A SERF-MCG heart motion-induced measurement error quantification system and method based on electrophysiological-mechanical-electromagnetic coupling

By calculating the difference in magnetic field distribution between static and dynamic cardiac geometry in the SERF-MCG system and extracting motion-induced magnetic field artifacts, the problem of distinguishing between changes in myocardial electrophysiological sources and changes in cardiac biomechanical motion measurements was solved, and the measurement error of SERF-MCG was quantified and artifacts were suppressed.

CN123058736APending Publication Date: 2026-11-10BEIHANG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611318286.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-28
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing SERF-MCG forward modeling methods have difficulty distinguishing between changes in myocardial electrophysiological sources and measurement changes caused by cardiac biomechanics. In particular, within the time window of low-amplitude magnetic field characteristics, changes in the magnetic field caused by cardiac biomechanics may obscure the original weak magnetic characteristics, affecting the interpretation of measurement results and error assessment.

Method used

Under the same myocardial electrophysiological source terms and sensor array conditions, the magnetic field distribution corresponding to static and dynamic cardiac geometry was calculated, and the influence of cardiac biomechanics on the measurement results was quantified by extracting motion-induced magnetic field artifacts by differential extraction.

Benefits of technology

The influence of cardiac biomechanics on SERF-MCG measurement results was quantified, the spatial distribution of motion-induced measurement errors was analyzed, and the superposition or cancellation effect of motion-induced magnetic field artifacts on the target magnetic field characteristics was evaluated, providing a basis for measurement error analysis and artifact suppression methods.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123058736A_ABST
    Figure CN123058736A_ABST
Patent Text Reader

Abstract

The application discloses a SERF-MCG heart motion induced measurement error quantification system and method based on electrophysiology-mechanics-electromagnetic coupling, which calculates a first magnetic field distribution under a static heart geometry and a second magnetic field distribution under a dynamic heart geometry under the same myocardial electrophysiology source term and the same SERF-MCG sensor array condition, and obtains a heart motion induced magnetic field artifact according to the difference between the second magnetic field distribution and the first magnetic field distribution; further extracts the amplitude, polarity, spatial topology, channel response and target time window response and other measurement error characteristics of the magnetic field artifact, and outputs a SERF-MCG motion induced measurement error evaluation result. The application can quantify the influence of the heart mechanics motion on the SERF-MCG magnetic field forward calculation result, and provides a basis for the measurement error evaluation and channel response difference analysis in the low-amplitude magnetic field characteristic time window.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the fields of biomagnetic measurement, magnetocardiography signal analysis and multiphysics forward modeling technology, specifically involving a SERF-MCG cardiac motion induction measurement error quantification system and method based on electrophysiological-mechanical-electromagnetic coupling. Background Technology

[0002] Magnetocardiography (MCG) is a biomagnetic measurement technique used to non-invasively record the weak magnetic field signals generated by cardiac electrical activity. With the development of atomic magnetometer technology, especially the application of spin exchange relaxation free magnetometers in weak magnetic field measurements, the SERF-MCG system can acquire cardiac magnetic signals near the body surface under multi-channel array conditions, providing a new measurement method for the spatial distribution analysis of cardiac electrical activity.

[0003] During SERF-MCG measurements, the magnetic field distribution acquired by the sensor array is not only related to the myocardial electrophysiological source but also closely related to the spatial relationship between the heart, the trunk conductors, and the sensor array. The heart is not a static structure during a cardiac cycle but undergoes mechanical movements such as contraction, relaxation, torsion, and changes in wall thickness. These cardiac mechanical movements alter the relative position and spatial coupling between the myocardial electrophysiological source and the sensor array, thus introducing measurement errors caused by geometric changes into the surface magnetic field measurement results.

[0004] Existing forward modeling methods for myocardial vascular mechanics (MCG) primarily focus on the mapping process from myocardial electrophysiological sources to the surface magnetic field, typically calculating the magnetic field distribution given cardiac geometry, trunk conductors, and sensor arrays. However, in many forward computational flows, cardiac geometry is often treated as a static structure, or the measurement changes caused by changes in electrophysiological sources and cardiac biomechanical motion are not explicitly distinguished. While such static forward models can describe the magnetic field distribution generated by myocardial electrical activity, they struggle to quantify the independent impact of cardiac contractile motion on SERF-MCG measurements.

[0005] Especially within the low-amplitude magnetic field characteristic time window, such as the ST segment, the target magnetic field characteristic itself is relatively weak. Changes in the magnetic field induced by cardiac biomechanics may superimpose or cancel out the target magnetic field characteristic. When their polarities are opposite, the motion-induced magnetic field change may obscure the original weak magnetic characteristic, causing a shift in the observed SERF-MCG magnetic field distribution or waveform, thus affecting the interpretation of the measurement results. Because these motion-induced changes have spatial distribution characteristics and channel differences, it is difficult to effectively distinguish and quantify them based solely on a single channel waveform or static forward results.

[0006] Therefore, a method is needed that can calculate the SERF-MCG magnetic field distribution corresponding to static and dynamic cardiac geometry under the same myocardial electrophysiological source terms and the same sensor array conditions, and extract cardiac motion-induced magnetic field artifacts by the difference between the two, so as to quantify the influence of cardiac biomechanics on SERF-MCG measurement results, and provide a basis for measurement error assessment, channel response difference analysis and subsequent artifact suppression method design within the weak magnetic characteristic time window. Summary of the Invention

[0007] To address the difficulty in distinguishing between changes in cardiac electrophysiological sources and changes in cardiac mechanical motion-induced measurements in existing SERF-MCG forward modeling and measurement error analysis, this invention provides a SERF-MCG cardiac motion-induced measurement error quantification system and method based on electrophysiological-mechanical-electromagnetic coupling. The aim is to calculate the first magnetic field distribution without cardiac mechanical motion and the second magnetic field distribution with cardiac mechanical motion under the same myocardial electrophysiological source terms and the same SERF-MCG sensor array conditions, and extract cardiac motion-induced magnetic field artifacts based on the difference between the second and first magnetic field distributions, thereby achieving a quantitative assessment of the SERF-MCG measurement error caused by cardiac mechanical motion.

[0008] To achieve the above objectives, the present invention provides the following solution: A SERF-MCG cardiac motion-induced measurement error quantification system based on electrophysiological-mechanical-electromagnetic coupling, the system comprising: a model parameter acquisition module, a myocardial electrophysiological source modeling module, a cardiac mechanical motion modeling module, a static magnetic field calculation module, a dynamic magnetic field calculation module, a motion artifact extraction module, a measurement error feature extraction module, and an error evaluation output module; The model parameter acquisition module is used to acquire the forward computation model parameters of the heart-trunk-sensor array. The model parameters include at least one of the following: heart geometric parameters, trunk conductor parameters, forward computation domain parameters, and SERF-MCG sensor array parameters. The myocardial electrophysiological source modeling module is used to construct an anisotropic electrophysiological source model of the myocardium based on cardiac geometric parameters, and to obtain myocardial electrophysiological source terms for SERF-MCG magnetic field forward calculation based on the anisotropic electrophysiological source model of the myocardium. The cardiac biomechanics modeling module is used to construct an active stress-driven cardiac biomechanics model based on the myocardial electrophysiological source model, and to obtain dynamic cardiac geometry at at least one target time point based on the cardiac biomechanics model. The static magnetic field calculation module is used to calculate the first magnetic field distribution at the SERF-MCG sensor array under static cardiac geometry conditions without introducing cardiac mechanical motion, based on myocardial electrophysiological source terms, trunk conductor parameters, forward computational domain parameters, and SERF-MCG sensor array parameters. The dynamic magnetic field calculation module is used to calculate the second magnetic field distribution at the SERF-MCG sensor array under dynamic cardiac geometry conditions that introduce cardiac biomechanical motion, based on myocardial electrophysiological source terms, trunk conductor parameters, forward computational domain parameters, and SERF-MCG sensor array parameters. The motion artifact extraction module is used to obtain cardiac motion-induced magnetic field artifacts based on the difference between the second magnetic field distribution and the first magnetic field distribution. The measurement error feature extraction module is used to extract the measurement error features of the cardiac motion-induced magnetic field artifact; The error assessment output module is used to output the SERF-MCG motion-induced measurement error assessment result based on the measurement error characteristics.

[0009] This invention also provides a method for quantifying the error of SERF-MCG cardiac exercise induction measurement based on electrophysiological-mechanical-electromagnetic coupling. The method is implemented using the aforementioned system and includes: Step 1: Obtain model parameters for the heart, trunk, and SERF-MCG sensor array. The model parameters include at least one of the following: heart geometric parameters, trunk body conductor parameters, forward computational domain parameters, and SERF-MCG sensor array parameters. Step 2: Construct a myocardial anisotropic electrophysiological source model based on the cardiac geometric parameters, and obtain the myocardial electrophysiological source terms for SERF-MCG magnetic field forward calculation based on the myocardial anisotropic electrophysiological source model; Step 3: Construct an active stress-driven cardiac biomechanical motion model based on the myocardial electrophysiological source model, and obtain the dynamic cardiac geometry at at least one target time point based on the cardiac biomechanical motion model. Step 4: Under static cardiac geometry conditions without introducing cardiac biomechanical motion, calculate the first magnetic field distribution at the SERF-MCG sensor array based on the myocardial electrophysiological source terms, the trunk conductor parameters, the forward computational domain parameters, and the SERF-MCG sensor array parameters. Step 5: Under the dynamic cardiac geometry conditions that introduce cardiac biomechanics, the second magnetic field distribution at the SERF-MCG sensor array is calculated based on the myocardial electrophysiological source terms, the trunk conductor parameters, the forward computation domain parameters, and the SERF-MCG sensor array parameters. Step 6: Obtain the cardiac motion-induced magnetic field artifact based on the difference between the second magnetic field distribution and the first magnetic field distribution; Step 7: Extract the measurement error characteristics of the cardiac motion-induced magnetic field artifact; Step 8: Output the SERF-MCG motion-induced measurement error evaluation result based on the measurement error characteristics.

[0010] Preferably, the method for constructing anisotropic electrophysiological source models of the myocardium based on the cardiac geometric parameters includes: The myocardial computation region is determined based on a pre-constructed cardiac geometry model, and a local anatomical coordinate system is constructed within the myocardial computation region; The direction of myocardial fibers is determined based on the local anatomical coordinate system, and myocardial anisotropic conduction parameters are constructed based on the direction of myocardial fibers. Based on the anisotropic conduction parameters of the myocardium, an electrophysiological propagation model of the myocardium was established to obtain the spatiotemporal distribution of transmembrane potentials within the computational region of the myocardium. Based on the spatiotemporal distribution of the transmembrane potential and the anisotropic conduction parameters of the myocardium, myocardial electrophysiological source terms for forward calculation of the SERF-MCG magnetic field are obtained.

[0011] Preferably, the method for constructing an actively stress-driven cardiac biomechanical motion model based on a myocardial electrophysiological source model includes: The electrophysiological activation state of the myocardium within the computational region was obtained based on the myocardial electrophysiological source model; Active myocardial stress is generated based on the electrophysiological activation state; Based on the active stress and dynamic parameters of myocardium, a cardiac mechanical motion model is established, and the displacement field or deformation field of myocardial tissue at the target time point is calculated. The cardiac geometry is updated based on the displacement field or deformation field to obtain the dynamic cardiac geometry after incorporating cardiac biomechanical motion.

[0012] Preferably, the method for obtaining dynamic cardiac geometry at at least one target time point based on the cardiac biomechanics model includes: The direction of the long axis of the heart is determined based on a pre-constructed geometric model of the heart, and an anatomical auxiliary reference system is established using the direction of the long axis of the heart as a reference direction. Based on the aforementioned anatomical auxiliary reference system, the cardiac calculation region is spatially divided to obtain a cardiac sub-region that includes at least the left ventricular region and the right ventricular region. Corresponding contraction parameters are set for the left ventricular region and the right ventricular region respectively, so that the left ventricular region and the right ventricular region have different active contraction intensities in the cardiac biomechanics model; The asymmetric ventricular contraction motion is calculated based on the different active contraction intensities to obtain dynamic cardiac geometry that includes the differences in motion between the left and right ventricles.

[0013] Preferably, the method for obtaining cardiac motion-induced magnetic field artifacts based on the difference between the second magnetic field distribution and the first magnetic field distribution includes: Select at least one target time point and obtain the first magnetic field distribution and the second magnetic field distribution corresponding to the target time point respectively; By performing channel-correspondence difference between the second magnetic field distribution and the first magnetic field distribution, motion-induced magnetic field artifacts at each SERF-MCG sensor channel are obtained. When the first magnetic field distribution and the second magnetic field distribution are multi-channel spatial magnetic field distributions, the differential results at each sensor channel are arranged according to the spatial position of the sensor array to form a spatial distribution of motion-induced magnetic field artifacts. When the first magnetic field distribution and the second magnetic field distribution are temporal magnetic field distributions at multiple target time points, the difference results at each target time point are temporally arranged to obtain the spatiotemporal distribution of motion-induced magnetic field artifacts.

[0014] Preferably, the method for extracting the measurement error characteristics of the cardiac motion-induced magnetic field artifact includes: Based on the difference results of the motion-induced magnetic field artifacts at each SERF-MCG sensor channel, the amplitude characteristics of the motion-induced magnetic field artifacts are extracted. Based on the directional relationship between the motion-induced magnetic field artifact and the first magnetic field distribution or target magnetic field characteristics, the polarity characteristics of the motion-induced magnetic field artifact are extracted. Based on the spatial distribution of the motion-induced magnetic field artifacts on the SERF-MCG sensor array, the spatial topological features of the motion-induced magnetic field artifacts are extracted. Based on the amplitude and waveform changes of motion-induced magnetic field artifacts at different SERF-MCG sensor channels, channel response characteristics are extracted; A set of SERF-MCG motion-induced measurement error features is formed based on at least one of amplitude features, polarity features, spatial topology features, and channel response features.

[0015] Preferably, the method for outputting the SERF-MCG motion-induced measurement error evaluation result based on the measurement error characteristics includes: Select a target time window from the SERF-MCG forward calculation results. The target time window includes the ST segment time window or the low-amplitude magnetic field characteristic time window. Acquire the first magnetic field distribution, the second magnetic field distribution, and the cardiac motion-induced magnetic field artifacts within the target time window; Based on the amplitude and polarity relationship between the cardiac motion-induced magnetic field artifact and the first magnetic field distribution, determine the superposition or cancellation effect of the cardiac motion-induced magnetic field artifact on the target magnetic field characteristics. When the cardiac motion-induced magnetic field artifact has opposite polarity to the target magnetic field feature, the degree of obscuring of the target magnetic field feature is calculated based on the amplitude of the cardiac motion-induced magnetic field artifact and the amplitude of the target magnetic field feature. Based on the degree of obscuring of the target magnetic field characteristics, the SERF-MCG motion-induced measurement error assessment result within the target time window is output.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) Under the same myocardial electrophysiological source terms and the same SERF-MCG sensor array, the present invention calculates the magnetic field distribution corresponding to static cardiac geometry and dynamic cardiac geometry respectively, and can separate the magnetic field measurement changes caused by cardiac mechanical motion from the SERF-MCG forward calculation results; (2) The present invention obtains cardiac motion-induced magnetic field artifacts by the difference between the second magnetic field distribution and the first magnetic field distribution, which can quantify the influence of cardiac biomechanics on the SERF-MCG magnetic field measurement results; (3) The present invention can extract the measurement error characteristics such as amplitude, polarity, spatial topology and channel response of cardiac motion-induced magnetic field artifacts, which helps to analyze the spatial distribution law of motion-induced measurement errors on the sensor array; (4) The present invention can evaluate the superposition or cancellation effect of motion-induced magnetic field artifacts on the target magnetic field characteristics for the ST segment time window or low amplitude magnetic field characteristic time window, thereby quantifying the weak magnetic feature masking effect caused by cardiac mechanical motion. (5) The present invention can compare the response of different SERF-MCG sensor channels to cardiac mechanical motion-induced measurement errors, providing a basis for SERF-MCG measurement error analysis, channel reliability evaluation and subsequent artifact suppression method design. Attached Figure Description

[0017] 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.

[0018] Figure 1 This is a flowchart of the SERF-MCG cardiac motion induction measurement error quantification method based on electrophysiological-mechanical-electromagnetic coupling of the present invention; Figure 2This is a schematic diagram of the module structure of the SERF-MCG cardiac motion induction measurement error quantification system of the present invention; Figure 3 This is a schematic diagram of the static magnetic field distribution, dynamic magnetic field distribution, cardiac motion-induced magnetic field artifacts and channel response differences of the present invention, wherein (A) is a schematic diagram of the static / dynamic magnetic field difference relationship, and (B) is a schematic diagram of the channel response differences. Detailed Implementation

[0019] 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.

[0020] 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.

[0021] This invention discloses a system and method for quantifying the error of SERF-MCG cardiac motion-induced measurement based on electrophysiological-mechanical-electromagnetic coupling. The system calculates the first magnetic field distribution without cardiac mechanical motion and the second magnetic field distribution with cardiac mechanical motion under the same myocardial electrophysiological source terms and the same SERF-MCG sensor array. The difference between the second and first magnetic field distributions is used to obtain the cardiac motion-induced magnetic field artifact, thereby achieving a quantitative assessment of the SERF-MCG motion-induced measurement error. Figure 1 As shown, the SERF-MCG cardiac motion-induced measurement error quantification method of the present invention includes: obtaining forward computation model parameters, constructing a myocardial anisotropic electrophysiological source model, constructing an active stress-driven cardiac biomechanical motion model, calculating the first magnetic field distribution under static cardiac geometry and the second magnetic field distribution under dynamic cardiac geometry respectively, obtaining cardiac motion-induced magnetic field artifacts through difference, and further extracting measurement error features and outputting motion-induced measurement error evaluation results.

[0022] Example 1: System Overall Structure like Figure 2As shown, this invention discloses a SERF-MCG cardiac motion-induced measurement error quantification system based on electrophysiological-mechanical-electromagnetic coupling, comprising a model parameter acquisition module, a myocardial electrophysiological source modeling module, a cardiac mechanical motion modeling module, a static magnetic field calculation module, a dynamic magnetic field calculation module, a motion artifact extraction module, a measurement error feature extraction module, and an error evaluation output module. The model parameter acquisition module acquires pre-constructed forward computation model parameters of the heart-trunk-sensor array. These forward computation model parameters include at least one of cardiac geometric model parameters, trunk geometric model parameters, myocardial electrophysiological parameters, cardiac mechanical motion parameters, and SERF-MCG sensor array parameters. The SERF-MCG sensor array parameters include at least one of the following: the number of sensor channels, the sensor array arrangement, the sensor channel positions, and the sensor measurement directions.

[0023] The model parameter acquisition module uses the torso coordinate system as a unified spatial coordinate system and transforms the heart and sensor coordinates to this unified spatial coordinate system through rigid body transformation. For the k-th sensor, its position in the unified coordinate system is represented as: (1) in, Let k be the position of the k-th sensor in a unified coordinate system. This refers to the sensor's position in its local coordinate system. For coordinate rotation matrix, This is a translation vector. The model parameter acquisition module generates the target time series according to a uniform time step Δt. =qΔt, and the unified parameters are passed to the myocardial electrophysiological source modeling module, cardiac biomechanical motion modeling module, static magnetic field calculation module and dynamic magnetic field calculation module respectively.

[0024] The myocardial electrophysiological source modeling module constructs an anisotropic electrophysiological source model of the myocardium based on cardiac geometric model parameters and myocardial electrophysiological parameters. Specifically, the module determines the myocardial calculation region based on the pre-constructed cardiac geometric model, constructs a local anatomical coordinate system within the calculation region, determines the direction of myocardial fibers based on the local anatomical coordinate system, and further constructs anisotropic conduction parameters of the myocardium. The module establishes a myocardial electrophysiological propagation model based on the anisotropic conduction parameters, calculates the spatiotemporal distribution of transmembrane potentials, and obtains myocardial electrophysiological source terms for forward calculation of the SERF-MCG magnetic field based on the spatiotemporal distribution of transmembrane potentials and the anisotropic conduction parameters of the myocardium. These myocardial electrophysiological source terms include at least a source current density determined by the local conduction parameters and the spatial gradient of the transmembrane potential, ensuring that the forward calculation of the SERF-MCG magnetic field is driven by myocardial electrophysiological activity.

[0025] The specific implementation process is as follows: First, based on the endocardial boundary and epicardial boundary Determine the myocardial computational region under the reference configuration Solve the wall thickness scalar field ψ( within the myocardial computation region). ): (2) (3) (4) Where, ψ( ) represents the wall thickness scalar field under the reference configuration. and These represent the endocardial boundary and the epicardial boundary, respectively. A local anatomical coordinate system is constructed based on the gradient of the wall thickness scalar field, with the wall thickness direction represented as: (5) Unit vector of the heart's major axis Project to Orthogonal local tangent planes are used to obtain the major axis direction: (6) The circumferential direction is represented as: (7) The orientation of myocardial fibers is determined based on the wall thickness direction, long axis direction, and circumferential direction. (8) Wherein, α(ψ) is the helix angle of the myocardial fiber that varies with the direction of wall thickness.

[0026] Construct the myocardial anisotropy tensor based on the direction of myocardial fibers: (9) in, The isotropic transmission coefficient, Add a conductivity coefficient in the fiber direction. It is a second-order unit tensor. It is the second-order tensor product along the direction of myocardial fibers.

[0027] The myocardial electrophysiology source modeling module is based on the anisotropic conduction tensor. Establish a myocardial electrophysiological propagation model and solve for transmembrane potentials within the myocardial computational region. ( The spatiotemporal distribution of ,t) and its state variable ξ. The transmembrane potential is obtained by response-diffusion calculations using an electrophysiological propagation model: (10) in, ( ,ξ) represents the ionic reaction term. ξ is a preset excitation term, and ξ is an electrophysiological state variable.

[0028] Based on the spatial gradient of the transmembrane potential and the local myocardial conduction tensor, the myocardial electrophysiological source terms under the reference configuration are obtained: (11) Among them, J impref (X,t) represents the source current density under the reference configuration, σ local (X) represents the local myocardial conduction tensor, V m (X,t) represents the transmembrane potential, and ∇ represents the spatial gradient operator relative to the reference coordinate X.

[0029] The myocardial anisotropic electrophysiological source model is used to calculate the transmembrane potential V. m (X,t) and electrophysiological state variables ξ, and obtain the source current density J under the reference configuration based on the spatial gradient of the transmembrane potential. impref (X,t); the myocardial electrophysiological source term is the source current density J. impref (X,t).

[0030] The cardiac biomechanical modeling module constructs an active stress-driven cardiac biomechanical model based on the myocardial electrophysiological source model. Specifically, the module generates active myocardial stress based on the transmembrane potential distribution or the electrophysiological activation state determined by the transmembrane potential distribution, and establishes a cardiac biomechanical model based on the active myocardial stress and myocardial kinetic parameters, calculating the displacement or deformation field of myocardial tissue at a target time point. The electrophysiological activation state includes the transmembrane potential distribution or the electrophysiological activation state determined by the transmembrane potential distribution, and the active myocardial stress characterizes the active contraction of myocardial tissue caused by electrophysiological activation. The module updates the cardiac geometry based on the displacement or deformation field, obtaining the dynamic cardiac geometry after incorporating cardiac biomechanical motion. This dynamic cardiac geometry is used by the dynamic magnetic field calculation module to calculate the second magnetic field distribution at the SERF-MCG sensor array.

[0031] The specific implementation process is as follows: Based on transmembrane potential ( Active stress is generated by the electrophysiological activation state determined by the transmembrane potential (t) or by the transmembrane potential. The active stress evolution equation is: (12) (13) in, The time constant for active contraction and relaxation. The peak shrinkage strength coefficient, It is an electrophysiological activation function. To determine the steepness of the activation function, This is the activation threshold. Myocardial tissue is located at a reference conformational position. Current position at target time t The motion mapping is as follows: (14) Where χ(X,t) is the motion mapping of myocardial tissue from the reference configuration to the current configuration at time t, X is the position vector under the reference configuration, x is the position vector under the current configuration, and u(X,t) is the myocardial displacement field. The corresponding deformation gradient and right Cauchy-Green deformation tensor are as follows: (15) (16) Where F is the deformation gradient tensor, I is the second-order unit tensor, ∇Xu is the gradient of the displacement field relative to the reference coordinate X, and C is the right Cauchy-Green deformation tensor. T It is the transpose of the deformable gradient tensor F.

[0032] The cardiac biomechanics modeling module constructs a total second Piola-Kirchhoff stress tensor S based on the myocardial passive strain energy function, active stress Ta, and approximately incompressible constraints. Then, neglecting body forces, it solves for the displacement field u(X,t) using the momentum balance equation under the reference configuration. Where F is the deformation gradient tensor, S is the total second Piola-Kirchhoff stress tensor including passive stress, active stress, and approximately incompressible constraint stress, FS is the first Piola-Kirchhoff stress tensor, ρ0 is the myocardial tissue density under the reference configuration, and ∂²u(X,t) / ∂t² is the second derivative of the myocardial displacement field with respect to time. The i-th mesh node in the reference heart geometry... Updated to Thus, the dynamic cardiac geometry at time t is obtained. X i Let x be the position vector of the i-th mesh node in the reference configuration. i (t) is the current position vector of the grid node at time t, and i is the grid node number.

[0033] The static magnetic field calculation module is used to calculate the static cardiac geometry without introducing cardiac mechanical motion. Under these conditions, myocardial electrophysiological source terms based on the reference configuration The parameters of the torso conductor and the SERF-MCG sensor array were used to calculate the first magnetic field distribution using a quasi-static electromagnetic field forward calculation method. First, in the torso conductor region... Internal solution for electric potential : (18) The total current density is obtained from the stated potential: (19) In the Coulomb gauge ∇· Solve for the magnetic vector potential and obtain the magnetic flux density when the magnetic flux density is 0: (20) (twenty one) Among them, A S Let μ0 be the magnetic vector potential under static cardiac geometry, and μ0 be the free permeability. This represents the total current density under static cardiac geometry conditions. The first magnetic field output of the k-th sensor channel is: (twenty two) in, For the conductivity tensor of the body conductor, The permeability of free space, and These represent the position and measurement direction of the k-th sensor, respectively. The static magnetic flux density, (t) represents the output of the first magnetic field.

[0034] The dynamic magnetic field calculation module incorporates dynamic cardiac geometry based on cardiac biomechanics. (t)=χ( Under the condition of ,t), based on the myocardial electrophysiological source terms under the reference configuration According to motion mapping The dynamic source item obtained by the transformation The second magnetic field distribution is calculated using the same quasi-static electromagnetic field solution process as the static magnetic field calculation, along with the parameters of the torso conductor, the forward computational domain, and the SERF-MCG sensor array.

[0035] Specifically, in equations (18) to (22) , , , , and Replace with , , , , and The second magnetic field is output by solving the dynamic cardiac geometry corresponding to the target time point: (twenty three) in, (t) represents the second magnetic field output of the k-th sensor channel. The first and second magnetic field distributions use the same reference configuration, electrophysiological activation state, trunk conductor parameters, forward computational domain parameters, sensor position, and sensor measurement direction; the difference lies in whether or not the dynamic cardiac geometry determined by the displacement field and the spatial mapping of the reference source term are introduced.

[0036] The motion artifact extraction module is used to obtain cardiac motion-induced magnetic field artifacts based on the difference between the second magnetic field distribution and the first magnetic field distribution. Since the first and second magnetic field distributions use the same myocardial electrophysiological source terms, the same trunk conductor parameters, the same forward computational domain parameters, and the same SERF-MCG sensor array parameters, the cardiac motion-induced magnetic field artifacts can characterize the SERF-MCG measurement changes caused by cardiac mechanical motion.

[0037] The measurement error feature extraction module is used to extract the measurement error features of cardiac motion-induced magnetic field artifacts. These measurement error features include at least one of the following: magnetic field amplitude, magnetic field polarity, spatial topology, channel response, and target time window response.

[0038] The error assessment output module is used to output the SERF-MCG exercise-induced measurement error assessment result based on the measurement error characteristics. The SERF-MCG exercise-induced measurement error assessment result is used to characterize the degree of influence of cardiac biomechanics on the SERF-MCG magnetic field measurement result.

[0039] In the SERF-MCG cardiac motion-induced measurement error quantification system, the static magnetic field calculation module and the dynamic magnetic field calculation module use the same myocardial electrophysiological source terms, the same trunk conductor parameters, the same forward computational domain parameters, and the same SERF-MCG sensor array parameters, so that the cardiac motion-induced magnetic field artifacts characterize the SERF-MCG measurement errors caused by cardiac mechanical motion.

[0040] Construction of a forward computational model of the electrophysiological-mechanical-electromagnetic coupling of the heart-trunk-sensor array: In this embodiment, the electrophysiological-mechanical-electromagnetic coupling forward model is a specific implementation of the heart-trunk-sensor array forward computation model described above, including a myocardial electrophysiological source modeling module, a cardiac mechanical motion modeling module, a static magnetic field calculation module, and a dynamic magnetic field calculation module. The model uses the heart, trunk conductors, and SERF-MCG sensor array as unified forward computation objects. First, the myocardial electrophysiological source model outputs reference source terms, then the cardiac mechanical motion model obtains the dynamic geometry, and finally, the first magnetic field distribution and the second magnetic field distribution are calculated respectively.

[0041] First, the myocardial electrophysiology source modeling module determines the myocardial calculation region based on a pre-constructed cardiac geometry model, and constructs a local anatomical coordinate system within this region. This local anatomical coordinate system describes the wall thickness direction, long axis direction, and circumferential direction of the myocardial tissue. Based on this local anatomical coordinate system, the myocardial fiber orientation is determined, and myocardial anisotropic conduction parameters are constructed according to these orientations.

[0042] In one implementation, the myocardial anisotropic conduction parameters can be represented as a local conduction tensor, which characterizes the different propagation characteristics of electrophysiological excitation in the myocardial fiber direction and non-fiber direction. The myocardial electrophysiological source modeling module calculates the spatiotemporal distribution of transmembrane potentials within the myocardial computational region based on the local conduction tensor.

[0043] Based on the spatiotemporal distribution of the transmembrane potential, the myocardial electrophysiological source modeling module obtains myocardial electrophysiological source terms for forward calculation of the SERF-MCG magnetic field. In one embodiment, the myocardial electrophysiological source term includes a source current density, which is determined by local conduction parameters and the spatial gradient of the transmembrane potential, and its expression is: (twenty four) in, This represents the source current density corresponding to the myocardial electrophysiological source term. Represents local conduction parameters. Relative to the reference configuration coordinates Spatial gradient operator, This represents the transmembrane potential. In this way, the SERF-MCG magnetic field forward calculation can be driven by myocardial electrophysiological activity, rather than using an empirically defined equivalent magnetic field source.

[0044] Secondly, the cardiac biomechanics modeling module generates active myocardial stress based on the transmembrane potential distribution or the electrophysiological activation state determined by the transmembrane potential distribution. This active myocardial stress characterizes the active contraction of myocardial tissue caused by electrophysiological activation. The cardiac biomechanics modeling module combines the active myocardial stress and myocardial kinetic parameters to calculate the displacement or deformation field of the myocardial tissue at a target time point.

[0045] In one implementation, the current position of the myocardial tissue can be determined by the position and displacement field in a reference configuration, expressed as follows: (25) in, Indicates the position in the reference configuration. Indicates the current position at the target time point. Indicates the time of myocardial tissue The displacement field below, This represents the motion mapping from the reference configuration to the current configuration.

[0046] The cardiac biomechanics modeling module updates the cardiac geometry based on the displacement or deformation field to obtain dynamic cardiac geometry. Further, in one embodiment, the cardiac biomechanics model also includes an anatomical auxiliary reference frame. This anatomical auxiliary reference frame determines the long axis direction of the heart based on a pre-constructed cardiac geometry model. It is established based on the long axis direction and used to spatially divide the cardiac computational region, obtaining cardiac sub-regions that include at least a left ventricular region and a right ventricular region. Corresponding contraction parameters are set for the left ventricular region and the right ventricular region, so that the left ventricular region and the right ventricular region have different active contraction intensities in the cardiac biomechanics model. Based on these different active contraction intensities, asymmetric ventricular contraction motion is calculated to obtain dynamic cardiac geometry that includes the difference in motion between the left and right ventricles. This dynamic cardiac geometry, including the difference in motion between the left and right ventricles, is used for dynamic magnetic field calculation to quantify the impact of asymmetric ventricular contraction motion on the SERF-MCG magnetic field measurement results.

[0047] The specific implementation process is as follows: Cardiac biomechanics models include anatomical auxiliary reference frames to differentiate the contractile strength of the left and right ventricles. Based on the apex... and the center point of the heart Determine the unit vector of the heart's major axis: (26) by With the origin of the coordinate system, As the first coordinate axis, and combined with Two orthogonal unit vectors are used to establish an anatomical auxiliary reference frame. For any point within the myocardial computation region... Its major axis coordinate is: (27) Based on the boundary between the left and right ventricles Spatial division of the cardiac computation region, at least obtaining the left ventricular region. and right ventricular region And set the corresponding peak shrinkage strength coefficient respectively. and Construct the continuous shrinkage coefficient based on the major axis coordinate: (28) in, For smooth step functions, This represents the width of the transition region between the left and right ventricles. Let ζ( Substituting into the active stress evolution equation, the left and right ventricles are made to have different active contraction intensities, and the dynamic cardiac geometry containing the difference in motion between the left and right ventricles is solved through the mechanical equilibrium equation.

[0048] Then, the static magnetic field calculation module and the dynamic magnetic field calculation module perform forward calculations of the SERF-MCG magnetic field, respectively. The static magnetic field calculation module calculates the first magnetic field distribution under static cardiac geometry conditions without incorporating cardiac biomechanics, based on myocardial electrophysiological source terms, trunk conductor parameters, forward computation domain parameters, and SERF-MCG sensor array parameters. The dynamic magnetic field calculation module calculates the second magnetic field distribution under dynamic cardiac geometry conditions incorporating cardiac biomechanics, based on the same myocardial electrophysiological source terms, trunk conductor parameters, forward computation domain parameters, and SERF-MCG sensor array parameters.

[0049] Specifically, the first magnetic field distribution and the second magnetic field distribution are calculated in the following manner: Based on the myocardial electrophysiological source terms, trunk conductor parameters, and sensor array parameters, a forward calculation model of the SERF-MCG magnetic field is established. Under static cardiac geometry conditions without introducing cardiac mechanical motion, the cardiac geometry is kept from being updated by mechanical displacement with the target time point, and the first magnetic field distribution at the SERF-MCG sensor array is calculated by the static magnetic field calculation module. Under the dynamic cardiac geometry conditions that incorporate cardiac biomechanics, the second magnetic field distribution at the SERF-MCG sensor array is calculated by the dynamic magnetic field calculation module based on the dynamic cardiac geometry corresponding to the target time point. The first magnetic field distribution and the second magnetic field distribution correspond to the same myocardial electrophysiological source terms, the same trunk conductor parameters, the same sensor array position parameters, and the same sensor measurement direction parameters; The first magnetic field distribution is used to characterize the SERF-MCG forward calculation results without considering cardiac mechanical motion, and the second magnetic field distribution is used to characterize the SERF-MCG forward calculation results considering cardiac mechanical motion.

[0050] In one implementation, the SERF-MCG forward magnetic field calculation employs a quasi-static electromagnetic field calculation method. Based on the source current density and the return current of the bulk conductor, the total current density is obtained, and the magnetic field distribution is further calculated. The magnetic field distribution can be determined by the magnetic vector potential, and its expression is: (29) in, Indicates the distribution of the magnetic field. It represents the magnetic vector potential.

[0051] In this embodiment, the difference between the first magnetic field distribution and the second magnetic field distribution lies in whether or not dynamic cardiac geometric changes induced by cardiac biomechanics are introduced. Since both use the same myocardial electrophysiological source terms, the same trunk conductor parameters, the same forward computational domain parameters, and the same SERF-MCG sensor array parameters, the magnetic field change subsequently obtained from the difference between the second and first magnetic field distributions can be used to characterize the SERF-MCG measurement error induced by cardiac biomechanics.

[0052] Extraction of motion-induced magnetic field artifacts: In this embodiment, the motion artifact extraction module extracts the SERF-MCG magnetic field measurement changes caused by cardiac biomechanical motion based on the first magnetic field distribution output by the static magnetic field calculation module and the second magnetic field distribution output by the dynamic magnetic field calculation module. Figure 3 As shown, this embodiment obtains cardiac motion-induced magnetic field artifacts by comparing the first magnetic field distribution under static cardiac geometry conditions with the second magnetic field distribution under dynamic cardiac geometry conditions and utilizing the difference between the two.

[0053] First, at least one target time point is selected. The target time point can be any analysis time point in the magnetocardiogram signal, or it can be a pre-specified time point for weak magnetic field characteristics. In one embodiment, the target time point is selected within the ST segment time window to analyze the influence of cardiac motion on low-amplitude magnetic field characteristics.

[0054] For any target time point The static magnetic field calculation module outputs the first magnetic field distribution at the SERF-MCG sensor array, denoted as: (30) in, Indicates the first Spatial location of each SERF-MCG sensor channel Indicates the target time point. This indicates that under static cardiac geometry conditions without introducing cardiac biomechanics, the first... Calculation results of the magnetic field at each sensor channel.

[0055] The dynamic magnetic field calculation module outputs the second magnetic field distribution at the SERF-MCG sensor array, denoted as: (31) in, This indicates that under the dynamic cardiac geometry condition that incorporates cardiac biomechanics, the first... Calculation results of the magnetic field at each sensor channel.

[0056] Subsequently, the motion artifact extraction module performs channel-correspondence difference analysis between the second magnetic field distribution and the first magnetic field distribution at the same target time point and the same sensor channel to obtain cardiac motion-induced magnetic field artifacts, the expression of which is: (32) in, Indicates the first Each sensor channel at the target time point Artifacts of the cardiac motion-induced magnetic field under the influence of the heart.

[0057] In the aforementioned differential process, the first and second magnetic field distributions use the same myocardial electrophysiological source terms, the same trunk conductor parameters, the same forward computational domain parameters, and the same SERF-MCG sensor array parameters. The difference lies in that the first magnetic field distribution does not introduce cardiac biomechanics, while the second magnetic field distribution does. Therefore, the difference between the two can be used to characterize the changes in magnetic field measurements caused by cardiac biomechanics.

[0058] When the SERF-MCG sensor array includes multiple sensor channels, the motion artifact extraction module arranges the difference results at each sensor channel according to the spatial position of the sensor array, forming a spatial distribution of motion-induced magnetic field artifacts. This spatial distribution can be used to characterize the differences in measurement errors caused by cardiac biomechanics at different sensor spatial positions.

[0059] When multiple target time points are selected, the motion artifact extraction module calculates the cardiac motion-induced magnetic field artifacts at each target time point and arranges them in chronological order to form the temporal variation results of the motion-induced magnetic field artifacts. Furthermore, the spatial distributions at multiple time points can be combined to obtain the spatiotemporal distribution of the motion-induced magnetic field artifacts.

[0060] The specific implementation process is as follows: When the SERF-MCG sensor array includes N sensor channels, the motion artifact extraction module at the target time point The difference between the second magnetic field output and the first magnetic field output at the same sensor channel is calculated to obtain the motion-induced magnetic field artifacts for each channel: (33) (34) The results from each channel are sorted according to the sensor location. Arrange them to form spatial artifact vectors, and then... ( Assign values ​​to the sensor array and The corresponding position; when the sensor array is a two-dimensional array, it is based on the sensor plane coordinates ( , The distribution map of motion-induced magnetic field artifacts formed by nodes.

[0061] When Q target time points are selected , ,..., At that time, the motion artifact extraction module calculates the spatial artifact vector corresponding to each target time point. ( They are arranged in chronological order to form a spatiotemporal artifact matrix: (35) in, This is the spatiotemporal artifact matrix within the target time series, where the k-th row represents the artifact time series of the k-th sensor channel, and the q-th column represents the target time point. The spatial artifact distribution below.

[0062] In one embodiment, the SERF-MCG sensor array is a multi-channel two-dimensional array. The motion artifact extraction module calculates separately for each sensor channel. The differential results of all channels are mapped to the corresponding spatial locations of the sensor array to obtain the distribution map of motion-induced magnetic field artifacts at the target time point.

[0063] In this way, this embodiment can separate the SERF-MCG magnetic field measurement changes caused by cardiac mechanical motion without changing the myocardial electrophysiological source terms and sensor array conditions, providing input for subsequent measurement error feature extraction and target time window error assessment.

[0064] Measurement error feature extraction and target time window error assessment: In this embodiment, the measurement error feature extraction module extracts the SERF-MCG motion-induced measurement error features based on the cardiac motion-induced magnetic field artifacts obtained by the motion artifact extraction module, and the error evaluation output module outputs the measurement error evaluation results within the target time window.

[0065] First, the measurement error feature extraction module operates within the target time window W=[ , Within, based on the motion-induced magnetic field artifact of the k-th sensor channel. (t), extract artifact peaks (W) and root mean square of the artifact (W): (36) (37) in, (W) represents the peak artifact value of the k-th sensor channel within the target time window. (W) is the root mean square of the artifact of the k-th sensor channel within the target time window.

[0066] Secondly, the measurement error feature extraction module extracts polarity features based on the sign relationship between the motion-induced magnetic field artifact and the target magnetic field features in the first magnetic field distribution. The polarity feature of the k-th sensor channel within the target time window W is expressed as: (38) in, (W) represents the signed target characteristic amplitude of the first magnetic field distribution within the target time window. (W) represents the signed characteristic amplitude of the motion-induced magnetic field artifact within the same target time window. When When (W)=1, it indicates that the two have the same polarity and produce a superposition effect; when When (W) = -1, it indicates that the two have opposite polarities and cancel each other out; when When (W)=0, no superposition or cancellation determination is made.

[0067] Next, the measurement error feature extraction module extracts spatial topological features based on the spatial distribution of motion-induced magnetic field artifacts on the SERF-MCG sensor array. For the target time window W, the positive and negative extreme values ​​of the artifacts in the k-th sensor channel are first calculated: (39) The set of positive and negative artifact channels is determined based on a preset threshold δ. and negative artifact channel set : (40) Wherein, δ>0 is the preset amplitude threshold used to determine the channels of significant positive and negative motion-induced magnetic field artifacts, and its dimension is the same as that of the motion-induced magnetic field artifact amplitude; δ is determined according to the system noise level or the preset detection sensitivity.

[0068] Based on the adjacency relationship of the sensor array, at least one of the following is calculated for the positive and negative artifact regions: position, area, connectivity, magnetic pole center, and zero flux line position, respectively, to form the spatial topological features within the target time window. .

[0069] Furthermore, the measurement error feature extraction module extracts channel response features based on the artifact peak value and root mean square value of artifacts at different SERF-MCG sensor channels, specifically including the following steps: (1) For the k-th sensor channel within the target time window W, calculate the artifact peak value according to equations (36) and (37). and the root mean square R of the artifact k (W). Among them, P k (W) is used to characterize the maximum artifact amplitude of the sensor channel within the target time window, R k (W) is used to characterize the overall artifact amplitude level of the sensor channel within the target time window; (2) Normalize the artifact peak values ​​of each sensor channel to obtain the relative response level of the k-th sensor channel: (41) Among them, C k (W) represents the relative response of the k-th sensor channel, N is the total number of sensor channels, and ε is a minimal positive number to prevent the normalization denominator from being zero. Its dimensions are the same as the artifact peak value P. k (W) Same; (3) According to C k (W) Sort the sensor channels from largest to smallest; or determine the set of high-response channels and the set of low-response channels based on a preset response threshold θC: (42) Where, θ C K is a preset relative response threshold between 0 and 1. H (W) represents the set of high-response channels, K L (W) represents the set of low-response channels; (4) Compare the C values ​​of different sensor channels k (W), P k (W) and R k (W) yields the sensor channel response ranking, high-response channel set, low-response channel set, and channel response difference evaluation results. For example... Figure 3 As shown in (B), the artifact peak, artifact root mean square, or relative response of the high response channel is greater than that of the low response channel, which is used to characterize the difference in the degree of influence of cardiac mechanical motion-induced measurement error on the SERF-MCG sensor channels at different spatial locations.

[0070] like Figure 3As shown, different SERF-MCG sensor channels exhibit varying degrees of response to cardiac motion-induced magnetic field artifacts. Channel response characteristics can be determined based on the artifact amplitude and waveform changes at each channel. These channel response characteristics characterize the sensitivity of different sensor channels to cardiac biomechanical motion-induced measurement errors. Specifically, a larger amplitude or more pronounced waveform change in a motion-induced magnetic field artifact at a particular sensor channel indicates a higher degree of influence from cardiac biomechanical motion; conversely, a smaller amplitude or weaker waveform change in a motion-induced magnetic field artifact at a particular sensor channel indicates a lower degree of influence from cardiac biomechanical motion.

[0071] After obtaining amplitude characteristics, polarity characteristics, spatial topology characteristics, and channel response characteristics, the measurement error feature extraction module combines the features of each channel according to the sensor channel number, and forms a SERF-MCG motion-induced measurement error feature set together with the spatial topology characteristics within the target time window: (43) in, This represents the set of motion-induced measurement error characteristics within the target time window W. The spatial topological features within the target time window are defined as N, where N is the total number of sensor channels. The error evaluation output module outputs the SERF-MCG motion-induced measurement error evaluation result based on the measurement error feature set. Specifically, (1) the target time window in the SERF-MCG forward calculation result is selected, where the target time window includes the ST segment time window or the low-amplitude magnetic field feature time window; (2) Obtain the first magnetic field distribution, the second magnetic field distribution, and the cardiac motion-induced magnetic field artifacts within the target time window; (3) Based on the amplitude and polarity relationship between the cardiac motion-induced magnetic field artifact and the first magnetic field distribution, determine the superposition or cancellation effect of the cardiac motion-induced magnetic field artifact on the target magnetic field characteristics; (4) When the cardiac motion-induced magnetic field artifact and the target magnetic field feature have opposite polarities, the degree of obscuring of the target magnetic field feature is calculated based on the amplitude of the cardiac motion-induced magnetic field artifact and the amplitude of the target magnetic field feature. (5) Based on the degree of shielding of each sensor channel (W), output the SERF-MCG motion-induced measurement error evaluation results within the target time window; the measurement error evaluation results include at least one of the following: the degree of occlusion of each sensor channel, the set of channels that have a canceling effect, the average occlusion degree of the array, and the maximum occlusion degree.

[0072] In one implementation, the target time window is determined using a preset time boundary or feature point detection method. When the target time window is the ST segment time window, it is determined based on the QRS complex endpoint in the synchronized electrocardiogram or magnetocardiogram signal. Determine the starting point of the ST segment, and =[ + , + Set as the target time window, where, and This is the preset time offset.

[0073] When the target time window is a low-amplitude magnetic field characteristic time window, within the preset candidate time range Inside, the amplitude of the sensor array is calculated based on the first magnetic field distribution: (44) Among them, b S (t)=[b1 S (t), b2 S (t), ..., b N S (t)] T Let b be the multi-channel output vector corresponding to the first magnetic field distribution at time t. k S (t) represents the first magnetic field output of the k-th sensor channel, b N S (t) represents the first magnetic field output of the Nth sensor channel, where N is the total number of sensor channels, and G... S (t) represents the synthesized amplitude of the sensor array at time t, where G is the first magnetic field distribution. max S =max t∈Wc G S (t) represents the candidate time range W c Inner G S The maximum value of (t) will satisfy G. S (t)≤λG max S The continuous time interval is determined as the characteristic time window of the low amplitude magnetic field, where λ is a preset proportional threshold between 0 and 1.

[0074] For the target time window W and the k-th sensor channel, determine the target characteristic time with the largest absolute value in the first magnetic field distribution. And at the same time point, the signed characteristic amplitudes of the first magnetic field distribution, the second magnetic field distribution, and the motion-induced magnetic field artifacts were extracted respectively: (45) (46) (47) Among them, b k D (t) represents the second magnetic field output of the k-th sensor channel under dynamic cardiac geometry conditions, G k D (W) represents the target characteristic time t. k * The signed characteristic amplitude G of the k-th sensor channel extracted from the second magnetic field distribution. k S (W) represents the signed target feature amplitude extracted from the first magnetic field distribution at the same target feature time, and G represents the target feature amplitude. k A (W) represents the signed characteristic amplitude of the motion-induced magnetic field artifact corresponding to the difference between the second magnetic field distribution and the first magnetic field distribution.

[0075] when (W)· When (W) < 0, the motion-induced magnetic field artifact is determined to have opposite polarity to the target magnetic field feature, and the degree of obscuring of the target magnetic field feature is calculated: (48) When 0 < When (W)≤1, it indicates that the target's magnetic field characteristics are partially or completely obscured; when When (W)≤0, it indicates that no shielding has occurred or the target's magnetic field characteristics have been enhanced. The error assessment output module is based on... (W) Outputs the occlusion level of each sensor channel, and further outputs at least one of the array average occlusion level and the maximum occlusion level.

[0076] Through the above method, this embodiment can output the SERF-MCG motion-induced measurement error evaluation result within the target time window, including at least one of the amplitude characteristics, polarity characteristics, spatial topological characteristics, channel response characteristics, and target magnetic field feature masking degree of the motion-induced magnetic field artifact. The measurement error evaluation result can be used to characterize the degree of influence of cardiac biomechanics on the forward measurement results of SERF-MCG. In one specific embodiment, the ST segment time window is selected as the target time window, and channel correspondence difference is performed on the first magnetic field distribution under static cardiac geometry conditions and the second magnetic field distribution under dynamic cardiac geometry conditions to obtain the cardiac motion-induced magnetic field artifact. The evaluation results show that within the ST segment time window, the cardiac motion-induced magnetic field artifact exhibits a structured spatial distribution and shows obvious positive and negative magnetic field region distribution characteristics, indicating that the artifact is not random noise, but a deterministic measurement error caused by changes in the source position and direction due to cardiac biomechanics.

[0077] Furthermore, when comparing different SERF-MCG sensor channels, the dynamic waveform in the low-response channel is close to the static waveform, indicating that this channel is less affected by cardiac motion-induced measurement errors. The dynamic waveform in the high-response channel shows a significant shift relative to the static waveform, indicating that this channel is more affected by cardiac motion-induced measurement errors. In one embodiment, the static magnetic field characteristic amplitude within the target time window is -47.4 pT. The cardiac motion-induced magnetic field artifact generates a positive component with the opposite polarity to the target magnetic field characteristic, with a peak value of 41.2 pT. After superposition, the dynamic magnetic field characteristic amplitude becomes -6.2 pT, corresponding to a target magnetic field characteristic attenuation of approximately 87.0%. Therefore, the measurement error assessment results can quantitatively demonstrate the masking effect of cardiac biomechanics on low-amplitude SERF-MCG magnetic field characteristics.

[0078] Example 2 This invention also provides a method for quantifying the error of SERF-MCG cardiac exercise induction measurement based on electrophysiological-mechanical-electromagnetic coupling, which utilizes the system described in Example 1 and includes the following steps: Step 1: The model parameter acquisition module acquires the model parameters of the heart, trunk, and SERF-MCG sensor array. The model parameters include at least one of the following: heart geometric parameters, trunk conductor parameters, forward computational domain parameters, and SERF-MCG sensor array parameters. Step 2: The myocardial electrophysiological source modeling module constructs an anisotropic electrophysiological source model of the myocardium based on the cardiac geometric parameters, and obtains the myocardial electrophysiological source terms for SERF-MCG magnetic field forward calculation based on the anisotropic electrophysiological source model of the myocardium. Step 3: The cardiac biomechanics modeling module constructs an active stress-driven cardiac biomechanics model based on the myocardial electrophysiological source model, and obtains the dynamic cardiac geometry at at least one target time point based on the cardiac biomechanics model. Step 4: Under static cardiac geometry conditions without introducing cardiac biomechanical motion, the static magnetic field calculation module calculates the first magnetic field distribution at the SERF-MCG sensor array based on the myocardial electrophysiological source terms, the trunk conductor parameters, the forward computation domain parameters, and the SERF-MCG sensor array parameters. Step 5: Under the dynamic cardiac geometry conditions that introduce cardiac biomechanics, the dynamic magnetic field calculation module calculates the second magnetic field distribution at the SERF-MCG sensor array based on the myocardial electrophysiological source terms, the trunk conductor parameters, the forward computational domain parameters, and the SERF-MCG sensor array parameters. Step 6: The motion artifact extraction module obtains cardiac motion-induced magnetic field artifacts based on the difference between the second magnetic field distribution and the first magnetic field distribution; Step 7: The measurement error feature extraction module extracts the measurement error features of the cardiac motion-induced magnetic field artifact. The measurement error features include at least one of magnetic field amplitude, magnetic field polarity, spatial topology, channel response, and target time window response. Step 8: The error assessment output module outputs the SERF-MCG motion-induced measurement error assessment result based on the measurement error characteristics.

[0079] In this embodiment, the model parameter acquisition module acquires the model parameters in the following manner: Obtain the parameters of a pre-constructed forward computation model of the heart-trunk-sensor array, wherein the forward computation model parameters include at least one of the following: heart geometric model parameters, trunk geometric model parameters, myocardial electrophysiological parameters, cardiac biomechanical motion 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 generates unified input parameters for myocardial electrophysiological source modeling, cardiac mechanical motion modeling, static magnetic field calculation, and dynamic magnetic field calculation based on the forward calculation model parameters.

[0080] In this embodiment, the myocardial anisotropic electrophysiological source model is constructed in the following manner: (1) Determine the myocardial calculation region based on the pre-constructed cardiac geometric model, and construct a local anatomical coordinate system within the myocardial calculation region; (2) Determine the direction of myocardial fibers according to the local anatomical coordinate system, and construct myocardial anisotropic conduction parameters according to the direction of myocardial fibers; (3) Based on the myocardial anisotropic conduction parameters, establish a myocardial electrophysiological propagation model to obtain the spatiotemporal distribution of transmembrane potentials within the myocardial calculation region; (4) Based on the spatiotemporal distribution of the transmembrane potential and the anisotropic conduction parameters of the myocardium, obtain the myocardial electrophysiological source terms for forward calculation of the SERF-MCG magnetic field; (5) The myocardial electrophysiological source term includes at least the source current density determined by local conduction parameters and transmembrane potential spatial gradient, such that the forward calculation of the SERF-MCG magnetic field is driven by myocardial electrophysiological activity.

[0081] In this embodiment, the cardiac biomechanics model is constructed in the following manner: (1) Obtain the electrophysiological activation state in the myocardial calculation region based on the myocardial electrophysiological source model, wherein the electrophysiological activation state includes the transmembrane potential distribution or the electrophysiological activation state determined by the transmembrane potential distribution; (2) Generate active myocardial stress based on the electrophysiological activation state, wherein the active myocardial stress is used to characterize the active contraction of myocardial tissue caused by electrophysiological activation; (3) Establish a cardiac mechanical motion model based on the active stress and dynamic parameters of the myocardium, and calculate the displacement field or deformation field of the myocardial tissue at the target time point; (4) Update the cardiac geometry based on the displacement field or deformation field to obtain the dynamic cardiac geometry after introducing cardiac mechanical motion; (5) The dynamic heart geometry is used by the dynamic magnetic field calculation module to calculate the second magnetic field distribution at the SERF-MCG sensor array.

[0082] In this embodiment, the cardiac biomechanics model further includes an anatomical auxiliary reference frame, which is constructed and used to characterize the asymmetric contractile motion of the ventricles in the following manner: (1) Determine the direction of the long axis of the heart based on the pre-constructed cardiac geometric model, and establish an anatomical auxiliary reference system with the direction of the long axis of the heart as the reference direction; (2) Based on the anatomical auxiliary reference system, the cardiac calculation region is spatially divided to obtain a cardiac sub-region that includes at least the left ventricular region and the right ventricular region; (3) Set corresponding contraction parameters for the left ventricular region and the right ventricular region respectively, so that the left ventricular region and the right ventricular region have different active contraction intensities in the cardiac biomechanics model; (4) Calculate the asymmetric ventricular contraction motion based on the different active contraction intensities to obtain dynamic cardiac geometry that includes the difference in motion between the left and right ventricles; (5) The dynamic cardiac geometry, which includes the difference in motion between the left and right ventricles, is used for dynamic magnetic field calculation to quantify the effect of asymmetric ventricular contraction on the SERF-MCG magnetic field measurement results.

[0083] In this embodiment, the first magnetic field distribution and the second magnetic field distribution are calculated in the following manner: (1) Based on the myocardial electrophysiological source terms, trunk conductor parameters and sensor array parameters, a forward calculation model of the SERF-MCG magnetic field is established; (2) Under the static cardiac geometry condition without introducing cardiac mechanical motion, the cardiac geometry is kept from being updated with mechanical displacement at the target time point, and the first magnetic field distribution at the SERF-MCG sensor array is calculated by the static magnetic field calculation module. (3) Under the dynamic cardiac geometry conditions of introducing cardiac biomechanics, the second magnetic field distribution at the SERF-MCG sensor array is calculated by the dynamic magnetic field calculation module according to the dynamic cardiac geometry corresponding to the target time point. (4) The first magnetic field distribution and the second magnetic field distribution correspond to the same myocardial electrophysiological source terms, the same trunk conductor parameters, the same sensor array position parameters, and the same sensor measurement direction parameters; (5) The first magnetic field distribution is used to characterize the SERF-MCG forward calculation results without considering cardiac mechanical motion, and the second magnetic field distribution is used to characterize the SERF-MCG forward calculation results considering cardiac mechanical motion.

[0084] In this embodiment, the cardiac motion-induced magnetic field artifact is obtained in the following manner: (1) Select at least one target time point and obtain the first magnetic field distribution and the second magnetic field distribution corresponding to the target time point respectively; (2) Perform channel correspondence difference between the second magnetic field distribution and the first magnetic field distribution to obtain motion-induced magnetic field artifacts at each SERF-MCG sensor channel; (3) When the first magnetic field distribution and the second magnetic field distribution are multi-channel spatial magnetic field distributions, the differential results at each sensor channel are arranged according to the spatial position of the sensor array to form a spatial distribution of motion-induced magnetic field artifacts; (4) When the first magnetic field distribution and the second magnetic field distribution are time-series magnetic field distributions at multiple target time points, the difference results at each target time point are time-series arranged to obtain the spatiotemporal distribution of motion-induced magnetic field artifacts; (5) The motion-induced magnetic field artifact is used to characterize the changes in magnetic field measurement caused by cardiac mechanical motion under the same myocardial electrophysiological source terms and the same SERF-MCG sensor array.

[0085] In this embodiment, the measurement error feature extraction module extracts measurement error features in the following manner: (1) Based on the difference results of the motion-induced magnetic field artifacts at each SERF-MCG sensor channel, extract the amplitude characteristics of the motion-induced magnetic field artifacts. The amplitude characteristics include at least one of the artifact amplitude, artifact peak value, or artifact amplitude variation at the target time point or within the target time window. (2) Based on the directional relationship between the motion-induced magnetic field artifact and the first magnetic field distribution or the target magnetic field feature, the polarity feature of the motion-induced magnetic field artifact is extracted. The polarity feature is used to characterize the superposition or cancellation relationship between the motion-induced magnetic field artifact and the target magnetic field feature. (3) Based on the spatial distribution of the motion-induced magnetic field artifacts on the SERF-MCG sensor array, extract the spatial topological features of the motion-induced magnetic field artifacts, wherein the spatial topological features include at least one of the distribution of positive and negative magnetic field regions, the distribution morphology of dipole patterns, and the topological changes of magnetic fields; (4) Based on the amplitude and waveform changes of motion-induced magnetic field artifacts at different SERF-MCG sensor channels, channel response features are extracted. These channel response features are used to characterize the sensitivity of different sensor channels to cardiac mechanical motion-induced measurement errors. (5) Based on at least one of the amplitude characteristics, polarity characteristics, spatial topology characteristics and channel response characteristics, a set of SERF-MCG motion-induced measurement error characteristics is formed.

[0086] In this embodiment, the error evaluation output module outputs the measurement error evaluation results within the target time window in the following manner: (1) Select the target time window from the SERF-MCG forward calculation results. The target time window includes the ST segment time window or the low amplitude magnetic field characteristic time window. (2) Obtain the first magnetic field distribution, the second magnetic field distribution, and the cardiac motion-induced magnetic field artifacts within the target time window; (3) Based on the amplitude and polarity relationship between the cardiac motion-induced magnetic field artifact and the first magnetic field distribution, determine the superposition or cancellation effect of the cardiac motion-induced magnetic field artifact on the target magnetic field characteristics; (4) When the cardiac motion-induced magnetic field artifact and the target magnetic field feature have opposite polarities, the degree of obscuring of the target magnetic field feature is calculated based on the amplitude of the cardiac motion-induced magnetic field artifact and the amplitude of the target magnetic field feature. (5) Based on the degree of obscuring of the target magnetic field characteristics, output the SERF-MCG motion-induced measurement error evaluation result within the target time window.

[0087] In this embodiment, the channel response characteristics are determined in the following manner: (1) Based on the motion-induced magnetic field artifacts at each SERF-MCG sensor channel, obtain the artifact amplitude and artifact waveform changes corresponding to different sensor channels; (2) Determine the degree of response of each SERF-MCG sensor channel to cardiac biomechanics motion-induced measurement error based on the artifact amplitude and the artifact waveform change; (3) Based on the response level of each SERF-MCG sensor channel, the motion-induced measurement error of different sensor channels is compared to obtain the channel response difference evaluation results; (4) The channel response difference assessment results are used to characterize the degree of difference in the influence of cardiac mechanical motion on sensor channels at different spatial locations in the SERF-MCG sensor array.

[0088] 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 cardiac motion induction measurement error quantification system based on electrophysiological-mechanical-electromagnetic coupling, characterized in that, The system includes: a model parameter acquisition module, a myocardial electrophysiological source modeling module, a cardiac biomechanics motion modeling module, a static magnetic field calculation module, a dynamic magnetic field calculation module, a motion artifact extraction module, a measurement error feature extraction module, and an error evaluation output module; The model parameter acquisition module is used to acquire the forward computation model parameters of the heart-trunk-sensor array. The model parameters include at least one of the following: heart geometric parameters, trunk conductor parameters, forward computation domain parameters, and SERF-MCG sensor array parameters. The myocardial electrophysiological source modeling module is used to construct an anisotropic electrophysiological source model of the myocardium based on cardiac geometric parameters, and to obtain myocardial electrophysiological source terms for SERF-MCG magnetic field forward calculation based on the anisotropic electrophysiological source model of the myocardium. The cardiac biomechanics modeling module is used to construct an active stress-driven cardiac biomechanics model based on the myocardial electrophysiological source model, and to obtain dynamic cardiac geometry at at least one target time point based on the cardiac biomechanics model. The static magnetic field calculation module is used to calculate the first magnetic field distribution at the SERF-MCG sensor array under static cardiac geometry conditions without introducing cardiac mechanical motion, based on myocardial electrophysiological source terms, trunk conductor parameters, forward computational domain parameters, and SERF-MCG sensor array parameters. The dynamic magnetic field calculation module is used to calculate the second magnetic field distribution at the SERF-MCG sensor array under dynamic cardiac geometry conditions that introduce cardiac biomechanical motion, based on myocardial electrophysiological source terms, trunk conductor parameters, forward computational domain parameters, and SERF-MCG sensor array parameters. The motion artifact extraction module is used to obtain cardiac motion-induced magnetic field artifacts based on the difference between the second magnetic field distribution and the first magnetic field distribution. The measurement error feature extraction module is used to extract the measurement error features of the cardiac motion-induced magnetic field artifact; The error assessment output module is used to output the SERF-MCG motion-induced measurement error assessment result based on the measurement error characteristics.

2. A method for quantifying the error of SERF-MCG cardiac motion induction measurement based on electrophysiological-mechanical-electromagnetic coupling, wherein the method is implemented using the system described in claim 1, characterized in that... The method includes: Step 1: Obtain model parameters for the heart, trunk, and SERF-MCG sensor array. The model parameters include at least one of the following: heart geometric parameters, trunk body conductor parameters, forward computational domain parameters, and SERF-MCG sensor array parameters. Step 2: Construct a myocardial anisotropic electrophysiological source model based on the cardiac geometric parameters, and obtain the myocardial electrophysiological source terms for SERF-MCG magnetic field forward calculation based on the myocardial anisotropic electrophysiological source model; Step 3: Construct an active stress-driven cardiac biomechanical motion model based on the myocardial electrophysiological source model, and obtain the dynamic cardiac geometry at at least one target time point based on the cardiac biomechanical motion model. Step 4: Under static cardiac geometry conditions without introducing cardiac biomechanical motion, calculate the first magnetic field distribution at the SERF-MCG sensor array based on the myocardial electrophysiological source terms, the trunk conductor parameters, the forward computational domain parameters, and the SERF-MCG sensor array parameters. Step 5: Under the dynamic cardiac geometry conditions that introduce cardiac biomechanics, the second magnetic field distribution at the SERF-MCG sensor array is calculated based on the myocardial electrophysiological source terms, the trunk conductor parameters, the forward computation domain parameters, and the SERF-MCG sensor array parameters. Step 6: Obtain the cardiac motion-induced magnetic field artifact based on the difference between the second magnetic field distribution and the first magnetic field distribution; Step 7: Extract the measurement error characteristics of the cardiac motion-induced magnetic field artifact; Step 8: Output the SERF-MCG motion-induced measurement error evaluation result based on the measurement error characteristics.

3. The method according to claim 2, characterized in that, Methods for constructing anisotropic electrophysiological source models of the myocardium based on the aforementioned cardiac geometric parameters include: The myocardial computation region is determined based on a pre-constructed cardiac geometry model, and a local anatomical coordinate system is constructed within the myocardial computation region; The direction of myocardial fibers is determined based on the local anatomical coordinate system, and myocardial anisotropic conduction parameters are constructed based on the direction of myocardial fibers. Based on the anisotropic conduction parameters of the myocardium, an electrophysiological propagation model of the myocardium was established to obtain the spatiotemporal distribution of transmembrane potentials within the computational region of the myocardium. Based on the spatiotemporal distribution of the transmembrane potential and the anisotropic conduction parameters of the myocardium, myocardial electrophysiological source terms for forward calculation of the SERF-MCG magnetic field are obtained.

4. The method according to claim 2, characterized in that, Methods for constructing an actively stress-driven cardiac biomechanical model based on myocardial electrophysiological source models include: The electrophysiological activation state of the myocardium within the computational region was obtained based on the myocardial electrophysiological source model; Active myocardial stress is generated based on the electrophysiological activation state; Based on the active stress and dynamic parameters of myocardium, a cardiac mechanical motion model is established, and the displacement field or deformation field of myocardial tissue at the target time point is calculated. The cardiac geometry is updated based on the displacement field or deformation field to obtain the dynamic cardiac geometry after incorporating cardiac biomechanical motion.

5. The method according to claim 2, characterized in that, Methods for obtaining dynamic cardiac geometry at at least one target time point based on the cardiac biomechanics model include: The direction of the long axis of the heart is determined based on a pre-constructed geometric model of the heart, and an anatomical auxiliary reference system is established using the direction of the long axis of the heart as a reference direction. Based on the aforementioned anatomical auxiliary reference system, the cardiac calculation region is spatially divided to obtain a cardiac sub-region that includes at least the left ventricular region and the right ventricular region. Corresponding contraction parameters are set for the left ventricular region and the right ventricular region respectively, so that the left ventricular region and the right ventricular region have different active contraction intensities in the cardiac biomechanics model; The asymmetric ventricular contraction motion is calculated based on the different active contraction intensities to obtain dynamic cardiac geometry that includes the differences in motion between the left and right ventricles.

6. The method according to claim 2, characterized in that, A method for obtaining cardiac motion-induced magnetic field artifacts based on the difference between the second magnetic field distribution and the first magnetic field distribution includes: Select at least one target time point and obtain the first magnetic field distribution and the second magnetic field distribution corresponding to the target time point respectively; By performing channel-correspondence difference between the second magnetic field distribution and the first magnetic field distribution, motion-induced magnetic field artifacts at each SERF-MCG sensor channel are obtained. When the first magnetic field distribution and the second magnetic field distribution are multi-channel spatial magnetic field distributions, the differential results at each sensor channel are arranged according to the spatial position of the sensor array to form a spatial distribution of motion-induced magnetic field artifacts. When the first magnetic field distribution and the second magnetic field distribution are temporal magnetic field distributions at multiple target time points, the difference results at each target time point are temporally arranged to obtain the spatiotemporal distribution of motion-induced magnetic field artifacts.

7. The method according to claim 2, characterized in that, The method for extracting the measurement error characteristics of the cardiac motion-induced magnetic field artifacts includes: Based on the difference results of the motion-induced magnetic field artifacts at each SERF-MCG sensor channel, the amplitude characteristics of the motion-induced magnetic field artifacts are extracted. Based on the directional relationship between the motion-induced magnetic field artifact and the first magnetic field distribution or target magnetic field characteristics, the polarity characteristics of the motion-induced magnetic field artifact are extracted. Based on the spatial distribution of the motion-induced magnetic field artifacts on the SERF-MCG sensor array, the spatial topological features of the motion-induced magnetic field artifacts are extracted. Based on the amplitude and waveform changes of motion-induced magnetic field artifacts at different SERF-MCG sensor channels, channel response characteristics are extracted; A set of SERF-MCG motion-induced measurement error features is formed based on at least one of amplitude features, polarity features, spatial topology features, and channel response features.

8. The method according to claim 2, characterized in that, The method for outputting SERF-MCG motion-induced measurement error evaluation results based on the aforementioned measurement error characteristics includes: Select a target time window from the SERF-MCG forward calculation results. The target time window includes the ST segment time window or the low-amplitude magnetic field characteristic time window. Acquire the first magnetic field distribution, the second magnetic field distribution, and the cardiac motion-induced magnetic field artifacts within the target time window; Based on the amplitude and polarity relationship between the cardiac motion-induced magnetic field artifact and the first magnetic field distribution, determine the superposition or cancellation effect of the cardiac motion-induced magnetic field artifact on the target magnetic field characteristics. When the cardiac motion-induced magnetic field artifact has opposite polarity to the target magnetic field feature, the degree of obscuring of the target magnetic field feature is calculated based on the amplitude of the cardiac motion-induced magnetic field artifact and the amplitude of the target magnetic field feature. Based on the degree of obscuring of the target magnetic field characteristics, the SERF-MCG motion-induced measurement error assessment result within the target time window is output.