y (W) represents the ordinate of the source centroid and the sink centroid of the magnetic field in the sensor measurement plane coordinate system, respectively. src,x (W) and c sink,x (W) represents the abscissa of both in the same coordinate system. Therefore, the two input quantities of atan2 constitute the ordinate and abscissa components of the direction vector from the centroid of the magnetic field sink to the centroid of the magnetic field source. All time windows adopt the same sensor coordinate system and the same positive angular direction convention. The magnetic field source / sink region, centroid position, positive and negative pole spatial distribution, and extreme value position belong to the spatial topological features of the magnetic field; θW and the direction determined by it belong to the angular features of the magnetic field.

[0053] Within the QRS composite wave time window and the T wave time window, the source centroid and sink centroid of the magnetic field are determined according to equation (31) respectively, and the two-dimensional direction vector from the sink centroid to the source centroid is denoted as q. QRS and q T , where q QRS Corresponding to the QRS composite wave time window, q T Corresponding to the T-wave time window. Calculate the QRS-T angle based on the angle between the two direction vectors: (32) Where, θ QRS-T q represents the undirected angle between the QRS magnetic field vector and the T-wave magnetic field vector; QRS q represents the magnetic field vector within the time window of the QRS composite wave. T q represents the magnetic field vector within the T-wave time window; QRS ·q T Represents the dot product of two vectors; ||q QRS || and ||q T || represents the Euclidean norm or magnitude of the two vectors, respectively; arccos represents the arccosine function. Dividing the dot product of the two vectors by the product of their magnitudes in the fraction is equivalent to normalizing the two vectors before calculating the cosine of the angle between them; therefore, the resulting angle lies within the range of 0 to π. During calculation, the arccosine input value should be truncated to fall within the range of -1 to 1.

[0054] The included angle within the range of 0 to π is obtained by normalizing the angle period. The degree of magnetic field angle collapse is calculated based on the QRS magnetic field angle and the T-wave magnetic field angle. (33) Among them, C collapse Indicates the degree of angular collapse of the magnetic field; θ T θ represents the principal direction angle of the magnetic field within the T-wave time window. QRS |θ represents the principal direction angle of the magnetic field within the time window of the QRS composite wave; T -θQRS | represents the absolute value of the difference between the two principal direction angles; π represents the radian constant corresponding to the semicircle. C collapse This is used to indicate the degree of collapse in the relative angular relationship between the QRS magnetic field direction and the T-wave magnetic field direction. Under normal circumstances, the T-wave direction and the QRS direction are nearly opposite, and the angle between them is close to π, resulting in a small degree of angular collapse. As the T-wave direction gradually rotates towards the QRS direction, the angle between them decreases, and the degree of angular collapse increases. The T-wave magnetic field angle offset is based on the normal reference angle under the same sensor array, the same sensor plane coordinate system, and the same T-wave time window, representing the periodic normalized difference of the target object's T-wave principal direction relative to the normal reference angle. The QRS-T angle anomaly indicates the degree of deviation of the target object's QRS-T angle from the preset normal reference value or reference distribution.

[0055] In addition to the angular features mentioned above, the magnetic field topology feature extraction module can also output at least one of the following: spatial distribution of magnetic field positive and negative poles, topological orientation of dipole-like structures, source / sink centroid positions, magnetic field extremum position offset, topological rotation direction from the QRS time window to the T-wave time window, and synchronous changes across multiple channels. These features are collectively used to characterize repolarization magnetic field vector reconstruction, spatial rearrangement of magnetic field positive and negative poles, or topological rotation of the magnetic field caused by subendocardial ischemia.

[0056] In one implementation, the magnetic field angle The QRS-T angle can be determined by the difference in planar coordinates between the source centroid and the sink centroid of the magnetic field. According to the QRS magnetic field angle Angle with T-wave magnetic field The difference is obtained after angular period normalization. In addition to angular features, features such as the spatial distribution of the positive and negative poles of the magnetic field, the topological direction of the dipole sample, the rotation direction of the magnetic field topology, the positional offset of the magnetic field extrema, and the synchronous changes of multi-channel waveforms can also be extracted. Magnetic field topological features are used to characterize the repolarization magnetic field vector reconstruction or spatial topological anomalies caused by subendocardial ischemia.

[0057] The decoding output module outputs the results of SERF-MCG magnetic field feature analysis or magnetic field topology anomaly indications related to subendocardial ischemia based on at least one of the following: vortex topology characterization, QRS-T angle, T-wave magnetic field angle shift, magnetic field topology rotation direction, and synchronous changes in multi-channel waveforms. The decoding results include, but are not limited to: ischemic injury current vortex field characterization information, vortex field modulus distribution information, vortex spatial distribution information, magnetic field topology rotation direction, T-wave magnetic field angle shift degree, QRS-T angle anomaly degree, subendocardial ischemia-related magnetic field feature indications, and SERF-MCG auxiliary assessment results.

[0058] In one specific application example, SERF-MCG data from healthy subjects and subjects with subendocardial ischemia can be input into the system. The QRS magnetic field vector and T-wave magnetic field vector are extracted separately, and the QRS-T angle distribution of the two types of subjects is compared. If the T-wave magnetic field vector of the target subject shows significant angle collapse or topological rotation anomaly relative to the QRS magnetic field vector, and its ischemic injury current vortex characteristics match those of the subendocardial ischemic region, the decoding output module outputs a subendocardial ischemia-related magnetic field topological anomaly indication. This indication is used to assist in SERF-MCG signal interpretation and feature analysis, and is not intended to replace the physician's final clinical diagnosis.

[0059] Through this embodiment, the hidden electrophysiological abnormalities caused by subendocardial ischemia can be characterized as source current vortex topology, and further quantitatively evaluated through features such as SERF-MCG magnetic field angle, spatial topology and QRS-T angle, thereby improving the physical consistency and traceability of SERF-MCG signal interpretation.

[0060] In summary, this invention can characterize the hidden electrophysiological abnormalities caused by subendocardial ischemia as source current vortex topology, and further quantitatively evaluate its macroscopic magnetic field performance through SERF-MCG magnetic field angle characteristics.

[0061] Example 2 like Figure 1 As shown, this invention provides a method for decoding the subendocardial ischemic magnetic field characteristics of SERF-MCG based on myocardial fiber anisotropy, which utilizes the system described in Example 1 and includes the following steps: Step 1: The model parameter acquisition module acquires at least one of the following parameters: cardiac geometric parameters, myocardial boundary parameters, trunk conductor parameters, forward computational domain parameters, SERF-MCG sensor array parameters, electrophysiological parameters, and subendocardial ischemic region parameters. Step 2: The myocardial anatomy coordinate construction module constructs a local myocardial anatomy coordinate system based on the cardiac geometric parameters and myocardial boundary parameters; Step 3: The myocardial fiber anisotropy construction module generates a myocardial fiber direction field based on the local myocardial anatomical coordinate system, and constructs anisotropic conduction parameters based on the myocardial fiber direction field; Step 4: The anisotropic electrophysiological propagation modeling module constructs a myocardial electrophysiological propagation model based on the anisotropic conduction parameters to obtain the spatiotemporal distribution of transmembrane potential and source current density; Step 5: The subendocardial ischemia source modeling module determines the subendocardial ischemia region based on the parameters of the subendocardial ischemia region, and performs pathological reconstruction on the electrophysiological parameters in the subendocardial ischemia region to obtain ischemia-related electrophysiological source items. Step 6: The eddy current feature extraction module extracts the source current eddy current topology features or ischemic injury current eddy current topology features based on the source current density or the injury current density determined by the ischemia-related electrophysiological source term. Step 7: The SERF-MCG magnetic field forward calculation module calculates the magnetic field distribution at the SERF-MCG sensor array; Step 8: The magnetic field topology feature extraction module extracts at least one of the following based on the magnetic field distribution: magnetic field angle features, magnetic field spatial topology features, QRS-T angle features, and magnetic field angle anomaly features. Step 9: The decoding output module outputs the decoding result of the subendocardial ischemic magnetic field feature based on at least one of the source current vortex topology features, the ischemic injury current vortex topology features, and the magnetic field topology features.

[0062] In this embodiment, the model parameter acquisition module acquires model parameters in the following manner: acquiring pre-constructed forward computation model parameters of the heart-trunk-sensor array, wherein the forward computation model parameters include at least one of the following: heart geometric model parameters, endocardial boundary parameters, epicardial boundary parameters, basal surface boundary parameters, trunk geometric model parameters, trunk conductor parameters, myocardial electrophysiological parameters, ischemic region parameters, and SERF-MCG sensor array parameters; wherein the SERF-MCG sensor array parameters include at least one of the following: number of sensor channels, sensor array arrangement, sensor channel position, and sensor measurement direction; and wherein the model parameter acquisition module forms unified input parameters for myocardial anatomical coordinate construction, myocardial fiber anisotropy construction, electrophysiological propagation modeling, ischemia source modeling, and magnetic field forward computation based on the forward computation model parameters.

[0063] In this embodiment, the local myocardial anatomical coordinate system is constructed in the following manner: (1) Solve the continuous scalar field from the endocardium to the epicardium or from the epicardium to the endocardium within the myocardial calculation region based on the endocardial boundary and the epicardial boundary, and determine the transmural direction based on the continuous scalar field; (2) Determine the cardiac anatomical major axis based on the apical reference point and the basal reference point, and use vector projection to obtain the normalized coordinates of the major axis within the myocardial calculation area; (3) Construct a local orthogonal basis according to the wall-penetrating direction and the major axis direction, wherein the local orthogonal basis includes the wall-penetrating basis vector, the major axis basis vector and the circumferential basis vector; (4) The local orthogonal basis is used to provide local anatomical reference for the calculation of myocardial fiber orientation field, anisotropic conduction parameters and electric vortex topology features.

[0064] In this embodiment, the myocardial fiber orientation field and anisotropic conduction parameters are constructed in the following manner: (1) Determine the transmural coordinates of the myocardium based on the endocardial distance, epicardial distance, or the continuous scalar field; (2) Determine the distribution of myocardial fiber helix angles based on the transmural coordinates and the normalized major axis coordinates; (3) Rotate the local orthogonal base around the transmural direction according to the helical angle of the myocardial fiber to obtain the myocardial fiber orientation field; (4) Construct an anisotropic diffusion tensor or anisotropic conductivity tensor based on the myocardial fiber direction field, so that electrophysiological excitation has different propagation characteristics in the myocardial fiber direction and non-fiber direction; (5) The anisotropic diffusion tensor or anisotropic conductivity tensor is used for electrophysiological propagation model and source current density calculation.

[0065] In this embodiment, the subendocardial ischemia source modeling module constructs the subendocardial ischemic region and ischemia-related electrophysiological source terms in the following manner: (1) Determine the spherical or ellipsoidal local candidate region based on the ischemic center, ischemic radius, and spatial constraint function, and constrain the local candidate region using the wall-penetrating coordinates and wall-penetrating threshold according to equation (20); (2) Based on the transmural coordinates, the candidate region of local ischemia is located in the subendocardial layer, thus obtaining the subendocardial ischemic region; (3) Pathological reconstruction of at least one of the following parameters in the subendocardial ischemic region: resting potential, excitation threshold, action potential amplitude parameter, recovery parameter, diffusion parameter and local conductivity; (4) Calculate the spatiotemporal distribution of transmembrane potential, source current density, or damage current density based on the reconstructed electrophysiological parameters; (5) The ischemia-related electrophysiological source term includes at least the injury current density caused by increased resting potential, decreased conduction parameters or changes in local conductivity.

[0066] In this embodiment, the eddy current feature extraction module extracts the source current eddy current topology features or ischemic injury current eddy current topology features in the following manner: (1) Calculate the source current density based on the anisotropic conduction parameters and the transmembrane potential spatial gradient; (2) Obtain the source current vortex field based on the spatial curl of the source current density, or obtain the ischemic injury current vortex field based on the spatial curl of the injury current density. (3) Based on the source current vortex field or the ischemic injury current vortex field, obtain at least one of the following computational characteristics: vortex field modulus distribution, directional distribution, spatial distribution, boundary annular distribution, local extreme value location, and vortex topological continuity. (4) The source current vortex topology features or ischemic injury current vortex topology features are used to characterize the redirection effect of myocardial fiber anisotropy on the propagation path of ischemia-related injury current.

[0067] In this embodiment, the SERF-MCG magnetic field forward calculation module calculates the magnetic field distribution in the following manner: The myocardial electrophysiological source term is established based on the source current density or the injury current density, the body conductor return current term is established based on the trunk body conductor parameters, and the magnetic field distribution at the SERF-MCG sensor array is calculated under quasi-static electromagnetic approximation conditions; the magnetic field distribution includes at least one of the following: multi-channel spatial magnetic field map at a single target time, temporal magnetic field distribution at multiple target times, magnetic field distribution corresponding to QRS composite wave, and magnetic field distribution corresponding to T wave.

[0068] In this embodiment, the magnetic field topology feature extraction module extracts magnetic field topology features in the following manner: (1) Determine the magnetic field source region and magnetic field sink region based on the magnetic field distribution at the SERF-MCG sensor array; (2) Determine the direction of the magnetic field vector or the magnetic field angle based on the positional relationship between the magnetic field source region and the magnetic field sink region; (3) Extract the QRS magnetic field vector and the T-wave magnetic field vector within the QRS composite wave time window and the T-wave time window, respectively; (4) Calculate at least one of the following based on the QRS magnetic field vector and the T-wave magnetic field vector: QRS-T angle, T-wave magnetic field angle offset, magnetic field angle collapse degree, and magnetic field topological rotation characteristics; (5) The magnetic field topology features are used to characterize the repolarization magnetic field vector reconstruction or spatial topology anomaly caused by subendocardial ischemia.

[0069] In this embodiment, the decoding output module outputs the decoding results of the subendocardial ischemic magnetic field characteristics in the following manner: Based on at least one of the source current vortex topology features, the ischemic injury current vortex topology features, the magnetic field angle features, the magnetic field spatial topology features, and the QRS-T angle features, a magnetic field feature decoding result is formed; the magnetic field feature decoding result includes at least one of the following: ischemic injury current vortex field characterization information, vortex field spatial distribution information, magnetic field topology rotation direction, QRS-T angle abnormality degree, T wave magnetic field angle offset degree, endocardial ischemia-related magnetic field feature indication information, and SERF-MCG auxiliary assessment results.

[0070] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A SERF-MCG subendocardial ischemic magnetic field feature decoding system based on myocardial fiber anisotropy, characterized in that, The system includes: a model parameter acquisition module, a myocardial anatomical coordinate construction module, a myocardial fiber anisotropy construction module, an anisotropic electrophysiological propagation modeling module, a subendocardial ischemia source modeling module, an electric vortex feature extraction module, a SERF-MCG magnetic field forward calculation module, a magnetic field topology feature extraction module, and a decoding output module. The model parameter acquisition module is used to acquire at least one of the following: cardiac geometric parameters, myocardial boundary parameters, trunk conductor parameters, forward computational domain parameters, SERF-MCG sensor array parameters, electrophysiological parameters, and subendocardial ischemic region parameters. The myocardial anatomy coordinate construction module is used to construct a local myocardial anatomy coordinate system based on cardiac geometric parameters and myocardial boundary parameters; The myocardial fiber anisotropy construction module is used to generate a myocardial fiber direction field based on the local myocardial anatomical coordinate system, and to construct anisotropic conduction parameters based on the myocardial fiber direction field. The anisotropic electrophysiological propagation modeling module is used to construct a myocardial electrophysiological propagation model based on anisotropic conduction parameters, and obtain the spatiotemporal distribution of transmembrane potential and source current density in the myocardial computational region. The subendocardial ischemia source modeling module is used to determine the subendocardial ischemia region within the myocardial calculation region based on the parameters of the subendocardial ischemia region, and to pathologically reconstruct the electrophysiological parameters within the subendocardial ischemia region to obtain ischemia-related electrophysiological source terms. The eddy current feature extraction module is used to extract source current eddy current topology features or ischemic injury current eddy current topology features based on the source current density or the injury current density determined by the ischemia-related electrophysiological source term. The SERF-MCG magnetic field forward calculation module is used to calculate the magnetic field distribution at the SERF-MCG sensor array based on at least one of the source current density, damage current density, trunk conductor parameters, forward calculation domain parameters, and SERF-MCG sensor array parameters. The magnetic field topology feature extraction module is used to extract magnetic field topology features based on the magnetic field distribution; The decoding output module is used to output the decoding result of the subendocardial ischemic magnetic field feature based on at least one of the source current vortex topology feature, the ischemic injury current vortex topology feature, and the magnetic field topology feature.

2. A method for decoding the subendocardial ischemic magnetic field characteristics of SERF-MCG based on myocardial fiber anisotropy, wherein the method is implemented using the system described in claim 1, characterized in that... The method includes: Step 1: Obtain at least one of the following parameters: cardiac geometry, myocardial boundary, trunk conductor, forward computation domain, SERF-MCG sensor array, electrophysiological parameters, and subendocardial ischemic region parameters. Step 2: Construct a local myocardial anatomical coordinate system based on cardiac geometric parameters and myocardial boundary parameters; Step 3: Generate the myocardial fiber orientation field based on the local myocardial anatomical coordinate system, and construct anisotropic conduction parameters based on the myocardial fiber orientation field; Step 4: Construct a myocardial electrophysiological propagation model based on anisotropic conduction parameters to obtain the spatiotemporal distribution of transmembrane potentials and source current density; Step 5: Determine the subendocardial ischemic region based on the parameters of the subendocardial ischemic region, and perform pathological reconstruction of the electrophysiological parameters in the subendocardial ischemic region to obtain ischemia-related electrophysiological source terms; Step 6: Extract source current vortex topology features or ischemic injury current vortex topology features based on the source current density or the injury current density determined by the ischemia-related electrophysiological source terms. Step 7: Calculate the magnetic field distribution at the SERF-MCG sensor array based on at least one of the source current density, damage current density, trunk conductor parameters, forward computation domain parameters, and SERF-MCG sensor array parameters. Step 8: Extract the topological features of the magnetic field based on the magnetic field distribution; Step 9: Based on at least one of the source current vortex topology features, ischemic injury current vortex topology features, and magnetic field topology features, output the decoding result of the subendocardial ischemic magnetic field features.

3. The method according to claim 2, characterized in that, Methods for constructing a local myocardial anatomical coordinate system based on cardiac geometric parameters and myocardial boundary parameters include: Based on the endocardial and epicardial boundaries, a scalar field that continuously changes from the endocardium to the epicardium is solved within the myocardial computation region, and the transmural direction is determined based on the gradient of the scalar field. The cardiac anatomical long axis is determined based on the apical reference point and the basal reference point, and the normalized coordinates of the long axis within the myocardial computational region are obtained using vector projection. A local orthogonal basis is constructed based on the wall-penetrating direction and the major axis direction. The local orthogonal basis includes the wall-penetrating basis vector, the major axis basis vector, and the circumferential basis vector.

4. The method according to claim 3, characterized in that, Methods for generating myocardial fiber orientation fields based on local myocardial anatomical coordinates and constructing anisotropic conduction parameters based on these orientation fields include: Determine the transmural coordinates of the myocardium based on the endocardial distance, epicardial distance, or a continuous scalar field. The distribution of myocardial fiber helix angles was determined based on myocardial transmural coordinates and normalized major axis coordinates. The local orthogonal basis is rotated around the transmural direction according to the helical angle of the myocardial fiber to obtain the myocardial fiber directional field; Anisotropic diffusion tensors or anisotropic conductivity tensors are constructed based on the myocardial fiber directional field, so that electrophysiological excitation has different propagation characteristics in the myocardial fiber direction and non-fiber direction.

5. The method according to claim 4, characterized in that, Methods for determining the subendocardial ischemic region based on its parameters, and for pathologically reconstructing the electrophysiological parameters within the subendocardial ischemic region to obtain ischemia-related electrophysiological parameters, include: Based on at least one of the ischemia center, ischemia radius, spatial constraint function and transmural threshold, determine the local ischemia candidate region within the myocardial computational region; Based on the transmural coordinates of the myocardium, the candidate region for local ischemia is located in the subendocardial layer, thus obtaining the subendocardial ischemic region; In the subendocardial ischemic region, at least one of the following parameters is pathologically reconstructed: resting potential, excitation threshold, action potential amplitude parameter, recovery parameter, diffusion parameter, and local conductivity. The spatiotemporal distribution of transmembrane potential, source current density, or injury current density are calculated based on the reconstructed electrophysiological parameters, i.e., ischemia-related electrophysiological source terms.

6. The method according to claim 5, characterized in that, Methods for extracting source current vortex topology features or ischemic injury current vortex topology features based on source current density or injury current density determined by ischemia-related electrophysiological source terms include: The source current density is calculated based on the anisotropic conduction parameters and the spatial gradient of the transmembrane potential. The source current vortex field can be obtained from the spatial curl of the source current density, or the ischemic injury current vortex field can be obtained from the spatial curl of the injury current density. Based on the vortex field of the source current or the vortex field of the ischemic injury current, at least one of the following computational characteristics is obtained: vortex field modulus distribution, directional distribution, spatial distribution, boundary annular distribution, local extremum location, and vortex topological continuity.

7. The method according to claim 6, characterized in that, Methods for extracting magnetic field topological features based on magnetic field distribution include: The magnetic field source region and magnetic field sink region are determined based on the magnetic field distribution at the SERF-MCG sensor array; The direction of the magnetic field vector or the magnetic field angle is determined based on the positional relationship between the magnetic field source region and the magnetic field sink region. The QRS magnetic field vector and T-wave magnetic field vector were extracted within the QRS composite wave time window and the T-wave time window, respectively. Calculate at least one of the following based on the QRS magnetic field vector and the T-wave magnetic field vector: QRS-T angle, T-wave magnetic field angle offset, magnetic field angle collapse degree, and magnetic field topological rotation characteristics.

8. The method according to claim 7, characterized in that, The magnetic field feature decoding results include at least one of the following: ischemic injury current vortex field characterization information, vortex field spatial distribution information, magnetic field topological rotation direction, QRS-T angle abnormality degree, T wave magnetic field angle deviation degree, endocardial ischemia-related magnetic field feature indication information, and SERF-MCG auxiliary assessment results.

Images