Cardiac electrophysiology simulation method and device, electronic equipment and storage medium

Abstract

This application relates to the field of cardiac electrophysiology simulation technology, and discloses a cardiac electrophysiology simulation method, device, electronic device, and storage medium. By constructing a myocardial tissue model, simulating electrophysiological performance, and screening target virtual populations for simulation, this application can more comprehensively capture and predict the risk of arrhythmias at the population level, improve prediction accuracy, and reduce deviations from clinical data.

Description

Technical Field

[0001] This application relates to the field of cardiac electrophysiology simulation technology, specifically to a cardiac electrophysiology simulation method, device, electronic device, and storage medium. Background Technology

[0002] In related technologies, drug safety assessment and antiarrhythmic drug development in cardiac electrophysiology research mainly rely on in vitro single-cell electrophysiological experiments or virtual population modeling methods at the single-cell level. These methods cannot adequately reflect the complexity at the tissue scale and behavioral changes under population genetic differences, leading to distortions in the simulation of the overall electrophysiological behavior of cardiac tissue. Summary of the Invention

[0003] This application provides a cardiac electrophysiology simulation method, device, electronic device, and storage medium that can improve the predictive accuracy of drug safety assessment.

[0004] In a first aspect, this application provides a cardiac electrophysiological simulation method, the method comprising: constructing a myocardial tissue model, wherein the myocardial tissue model includes the endocardium, media, and epicardium, and the endocardium, media, and epicardium are each composed of multiple discrete myocardial cell units; forming an initial virtual individual model based on the physiological value range of preset parameters in the discrete myocardial cell units, wherein the preset parameters include ion channel parameters and transmembrane transporter parameters; simulating the electrophysiological performance of each initial virtual individual model, and performing multi-level screening based on the electrophysiological performance to obtain an initial virtual population; randomly sampling and re-screening the initial virtual population based on the corrected QT to obtain a target virtual population; and performing cardiac electrophysiological simulation based on the target virtual population.

[0005] Based on the aforementioned technical methods, by constructing multi-scale myocardial tissue models and simulating virtual populations, alignment with real clinical data can be achieved. This allows for a comprehensive and accurate capture and prediction of population-level arrhythmia risk, providing more reliable data for drug safety assessment.

[0006] In one optional implementation, the electrophysiological performance of each initial virtual individual model is simulated, and multi-level screening is performed based on the electrophysiological performance to obtain an initial virtual population. This includes: under preset conditions, simulating the electrophysiological performance of the human body under resting, stress, and rhythmic change states, and screening the initial virtual individual models that can maintain physiological homeostasis to obtain screened virtual individual models. The preset conditions include sympathetic nerve excitation, shortened cardiac cycle, and prolonged cardiac cycle.

[0007] Based on the above-mentioned technical means, it is possible to more accurately reflect the physiological response and homeostasis maintenance ability of the real human body in a dynamic environment.

[0008] In one optional implementation, screening of initial virtual individual models capable of maintaining physiological homeostasis includes: performing at least one of cell-level screening, tissue-level screening, and pseudo-ECG-level screening on the initial virtual individual models. Cell-level screening includes detecting whether the action potentials of discrete myocardial cell units and the calcium ion concentrations within discrete myocardial cell units meet the corresponding screening conditions. Tissue-level screening includes applying standard stimulation to the myocardial tissue model and detecting whether action potential propagation blockage or repolarization failure occurs. Pseudo-ECG-level screening includes detecting whether the QRS waveform or QT interval in the generated pseudo-ECG signal exceeds the normal physiological range.

[0009] Based on the above technical means, by integrating specific detection standards at different scales, the verification of the model's comprehensiveness and reliability is strengthened, effectively avoiding prediction bias caused by insufficient evaluation at a single scale, so that the obtained initial virtual population can more accurately reflect the electrophysiological characteristics of the real population.

[0010] In one alternative implementation, the calculation of the pseudo-ECG signal includes: , in, x is the location of a discrete cardiomyocyte unit. and The location of the electrode used to measure the ECG signal, where D is the conductivity. Let be the gradient of the potential field.

[0011] Based on the above technical means, when constructing a myocardial tissue model for cardiac electrophysiological simulation, discrete myocardial cell units are constructed based on an ion channel dynamics model, so that each cell unit can accurately simulate its own ion channel dynamics and action potential behavior, thereby providing a high-fidelity electrophysiological basis at the cellular level.

[0012] In one optional implementation, discrete cardiomyocyte units are constructed based on an ion channel kinetics model, and the discrete cardiomyocyte units are electrically coupled through a diffusion coefficient D; the transmembrane current in the cardiomyocyte tissue model includes at least a fast sodium current, an L-type calcium current, a delayed rectified potassium current, an inward rectified potassium current, a sodium-calcium exchange current, and a sodium-potassium pump current; the transmembrane voltage in the cardiomyocyte tissue model is obtained by: ,in, For transmembrane current, C represents the stimulation current that controls the rhythm. m It is the capacitance of the myocardial fiber membrane.

[0013] Based on the above technical means, the accuracy of simulation results and clinical prediction ability have been significantly improved, and the problems of inaccurate and inconsistent parameter generation have been effectively solved.

[0014] In one optional implementation, an initial virtual individual model is formed based on the physiological value range of preset parameters in discrete cardiomyocyte units, including: determining the physiological value range of ion channel parameters and transmembrane transporter parameters through database data, wherein the database data includes publicly available human electrophysiological experimental data and literature data; randomly generating endocardial parameter data based on the physiological value range; generating mid-layer parameter data and epicardial parameter data respectively from the generated endocardial parameter data according to an independent scaling factor; and forming the initial virtual individual model based on the endocardial parameter data, mid-layer parameter data, and epicardial parameter data.

[0015] Based on the above technical means, an initial virtual individual model is formed, providing a basic model with physiological basis, individual specificity and complete structure for subsequent cardiac electrophysiological simulation. This significantly improves the accuracy of simulation results and clinical prediction ability, and effectively solves the problems of inaccurate and inconsistent parameter generation.

[0016] In one optional implementation, based on the corrected QT, the initial virtual population is randomly sampled and re-screened to obtain the target virtual population, including: obtaining a standard corrected QT distribution as the target reference distribution; and selecting a set of virtual individuals consistent with the target reference distribution from the initial virtual population through a sampling strategy to obtain the target virtual population.

[0017] Based on the aforementioned technical means, cardiac electrophysiological simulations based on this target virtual population can more accurately predict tissue-level arrhythmic events that may occur in clinical practice, such as premature ventricular contractions, T-wave alternation, or repolarization failure, thereby significantly improving the accuracy and reliability of drug safety assessment and antiarrhythmic drug development.

[0018] Secondly, this application provides a cardiac electrophysiological simulation device, comprising: a construction module for constructing a myocardial tissue model, wherein the myocardial tissue model includes the endocardium, media, and epicardium, each of which is composed of multiple discrete myocardial cell units; a formation module for forming an initial virtual individual model based on the physiological value range of preset parameters in the discrete myocardial cell units, wherein the preset parameters include ion channel parameters and transmembrane transporter parameters; a screening module for simulating the electrophysiological performance of each initial virtual individual model and performing multi-level screening based on the electrophysiological performance to obtain an initial virtual population; a sampling module for randomly sampling and re-screening the initial virtual population based on the corrected QT to obtain a target virtual population; and a simulation module for performing cardiac electrophysiological simulation based on the target virtual population.

[0019] Thirdly, this application provides an electronic device, including: a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to perform the cardiac electrophysiological simulation method of the first aspect or any corresponding embodiment described above.

[0020] Fourthly, this application provides a computer-readable storage medium storing computer instructions for causing a computer to execute the cardiac electrophysiological simulation method of the first aspect or any corresponding embodiment described above.

[0021] Fifthly, this application provides a computer program product, including computer instructions for causing a computer to execute the cardiac electrophysiological simulation method described in the first aspect or any corresponding embodiment. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the specific embodiments or related technologies of this application, the drawings used in the description of the specific embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 This is a schematic flowchart of a first method for simulating cardiac electrophysiology according to an embodiment of this application; Figure 2 This is a schematic diagram of a second flowchart of a cardiac electrophysiology simulation method according to an embodiment of this application; Figure 3 This is a schematic diagram of the third process of the cardiac electrophysiology simulation method according to the embodiments of this application; Figure 4 This is a schematic diagram of the fourth process of the cardiac electrophysiology simulation method according to the embodiments of this application; Figure 5 This is a structural block diagram of a cardiac electrophysiology simulation device according to an embodiment of this application; Figure 6 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] It is understood that before using the technical solutions disclosed in the various embodiments of this application, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this application in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0026] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0027] In some embodiments, this application proposes a cardiac electrophysiological simulation method, such as Figure 1 As shown, it includes: Step S101: Construct a myocardial tissue model. The myocardial tissue model includes the endocardium, media, and epicardium, each of which is composed of multiple discrete myocardial cell units.

[0028] Specifically, a myocardial tissue model refers to a computational model used to simulate the electrophysiological activity of the heart. Structurally, it simulates the layered characteristics of real heart tissue, such as including the endocardium, media, and epicardium, and is composed of smaller discrete cardiomyocyte units to capture the complexity at the tissue scale.

[0029] Discrete cardiomyocyte units refer to the basic computational units that make up a model of myocardial tissue. Each unit represents one or a group of cardiomyocytes, and their electrophysiological behavior is described by specific mathematical equations. They are the basis for simulating cardiac electrical activity.

[0030] Cardiac tissue models are fundamental for cardiac electrophysiological simulations, structurally mimicking the layered characteristics of real heart tissue. Specifically, a cardiac tissue model can be constructed as a structure composed of multiple discrete cardiac cell units. For example, a simple two-dimensional mesh model can be used, where each mesh point represents a discrete cardiac cell unit, and these units are connected according to predefined rules, such as connections between adjacent units, forming cardiac tissue. The cardiac tissue model can be further divided into three regions: the endocardium, media, and epicardium, each composed of multiple of the aforementioned discrete cardiac cell units. These regional divisions can be based on predefined geometric boundaries or cell type distributions. For example, during model construction, it is possible to manually specify whether cell units in different regions belong to the endocardium, media, or epicardium, and assign corresponding region attributes to each unit.

[0031] Step S102: Based on the physiological value range of preset parameters in the discrete myocardial cell unit, an initial virtual individual model is formed.

[0032] Specifically, preset parameters refer to the variable parameters used to define the electrophysiological characteristics of discrete cardiomyocyte units when constructing a virtual individual model. Their value range reflects biological individual differences. These parameters specifically include ion channel parameters and transmembrane transporter parameters.

[0033] Ion channel parameters refer to parameters that control the opening, closing, and conduction properties of ion channels on the myocardial cell membrane, such as the conductivity or activation / deactivation mechanical parameters of sodium, calcium, and potassium ion channels.

[0034] Transmembrane transporter parameters refer to parameters that control the activity of various transmembrane transport proteins on the cardiomyocyte membrane, such as the transport rate or affinity parameters of the sodium-calcium exchanger and the sodium-potassium pump.

[0035] The initial virtual individual model refers to a set of parameter configurations randomly generated according to the physiological value range of preset parameters. It represents a virtual individual with specific electrophysiological characteristics and is the basis for constructing a virtual population.

[0036] This step aims to simulate individual differences present in real-world populations. Specifically, the physiological ranges of key electrophysiological parameters in discrete cardiomyocyte units can be determined first. For example, upper and lower limits for ion channel parameters, such as sodium channel conductance and calcium channel conductance, and transmembrane transporter parameters, such as sodium-calcium exchanger activity and sodium-potassium pump activity, can be set by consulting publicly available biological literature or expert experience. Subsequently, within these pre-defined physiological ranges, a set of parameter values ​​can be randomly selected for each discrete cardiomyocyte unit to form an initial virtual individual model. For example, a uniformly distributed random sampling method can be used to randomly generate a parameter value within the physiological range of each parameter, and these parameter values ​​can be combined to define a unique virtual individual.

[0037] Step S103: Simulate the electrophysiological performance of each initial virtual individual model, and perform multi-level screening based on the electrophysiological performance to obtain the initial virtual population.

[0038] The initial virtual population refers to the set of virtual individuals obtained by simulating the electrophysiological performance of multiple initial virtual individual models and conducting preliminary screening. These individuals maintain basic physiological homeostasis to a certain extent.

[0039] Specifically, this step aims to select physiologically plausible individuals from a large number of randomly generated virtual individuals. Specifically, each initial virtual individual model can be subjected to electrophysiological simulation to observe its electrical activity under specific conditions. For example, the action potential of a single myocardial cell unit at a fixed stimulation frequency can be simulated, and its action potential duration (APD) or repolarization morphology can be recorded. Subsequently, screening is performed based on these electrophysiological manifestations. For example, a simple screening criterion can be set, such as requiring the simulated action potential duration to fall within a predetermined physiological range, or requiring that cells do not exhibit spontaneous excitation. Initial virtual individual models that meet these basic physiological homeostasis conditions will be retained, thus constituting the initial virtual population.

[0040] Step S104: Based on the corrected QT, randomly sample and re-screen the initial virtual population to obtain the target virtual population.

[0041] Among them, the corrected QT, which refers to the QT interval in an electrocardiogram (ECG) after heart rate correction, is an important clinical indicator for assessing the cardiac repolarization process and is used to reflect the average level of the action potential duration of ventricular myocytes.

[0042] The target virtual population refers to a set of virtual individuals whose corrected QT distribution matches the clinical distribution of the real population, based on the initial virtual population, through random sampling and re-screening processes based on corrected QT. This set is used for subsequent more accurate cardiac electrophysiological simulations.

[0043] Specifically, this step aims to make the electrophysiological characteristics of the virtual population more closely resemble clinical data from real-world populations. Specifically, each virtual individual model in the initial virtual population can be simulated to calculate its corresponding corrected QT value. For example, the electrical activity of a myocardial tissue model can be simulated, and pseudo-ECG signals can be extracted from the simulation results to calculate the corrected QT interval. Subsequently, these corrected QT values ​​can be analyzed and compared with a pre-defined reference QT distribution. For example, virtual individuals whose corrected QT values ​​fall within a specific normal range can be selected, or a subset of individuals can be selected from the initial virtual population using a random sampling method, such that the corrected QT distribution of these selected individuals is statistically similar to the target reference distribution, such as the QT distribution of a known clinical population. In this way, a more representative target virtual population can be obtained.

[0044] Step S105: Perform cardiac electrophysiological simulation based on the target virtual population.

[0045] Among them, cardiac electrophysiological simulation refers to the use of computational models to simulate the electrical activity of the heart under different physiological or pathological conditions, including the generation and propagation of action potentials and the formation of electrocardiograms, in order to predict cardiac function or drug effects.

[0046] Specifically, the acquired target virtual population can be used to conduct cardiac electrophysiological simulation experiments under various simulated conditions. For example, the effects of drugs on the QT interval of different individuals within the target virtual population can be simulated, or the risk of conditional arrhythmias under specific pathological conditions, such as ischemia or myocardial infarction, can be simulated. By simulating the target virtual population, more clinically meaningful and individually differentiated predictive results can be obtained, thereby supporting applications such as drug development and risk assessment.

[0047] It is understood that the embodiments of this application, by constructing a multi-scale myocardial tissue model and simulating a virtual population, can be aligned with real clinical data. Therefore, it is possible to comprehensively and accurately capture and predict the risk of arrhythmias at the population level, providing a more reliable tool for drug safety assessment.

[0048] In some of the embodiments described above in this application, a method is proposed to simulate electrophysiological performance and perform multi-level screening to obtain an initial virtual population. However, in this process, the electrophysiological performance under different physiological states is not considered, which may result in the selected model failing to accurately reflect the ability of the real human body to maintain physiological homeostasis in a dynamic environment, thereby affecting the reliability of subsequent risk assessment.

[0049] In response, this application further proposes a method for simulating cardiac electrophysiology, such as... Figure 2 As shown, the method includes: Step S201: Construct a myocardial tissue model. This model includes the endocardium, media, and epicardium, each composed of multiple discrete myocardial cell units. (See details...) Figure 1 Step S101 in the embodiment will not be described again here.

[0050] Step S202: Based on the physiological value range of preset parameters in the discrete cardiomyocyte units, an initial virtual individual model is formed. (See details...) Figure 1 Step S102 in the embodiment will not be described again here.

[0051] Step S203 involves simulating the electrophysiological performance of each initial virtual individual model and performing multi-level screening based on these performances to obtain the initial virtual population. Specifically, this includes: Step S2031: Under preset conditions, simulate the electrophysiological performance of the human body under resting, stress and rhythm change states respectively, and screen the initial virtual individual models that can maintain physiological homeostasis to obtain screened virtual individual models. The preset conditions include sympathetic nerve excitation, shortened cardiac cycle and prolonged cardiac cycle.

[0052] Specifically, in simulating the electrophysiological performance of the human body under resting, stress, and rhythmic changes, this step aims to comprehensively evaluate the electrophysiological response of the initial virtual individual model under different physiological conditions. On the one hand, ion channel conductance and transmembrane transporter activity can be set to healthy physiological values ​​to simulate cardiac electrical activity at rest; stress can be simulated by adjusting parameters of ion channels such as L-type calcium channels and potassium channels to reflect cardiomyocyte responses to adrenergic stimulation; and the adaptability of the model at different heart rates can be evaluated by changing the cycle length of external stimuli to simulate cardiac cycle shortening or lengthening. On the other hand, these different physiological states can also be simulated by adjusting the frequency and intensity of external stimuli in the simulated environment, as well as simulating changes in intracellular and extracellular ion concentrations. For example, cardiac cycle shortening or lengthening can be simulated by changing the cycle length of external stimuli, and sympathetic nerve excitation can be simulated by introducing simulated neurotransmitters, such as changes in norepinephrine concentration, thereby observing the electrophysiological performance of cardiomyocytes, including action potential morphology, conduction velocity, and repolarization characteristics.

[0053] When screening initial virtual individual models capable of maintaining physiological homeostasis, this step ensures that the selected virtual individual models can maintain normal electrophysiological function under dynamic physiological conditions. On one hand, the screening process can set a series of physiological homeostasis indicators, such as requiring action potential duration (APD) to fluctuate within the physiological range, the absence of early afterdepolarization (EAD) or delayed afterdepolarization (DAD), and stable repolarization at different stimulation frequencies. On the other hand, a strict set of screening criteria can be defined. For example, during simulation, if the model's resting membrane potential drift exceeds a preset threshold, the action potential amplitude decreases significantly, or non-physiological phenomena such as persistent reentry excitation occur, the model will be excluded. Furthermore, the ability to maintain physiological homeostasis can be determined by evaluating the model's repolarization reserve capacity under different cardiac cycles, i.e., its ability to maintain normal repolarization during heart rate changes.

[0054] The preset conditions include sympathetic nerve excitation, shortened cardiac cycle, and prolonged cardiac cycle. These preset conditions are used to simulate specific scenarios under different physiological states to more comprehensively evaluate the robustness of the model. On the one hand, sympathetic nerve excitation can be simulated by increasing the conductance of L-type calcium channels and regulating potassium channel activity to reflect the heart's response under stress or exercise. Shortening the cardiac cycle, such as from 1000 ms to 500 ms, is used to simulate tachycardia and evaluate the model's response to high-frequency stimulation. Prolonging the cardiac cycle, such as from 1000 ms to 1500 ms, is used to simulate bradycardia and evaluate the model's repolarization characteristics and stability under low-frequency stimulation. On the other hand, sympathetic nerve excitation can be simulated by mimicking the signaling pathway activated by adrenaline receptors, leading to an increase in cyclic adenosine monophosphate (cAMP) levels, which in turn affects the phosphorylation state of various ion channels. Shortening and prolonging the cardiac cycle can be achieved by directly adjusting the frequency of external electrical stimulation. For example, by setting different stimulation intervals, the action potential remodeling and conduction characteristics of cardiomyocytes at different heart rates can be observed.

[0055] Step S204: Based on the calibrated QT, randomly sample and re-screen the initial virtual population to obtain the target virtual population. See details... Figure 1 Step S104 in the embodiment will not be described again here.

[0056] Step S205: Perform cardiac electrophysiological simulation based on the target virtual population. See details... Figure 1 Step S105 in the embodiment will not be described again here.

[0057] It is understood that, through the above-described technical solution in this embodiment, this application overcomes the limitations of traditional methods in obtaining the initial virtual population by failing to fully consider the electrophysiological performance under different physiological states. This ensures that the obtained initial virtual population not only performs normally under a single steady state but also maintains its electrophysiological homeostasis when facing multiple physiological challenges such as sympathetic nerve excitation, shortened cardiac cycle, and prolonged cardiac cycle. This makes the selected virtual individual models more biologically realistic and clinically relevant, and more accurately reflects the physiological response and homeostasis maintenance ability of the real human body in a dynamic environment. Subsequent cardiac electrophysiological simulations based on this initial virtual population will be able to more reliably predict the risk of arrhythmias in individuals and populations, thereby significantly improving the accuracy and reliability of drug safety assessment and antiarrhythmic drug development. In some of the embodiments described above in this application, it is proposed to screen initial virtual individual models that can maintain physiological homeostasis to obtain an initial virtual population. However, in this process, the screening method may lack a specific multi-scale evaluation mechanism and cannot simultaneously cover electrophysiological defects at the cellular, tissue, and clinical electrocardiogram levels. This results in the model being unable to accurately identify tissue-level arrhythmic events such as repolarization failure, action potential propagation blockage, or QT interval abnormalities, thereby causing a deviation between the predicted results and the actual clinical risk.

[0058] In response, this application further proposes a screening method for initial virtual individual models capable of maintaining physiological homeostasis, including: screening the initial virtual individual model at least at the cellular level, tissue level, and pseudo-ECG level. Cellular level screening includes detecting whether the action potentials and calcium ion concentrations within discrete myocardial cell units meet the corresponding screening criteria. Tissue level screening includes applying standard stimulation to the myocardial tissue model and detecting whether action potential propagation blockage or repolarization failure occurs. Pseudo-ECG level screening includes detecting whether the QRS waveform or QT interval in the generated pseudo-ECG signal exceeds the normal physiological range.

[0059] Specifically, a multi-level screening mechanism can be employed when screening initial virtual individual models. This mechanism involves at least one of three screening levels: cellular, tissue, and pseudo-ECG. This screening method allows for selective execution of one or more levels based on actual needs and research objectives. For example, it can involve only cellular screening, simultaneous cellular and tissue screening, or a combination of all three. For instance, a sequential screening approach can be used: first, cellular screening is performed; models that pass this stage proceed to tissue screening; and finally, models that pass tissue screening undergo pseudo-ECG screening. Alternatively, a parallel screening approach can be used, where models are screened at different levels simultaneously, and a comprehensive judgment is made based on the results of each level. This ensures that the model meets physiological homeostasis requirements at different scales.

[0060] Cell-level screening is a fundamental level of validation, its core being the detection of action potentials and intracellular calcium ion concentrations within discrete cardiomyocyte units to ensure they meet relevant screening criteria. Action potentials directly reflect the excitability of cardiomyocytes, and their morphology, duration, peak potential, resting potential, and repolarization rate must all be within physiological ranges. For example, the normal range for action potential duration can be set as 200-350 milliseconds, and the resting membrane potential should be between -80mV and -90mV. Calcium ion concentration reflects the excitation-contraction coupling function of cardiomyocytes, and its peak concentration, time to peak concentration, and clearance rate must also meet physiological standards. For instance, the peak value of intracellular calcium transients should be between 0.5-1.5 μM, and the decay time should be within a reasonable range. By detecting these key parameters, virtual individual models exhibiting functional abnormalities or homeostatic imbalances at the ion channel level can be effectively identified.

[0061] Tissue-level screening focuses on evaluating the electrophysiological behavior of myocardial tissue models at the macroscopic level. This screening involves applying standard stimulation to the myocardial tissue model to detect action potential propagation blockade or repolarization failure. The standard stimulation can be a single suprathreshold stimulus to assess the conduction velocity and integrity of the action potential; or it can be an S1-S2 stimulation protocol, where a second stimulus (S2) is applied at different time intervals after a first stimulus (S1) to assess the tissue's effective refractory period and repolarization reserve. Action potential propagation blockade manifests as the inability of the stimulus to conduct effectively in the tissue, or a conduction velocity significantly lower than physiological values; repolarization failure may manifest as a persistently prolonged action potential in a localized area, preventing normal repolarization, or persistent reentrant excitation under short coupling interval stimulation. For example, it can be detected whether the action potential conduction velocity in the myocardial tissue model is within the physiological range of 0.3-0.6 m / s, and whether there is a significant prolongation or shortening of the action potential duration (APD) in a localized area, leading to repolarization heterogeneity.

[0062] The pseudo-ECG screening correlates the model output with clinical ECG characteristics, including detecting whether the QRS waveform or QT interval in the generated pseudo-ECG signal exceeds the normal physiological range. The pseudo-ECG signal is calculated and simulated by the potential field in a myocardial tissue model, reflecting the overall electrical activity of the heart. The QRS waveform represents the ventricular depolarization process, and its width, amplitude, and morphology should conform to the normal physiological range; for example, the QRS width should be between 80-120 milliseconds. The QT interval reflects the total duration of ventricular depolarization and repolarization, and its length is an important indicator of arrhythmia risk; for example, the corrected QT interval (QTc) should be between 350-450 milliseconds. By detecting these pseudo-ECG characteristics, it can be ensured that the virtual individual model maintains consistency with real human physiological data when simulating clinical ECG manifestations.

[0063] It is understood that, through the above-described technical solution in this embodiment, this application introduces a hierarchical screening mechanism, which addresses the shortcomings of single-scale assessment and ensures that the model fully conforms to physiological standards at the cellular, tissue, and clinical levels, thereby improving the accuracy of arrhythmia prediction. Specifically, cellular-level screening is performed on the initial virtual individual model by detecting whether the action potential and calcium ion concentration of discrete myocardial cell units meet the conditions. Based on these basic electrophysiological parameters, cell homeostasis is verified to prevent potential risks caused by ion channel abnormalities. Tissue-level screening assesses tissue-scale events based on the dynamic response of the myocardial tissue model by applying standard stimulation and detecting action potential propagation blockage or repolarization failure, directly capturing excitation-conduction coupling defects. Pseudo-ECG-level screening detects whether the QRS waveform or QT interval in the pseudo-ECG signal exceeds the normal range. Based on the generated pseudo-ECG signal and its association with clinical ECG characteristics, the model output is correlated with real arrhythmia indicators. By integrating specific detection standards at different scales, the verification of the model's comprehensiveness and reliability is strengthened, effectively avoiding prediction bias caused by insufficient evaluation at a single scale. This allows the initial virtual population to more accurately reflect the electrophysiological characteristics of the real population, providing a more reliable foundation for subsequent cardiac electrophysiological simulation. In some of the schemes described above in this application, a pseudo-ECG screening method is proposed to detect whether the QRS waveform or QT interval in the generated pseudo-ECG signal exceeds the normal physiological range. However, in this process, if the calculation method of the pseudo-ECG signal is not specific or based on physical principles, it may lead to inaccurate calculation and fail to reliably reflect the clinical electrocardiographic characteristics, thereby affecting the effectiveness of the screening and its alignment with real data.

[0064] In this regard, this application further proposes that the calculation of the pseudo-ECG signal includes: , in, x is the location of a discrete cardiomyocyte unit. and The location of the electrode used to measure the ECG signal, where D is the conductivity. Let be the gradient of the potential field.

[0065] Specifically, the ECG calculation formula is based on the body conduction model in electrophysiology. It simulates ECG signals at the body surface or specific locations by spatially integrating the electrical activity within the myocardial tissue. This formula establishes a physical connection between microscopic cellular-level electrical activity and macroscopic ECG signals, ensuring the physiological accuracy of the calculation results. In practical applications, this integration is typically achieved through numerical methods, such as discretizing the myocardial tissue model into a finite element or finite difference mesh, and then summing the contribution of each mesh element.

[0066] Here, x represents the position of the discrete cardiomyocyte unit, indicating the specific coordinates of each discrete cardiomyocyte unit in the three-dimensional space within the cardiomyocyte tissue model. By clearly defining the position of each cardiomyocyte unit, its contribution to the overall ECG signal can be accurately tracked and calculated, thus ensuring spatial resolution and model realism. For example, a Cartesian coordinate system (x, y, z) can be used to define the center position of each unit, or a curvilinear coordinate system can be used in more complex geometric models.

[0067] and The position of the electrodes used to measure ECG signals refers to the fixed coordinates in three-dimensional space of the electrodes used to "measure" ECG signals in a simulated environment. The electrode positions directly simulate the placement of ECG leads in clinical settings, ensuring that the calculated pseudo-ECG signals can be directly compared and verified with actual clinical ECG data. For example, custom electrode array positions can be set according to specific research needs.

[0068] D represents electrical conductivity, indicating the electrical conduction capacity of myocardial tissue and its surrounding medium. The magnitude and directionality of conductivity directly affect the propagation speed and attenuation of electrical signals in the tissue, thus determining the amplitude and morphology of the ECG signal. For example, in simulations, different conductivity tensors can be set according to the orientation of myocardial fibers to more realistically reflect the propagation of electrical signals in the myocardium.

[0069] V represents the gradient of the electrical potential field, indicating the rate of change of the electrical potential field within myocardial tissue in space. Its magnitude reflects the drasticness of the potential change. For example, the gradient of the electrical potential field can be approximated by spatially differencing the transmembrane voltage of discrete myocardial cell units.

[0070] (1 / r) represents the gradient of the distance factor 1 / r, where r is the distance from the cardiomyocyte unit x to the measuring electrode location. This term reflects the physical property that the electrical potential decays with distance as the electrical signal propagates from the source point, i.e., the cardiomyocyte unit, to the observation electrode. It ensures that cell units closer to the electrode contribute more to the ECG signal, while those farther away contribute less. For example, for each cardiomyocyte unit, its Euclidean distance r to each electrode can be calculated, and then the gradient of 1 / r can be calculated.

[0071] It is understood that, through the above-described technical solutions in this embodiment, the physics-based pseudo-ECG signal calculation method provided in this application ensures the high scientific rigor and accuracy of the pseudo-ECG signal generation process. This method precisely simulates the conversion process from internal electrical activity to surface potential within myocardial tissue, enabling the generated pseudo-ECG signal to accurately reflect the electrophysiological state of the myocardium. This improves the reliability of pseudo-ECG-level screening. Pseudo-ECG calculation enhances the comprehensiveness and accuracy of screening, resulting in a higher consistency between the final obtained initial virtual population and target virtual population and real human data in terms of electrophysiological characteristics.

[0072] In some of the embodiments described above in this application, a myocardial tissue model is proposed for cardiac electrophysiological simulation. However, in its implementation, due to the lack of detailed myocardial cell unit construction mechanism, electrical coupling mode and comprehensive coverage of transmembrane current, the model cannot accurately simulate tissue-scale repolarization gradient, excitation-conduction coupling and individual variability, thereby affecting the accuracy of simulation results and clinical predictive ability, and causing deviation between model prediction and real arrhythmia events.

[0073] In response, this application proposes a cardiac electrophysiological simulation method, wherein discrete cardiomyocyte units are constructed based on an ion channel dynamics model, and the discrete cardiomyocyte units are electrically coupled through a diffusion coefficient D; the transmembrane current in the cardiomyocyte tissue model includes at least a fast sodium current, an L-type calcium current, a delayed rectified potassium current, an inward rectified potassium current, a sodium-calcium exchange current, and a sodium-potassium pump current; the transmembrane voltage in the cardiomyocyte tissue model is obtained by:

[0074] in, For transmembrane current, C represents the stimulation current that controls the rhythm. m It is the capacitance of the myocardial fiber membrane.

[0075] Specifically, discrete cardiomyocyte units are constructed based on ion channel dynamics models, which are mathematical models used to describe the opening, closing, and transmembrane flow of various ion channels on the cardiomyocyte membrane. These models typically consist of a series of differential equations that simulate the generation, propagation, and repolarization of action potentials. In practical applications, Hodgkin-Huxley models, such as the Luo-Rudy model, the Ten Tusscher model, or the O'Hara-Rudy model, can be used. These models detail the gating dynamics and current characteristics of various ion channels, such as sodium, calcium, and potassium channels. Alternatively, simplified ion channel models, such as variants based on the FitzHugh-Nagumo or Beeler-Reuter models, can be employed, striking a balance between computational efficiency and physiological detail. In this way, each discrete cardiomyocyte unit can be realistically simulated to reflect its own electrophysiological behavior, providing an accurate cellular-level basis for complex interactions at the tissue level.

[0076] Discrete cardiomyocyte units are electrically coupled via a diffusion coefficient D. Electrical coupling refers to the electrical connection formed between adjacent cardiomyocytes through gap junctions, allowing ionic currents to flow between cells, thus enabling rapid propagation of action potentials. The diffusion coefficient D is a parameter measuring the strength and efficiency of this electrical coupling, reflecting the speed and range of electrical signal propagation in the tissue. In the model, the potential difference between adjacent discrete cardiomyocyte units can be combined with the diffusion coefficient D using the finite difference method or the finite element method to simulate the current diffusion process. For example, on a discrete grid, the potential change of each cell is affected by the potentials of its neighboring cells, the degree of which is determined by the diffusion coefficient D. Another approach is to use a resistive network model, treating each cardiomyocyte unit as a node, with adjacent nodes connected by an equivalent resistance associated with the diffusion coefficient D, thereby simulating current conduction between cells. This solves the simulation problem of excitation-conduction coupling in myocardial tissue, enabling action potentials to propagate effectively between discrete cardiomyocyte units, thus forming tissue-scale electrophysiological activity.

[0077] Furthermore, the transmembrane currents in the myocardial tissue model include at least the rapid sodium current, L-type calcium current, delayed rectified potassium current, inward rectified potassium current, sodium-calcium exchange current, and sodium-potassium pump current. Transmembrane currents are ion flows generated by the activity of various ion channels and transporters on the myocardial cell membrane. The dynamic balance of these currents determines the action potential morphology and rhythm of the myocardial cells. These listed currents are the most critical and dominant current types in myocardial electrophysiological activity. In the ion channel kinetic model, independent mathematical equations are established for each current type, such as the rapid sodium current INa, L-type calcium current ICaL, delayed rectified potassium current IKr / IKs, inward rectified potassium current IK1, sodium-calcium exchange current INaCa, and sodium-potassium pump current INaK, and calculations are performed based on their gating characteristics and ion concentration gradients. Alternatively, parameters for each current can be fitted and calibrated based on experimental data and biophysical principles to ensure that their behavior under different physiological conditions is consistent with the response of real myocardial cells. By covering these key current types, the myocardial tissue model can comprehensively and accurately reflect the electrophysiological processes of cardiomyocytes, covering key mechanisms of action potential formation, repolarization, and ion homeostasis maintenance, thereby improving the model's accuracy in predicting arrhythmic events.

[0078] Obtaining transmembrane voltage in a myocardial tissue model includes:

[0079] This equation is a partial differential equation describing the change of the myocardial cell membrane potential V with time t. It comprehensively considers the transmembrane ion current. External stimulation current membrane capacitor And intercellular electrical coupling, determined by the diffusion coefficient D and the second derivative of the potential field. This represents the effect on membrane potential. The partial differential equation can be solved numerically, for example using the finite difference method or the finite element method, iteratively calculating the transmembrane voltage of each discrete cardiomyocyte unit on discrete time steps and spatial grids. Alternatively, dedicated simulation software platforms or custom numerical solvers can be used to achieve efficient parallel computation of the equation to simulate the electrophysiological activity of large-scale myocardial tissue models. This equation provides a mathematical framework for accurately calculating the dynamic changes of transmembrane voltage in myocardial tissue models, organically combining cellular-level ion dynamics with tissue-level electrical signal propagation.

[0080] It is understood that, through the above-described technical solution in this embodiment, when constructing a myocardial tissue model for cardiac electrophysiological simulation, discrete myocardial cell units are built based on an ion channel dynamics model. This allows each cell unit to accurately simulate its own ion channel dynamics and action potential behavior, thus providing a high-fidelity electrophysiological basis at the cellular level. Simultaneously, the diffusion coefficient D enables electrical coupling between discrete myocardial cell units, effectively simulating intercellular electrical signal propagation and resolving the lack of excitation-conduction coupling at the tissue scale, allowing action potentials to propagate realistically throughout the entire myocardial tissue model. Furthermore, the myocardial tissue model encompasses key transmembrane currents such as rapid sodium flow, L-type calcium flow, delayed rectified potassium flow, inward rectified potassium flow, sodium-calcium exchange flow, and sodium-potassium pump flow, ensuring comprehensive coverage of myocardial cell electrophysiological processes and avoiding simulation biases caused by the omission of important current mechanisms. Finally, by employing a transmembrane voltage acquisition equation, the cellular-level ion dynamics and tissue-level electrical signal propagation are precisely integrated, enabling accurate calculation of the dynamic changes in transmembrane voltage within the myocardial tissue model. In some of the embodiments described above in this application, an initial virtual individual model is proposed to construct a virtual population for cardiac electrophysiological simulation. However, in the process of its implementation, a key issue is how to ensure that the ion channel parameters and transmembrane transporter parameters are based on real physiological data, and to reasonably generate the parameters of the endocardium, media and epicardium to reflect individual differences and avoid model bias.

[0081] In response, this application further proposes a method for simulating cardiac electrophysiology, such as... Figure 3 As shown, the method includes: Step S301: Construct a myocardial tissue model. This model includes the endocardium, media, and epicardium, each composed of multiple discrete myocardial cell units. (See details...) Figure 1 Step S101 in the embodiment will not be described again here.

[0082] Step S302: Based on the physiological value range of preset parameters in the discrete cardiomyocyte units, an initial virtual individual model is formed. Specifically, this includes: Step S3021: Determine the physiological range of ion channel parameters and transmembrane transporter parameters using database data, which includes publicly available human electrophysiological experimental data and literature data.

[0083] Specifically, ion channel parameters and transmembrane transporter parameters are key variables describing the electrophysiological activity of cardiomyocytes. They determine the rate and characteristics of ion flow across the cell membrane. For example, ion channel parameters may include channel conductivity, activation / inactivation time constant, and voltage dependence; transmembrane transporter parameters may include the transporter's maximum transport rate and affinity. The accuracy of these parameters directly affects the physiological realism of the simulation model. The physiological range refers to the allowable variation of these parameters under normal human physiological conditions. Determining their physiological range is fundamental to ensuring that the simulation model conforms to biological reality. For example, the mean and standard deviation of parameters can be determined through statistical analysis of a large amount of experimental data, thereby defining a reasonable physiological range; alternatively, upper and lower limits of parameters can be set based on known physiological principles and clinical observations. Database data is the source of these physiological ranges. This can be a biomedical database specifically established for cardiac electrophysiological research, such as a public database containing the properties of various ion channels and transporters; or it can be experimental data accumulated within research institutions or pharmaceutical companies. Publicly available human electrophysiological experimental data refers to the electrophysiological response data of human cardiomyocytes or tissues obtained through in vitro cell experiments, in vitro tissue perfusion experiments, or in vivo clinical studies. This data is usually peer-reviewed and publicly published. Literature data refers to research results published in scientific journals, conference proceedings, or patent documents, which include the measurement, modeling, or inference of ion channel and transmembrane transporter parameters.

[0084] Step S3022: Based on the physiological value range, randomly generate endocardial parameter data.

[0085] Specifically, random generation refers to assigning values ​​to various parameters of the endocardium using random sampling within a defined physiological range. Randomness is a means of introducing individual differences. For example, uniform random sampling can be used, selecting parameter values ​​with equal probability within a given range; or Gaussian random sampling can be used to simulate the natural distribution of parameters in a population, where the mean and standard deviation can be set according to the physiological range. The endocardium is the innermost layer of the ventricular wall, and its electrophysiological properties differ from those of the media and epicardium. Generating endocardial parameter data is the first step in constructing a multilayer myocardial tissue model, providing a foundation for subsequently generating data for other layers.

[0086] Step S3023: The generated endocardial parameter data is used to generate mid-layer parameter data and epicardial parameter data according to an independent scaling factor.

[0087] Specifically, independent scaling factors are multipliers or additives used to adjust endocardial parameter data to generate mid- and epicardial parameter data. These factors reflect the systematic differences in electrophysiological properties between different layers of the myocardium. For example, the conductivity of certain ion channels may be 10% higher in the mid-middle layer than in the endocardium, corresponding to a scaling factor of 1.1; or, certain parameters may have a fixed offset between the mid- and epicardial layers. These scaling factors can be determined based on known physiological research findings or empirical data. The mid-middle layer is the thickest part of the ventricular wall, and the epicardium is the outermost layer. They each have unique electrophysiological properties, such as differences in action potential duration (APD). By applying independent scaling factors, it is ensured that the generated mid- and epicardial parameter data, while maintaining correlation with endocardial parameter data, also reflect their unique physiological characteristics, thereby constructing a virtual myocardial tissue model with physiological gradient differences.

[0088] Step S3024: Based on the endocardial parameter data, mid-layer parameter data, and epicardial parameter data, an initial virtual individual model is formed.

[0089] Step S303 involves simulating the electrophysiological performance of each initial virtual individual model and performing multi-level screening based on these performances to obtain the initial virtual population. Specifically, this includes: Step S3031: Under preset conditions, simulate the electrophysiological performance of the human body under resting, stress, and rhythmic change states, and screen the initial virtual individual models that can maintain physiological homeostasis to obtain screened virtual individual models. The preset conditions include sympathetic nerve excitation, shortened cardiac cycle, and prolonged cardiac cycle. (See details...) Figure 2 Step S2031 in the embodiment will not be described again here.

[0090] Step S304: Based on the calibrated QT, randomly sample and re-screen the initial virtual population to obtain the target virtual population. See details... Figure 1 Step S104 in the embodiment will not be described again here.

[0091] Step S305: Perform cardiac electrophysiological simulation based on the target virtual population. See details... Figure 1 Step S105 in the embodiment will not be described again here.

[0092] It is understood that, through the above-described technical solution in this embodiment, firstly, the database constructed using publicly available human electrophysiological experimental data and literature data ensures that the physiological value ranges of the determined ion channel parameters and transmembrane transporter parameters have high physiological authenticity and reliability, thereby avoiding model bias caused by the arbitrariness of parameter setting. Secondly, based on these real physiological value ranges, endocardial parameter data is randomly generated, effectively introducing the inherent variability between virtual individuals, enabling each initial virtual individual model to simulate unique electrophysiological characteristics, thus more realistically reflecting individual differences in the population. Furthermore, by introducing an independent scaling factor, the endocardial parameter data is reasonably extended to the middle and epicardium, ensuring that the gradient differences in electrophysiological characteristics between different layers within the myocardial tissue are accurately reflected, while maintaining the physiological coordination between parameters of each layer. Finally, integrating these layered parameter data to form an initial virtual individual model provides a physiologically based, individual-specific, and structurally complete basic model for subsequent cardiac electrophysiological simulation, significantly improving the accuracy of simulation results and clinical predictive ability, and effectively solving the problems of inaccurate and inconsistent parameter generation.

[0093] In some of the schemes described above in this application, a random sampling and re-screening of the initial virtual population based on the corrected QT is proposed to obtain the target virtual population. However, in this process, the QTc distribution of the initial virtual population may lack consistency with the real clinical QTc distribution, resulting in a deviation between the cardiac electrophysiological simulation results and clinical data, which in turn affects the accuracy and reliability of arrhythmia risk prediction.

[0094] In response, this application further proposes a method for simulating cardiac electrophysiology, such as... Figure 4 As shown, the method includes: Step S401: Construct a myocardial tissue model. This model includes the endocardium, media, and epicardium, each composed of multiple discrete myocardial cell units. (See details...) Figure 1 Step S101 in the embodiment will not be described again here.

[0095] Step S402: Based on the physiological value range of preset parameters in the discrete cardiomyocyte units, an initial virtual individual model is formed. Specifically, this includes: Step S4021: Determine the physiological range of ion channel parameters and transmembrane transporter parameters using database data, which includes publicly available human electrophysiological experimental data and literature data. (See details...) Figure 3 Step S3021 in the embodiment will not be described again here.

[0096] Step S4022: Based on the physiological value range, randomly generate endocardial parameter data. See details... Figure 3Step S3022 in the embodiment will not be described again here.

[0097] Step S4023: The generated endocardial parameter data is used to generate mid-layer parameter data and epicardial parameter data separately, according to an independent scaling factor. See details... Figure 3 Step S3023 in the embodiment will not be repeated here.

[0098] Step S4024: Based on the endocardial parameter data, mid-layer parameter data, and epicardial parameter data, an initial virtual individual model is formed. See details... Figure 3 Step S3024 in the embodiment will not be described again here.

[0099] Step S403 involves simulating the electrophysiological performance of each initial virtual individual model and performing multi-level screening based on these performances to obtain the initial virtual population. Specifically, this includes: Step S4031: Under preset conditions, simulate the electrophysiological performance of the human body under resting, stress, and rhythmic change states, and screen the initial virtual individual models that can maintain physiological homeostasis to obtain screened virtual individual models. The preset conditions include sympathetic nerve excitation, shortened cardiac cycle, and prolonged cardiac cycle. (See details...) Figure 2 Step S2021 in the embodiment will not be repeated here.

[0100] Step S404: Based on the calibrated QT, randomly sample and re-screen the initial virtual population to obtain the target virtual population. Specifically, this includes: Step S4041: Obtain the standard corrected QT distribution as the target reference distribution.

[0101] The "standard corrected QT distribution" refers to a statistical distribution that represents the QTc characteristics of the real population, obtained through large-scale clinical studies or authoritative medical databases. This distribution is typically obtained by analyzing ECG data from a large number of healthy individuals or populations with specific diseases, using appropriate correction formulas such as Bazett's formula or Fridericia's formula to correct the QT interval for heart rate, and then statistically analyzing its distribution characteristics. For example, this distribution can be obtained by fitting publicly available human electrophysiological experimental data and literature data through statistical analysis methods; or, published QTc distribution data can be directly obtained from authoritative medical databases or publicly available clinical trial reports as a standard. Obtaining this standard corrected QT distribution aims to provide an objective and reliable clinical reference benchmark for subsequent virtual population screening, thereby avoiding biases caused by subjective settings and ensuring the clinical relevance of the simulation results.

[0102] Step S4042: Using a sampling strategy, select a set of virtual individuals from the initial virtual population that are consistent with the target reference distribution to obtain the target virtual population.

[0103] The "sampling strategy" refers to a method of selectively selecting virtual individuals from the initial virtual population based on the degree of matching between their QTc values ​​and a pre-defined standard-corrected QT distribution. This strategy aims to ensure that the final target virtual population has a QTc distribution highly consistent with that of the real clinical population. Specifically, various statistical methods can be employed. For example, methods such as rejection sampling or Markov chain Monte Carlo can be used to compare the QTc value of each individual in the initial virtual population with the probability density function of the target reference distribution to determine whether to select that individual. Alternatively, stratified sampling or paired sampling can be used to divide the initial virtual population according to QTc intervals and extract individuals from each interval according to the proportion of the target reference distribution, ensuring the representativeness of each QTc interval. By employing this sampling strategy, biases caused by potential individual variability in the initial virtual population can be effectively eliminated, resulting in a more representative set of selected virtual individuals.

[0104] The "target virtual population" is a set of virtual individuals obtained after sampling and re-screening based on the standard-corrected QT distribution. This population exhibits a high degree of consistency with the distribution of the real clinical population in terms of QTc distribution, thus providing a more reliable and clinically representative model basis for subsequent cardiac electrophysiological simulations.

[0105] Step S405: Perform cardiac electrophysiological simulation based on the target virtual population. See details in [link / reference]. Figure 1 Step S105 in the embodiment will not be described again here.

[0106] It is understood that, through the above-described technical solution in this embodiment, this application effectively solves the problem of discrepancy between the QTc distribution of the initial virtual population and clinical data by introducing a standard corrected QT distribution as a target reference and combining it with a refined sampling strategy. Specifically, firstly, a standard corrected QT distribution is obtained, providing an objective and unified clinical benchmark for screening and avoiding errors caused by subjective settings. Based on this, a sampling strategy is used to dynamically select a set of virtual individuals consistent with the target reference distribution from the initial virtual population, ensuring representativeness at the population level and effectively eliminating bias caused by individual variations. Ultimately, the resulting target virtual population highly matches the real clinical population in terms of QTc distribution, improving the clinical relevance of cardiac electrophysiology simulation. In view of this, cardiac electrophysiology simulation based on this target virtual population can more accurately predict tissue-level arrhythmic events that may occur in clinical practice, such as premature ventricular contractions, T-wave alternation, or repolarization failure, thereby significantly improving the accuracy and reliability of drug safety assessment and antiarrhythmic drug development.

[0107] In some alternative embodiments, the myocardial tissue model includes a one-dimensional myocardial cable model.

[0108] Myocardial tissue models, including one-dimensional myocardial cable models, simplify the electrophysiological activity of the heart into a single-dimensional propagation model. This model simulates the conduction and repolarization of action potentials by abstracting myocardial tissue as an electrical "cable." Its core principle lies in balancing computational efficiency with physiological realism, making it particularly suitable for studying the propagation characteristics of electrical signals along a specific direction. One implementation involves constructing the myocardial tissue model as a structure composed of multiple discrete myocardial cell units connected in series along a one-dimensional direction. Each discrete myocardial cell unit is electrically coupled to adjacent units, for example, through resistance or conductivity D, thereby simulating intercellular electrical signal transmission. This effectively simulates the propagation speed of action potentials, conduction blockage, and the formation of repolarization gradients.

[0109] Employing a myocardial tissue model, including a one-dimensional myocardial cable model, reduces the computational resources and time required for cardiac electrophysiological simulation. It efficiently simulates the propagation of action potentials in myocardial tissue, including key electrophysiological phenomena such as conduction velocity, repolarization gradient, and excitation-conduction coupling. This enables more efficient large-scale simulation calculations when simulating the electrophysiological performance of each initial virtual individual model, supporting multi-level screening of the initial virtual population and ultimately obtaining a representative target virtual population.

[0110] In one example, the following provides a more detailed explanation of the above technical solution through a more specific example: A drug development team is working on a novel cardiac drug and needs to assess its potential effects on human cardiac electrophysiology and its risk of causing arrhythmias. Traditional in vitro single-cell experiments or animal models cannot fully reflect the complexity at the human tissue scale, are difficult to accurately predict clinical arrhythmia events, and their predictions deviate from real clinical data.

[0111] To address these issues, the team employed a cardiac electrophysiological simulation method. First, a myocardial tissue model was constructed. This model is a one-dimensional myocardial cable model comprising three regions: the endocardium, media, and epicardium. Each region is composed of multiple discrete myocardial cell units. These discrete myocardial cell units are constructed based on an ion channel kinetic model and electrically coupled via a diffusion coefficient D. The transmembrane currents in the myocardial tissue model include at least rapid sodium flow, L-shaped calcium flow, delayed rectified potassium flow, inward rectified potassium flow, sodium-calcium exchange flow, and sodium-potassium pump flow. The construction of this tissue model overcomes the limitations of related techniques in reflecting the complexity of tissue scale.

[0112] Next, based on the physiological value ranges of preset parameters in discrete cardiomyocyte units, an initial virtual individual model was formed. The team determined the physiological value ranges of ion channel parameters and transmembrane transporter parameters by analyzing publicly available human electrophysiological experimental data and literature data. Based on these ranges, endocardial parameter data were randomly generated. Subsequently, the generated endocardial parameter data was used to generate mid-layer and epicardial parameter data according to independent scaling factors. Finally, an initial virtual individual model was formed based on the endocardial, mid-layer, and epicardial parameter data. By repeating this process, a large number of initial virtual individual models with physiological variability were generated, simulating individual differences in the real population.

[0113] Then, the electrophysiological performance of each initial virtual individual model is simulated, and multi-level screening is performed based on the electrophysiological performance to obtain the initial virtual population. Under preset conditions, such as sympathetic nerve excitation, shortened cardiac cycle, and prolonged cardiac cycle, the electrophysiological performance of the human body under resting, stress, and rhythmic change states is simulated, respectively. The initial virtual individual models that can maintain physiological homeostasis are screened to obtain screened virtual individual models. The screening process includes at least one of the following: cell-level screening, tissue-level screening, and pseudo-ECG-level screening. In cell-level screening, the action potentials of discrete myocardial cell units and the calcium ion concentrations within discrete myocardial cell units are detected to see if they meet the corresponding screening conditions. In tissue-level screening, standard stimulation is applied to the myocardial tissue model to detect whether action potential propagation blockage or repolarization failure occurs. This solves the problem that related technologies are difficult to predict tissue-level arrhythmic events. In pseudo-ECG-level screening, the QRS waveform or QT interval in the generated pseudo-ECG signal is detected to see if it exceeds the normal physiological range. The multi-level screening mechanism links single-cell level indicators with tissue-level events and clinical electrocardiogram characteristics, making up for the shortcomings of related technical substitute indicators being disconnected from clinical characteristics.

[0114] Building upon this foundation, the initial virtual population was randomly sampled and re-screened based on the corrected QT to obtain the target virtual population. The team obtained a standard corrected QT distribution as the target reference distribution, derived from large-scale clinical data. Through a sampling strategy, a set of virtual individuals consistent with the target reference distribution was selected from the initial virtual population to obtain the target virtual population. This step ensured a high degree of match between the QT characteristics of the virtual population and the real clinical population, significantly reducing the bias between model predictions and clinical data, and improving the accuracy of risk prediction.

[0115] Finally, cardiac electrophysiological simulations were conducted based on the target virtual population. The team integrated the pharmacological parameters of the new drug candidate into the model of the target virtual population to simulate the electrophysiological responses of the heart under drug action. By analyzing the electrophysiological performance of the target virtual population under drug action, such as changes in the QT interval and the incidence of arrhythmic events, the proarrhythmic risk of the drug can be accurately assessed. This method can provide more comprehensive and accurate risk prediction than traditional methods, providing strong support for the preclinical evaluation of drugs.

[0116] This embodiment also provides a cardiac electrophysiology simulation device for implementing the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0117] This embodiment provides a cardiac electrophysiology simulation device, such as Figure 5 As shown, it includes: Module 501 is used to construct a myocardial tissue model, wherein the myocardial tissue model includes the endocardium, media, and epicardium, and the endocardium, media, and epicardium are all composed of multiple discrete myocardial cell units; The forming module 502 is used to form an initial virtual individual model based on the physiological value range of preset parameters in the discrete cardiomyocyte unit, wherein the preset parameters include ion channel parameters and transmembrane transporter parameters; The screening module 503 is used to simulate the electrophysiological performance of each initial virtual individual model and perform multi-level screening based on the electrophysiological performance to obtain the initial virtual population. Sampling module 504 is used to randomly sample and re-screen the initial virtual population based on the calibration QT to obtain the target virtual population; Simulation module 505 is used for cardiac electrophysiological simulation based on a target virtual population.

[0118] In some alternative implementations, the filtering module 503 includes: The first unit is used to simulate the electrophysiological performance of the human body under preset conditions, namely resting, stress and rhythm change states, and to screen the initial virtual individual models that can maintain physiological homeostasis to obtain the screened virtual individual models. The preset conditions include sympathetic nerve excitation, shortened cardiac cycle and prolonged cardiac cycle.

[0119] The cardiac electrophysiology simulation device provided in this application can execute the cardiac electrophysiology simulation method provided in any embodiment of this application, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0120] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0121] The following is a detailed reference. Figure 6 This diagram illustrates a suitable structural schematic for implementing the electronic device described in the embodiments of this application. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 601, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 602 or a program loaded from memory 608 into random access memory (RAM) 603. The RAM 603 also stores various programs and data required for the operation of the electronic device. The processor 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0122] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 608 including, for example, magnetic tapes, hard disks, etc.; and communication devices 609. Communication device 609 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 6 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0123] Specifically, according to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a memory 608, or installed from a ROM 602. When the computer program is executed by the processor 601, it performs the functions defined in the cardiac electrophysiology simulation method of this application.

[0124] Figure 6The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0125] This application also provides a computer-readable storage medium. The methods described in this application can be implemented in hardware or firmware, or implemented as recordable on a storage medium, or implemented as computer code downloaded over a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and subsequently stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the cardiac electrophysiology simulation method shown in the above embodiments is implemented.

[0126] A portion of this application can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to this application through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0127] Although embodiments of this application have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of this application, and all such modifications and variations fall within the scope defined by the appended claims.

Claims

1. A method for simulating cardiac electrophysiology, characterized in that, The method includes: A myocardial tissue model is constructed, wherein the myocardial tissue model includes the endocardium, the media, and the epicardium, and the endocardium, the media, and the epicardium are all composed of multiple discrete myocardial cell units; An initial virtual individual model is formed based on the physiological value range of preset parameters in the discrete cardiomyocyte unit, wherein the preset parameters include ion channel parameters and transmembrane transporter parameters; The electrophysiological performance of each initial virtual individual model is simulated, and multi-level screening is performed based on the electrophysiological performance to obtain the initial virtual population; Based on the corrected QT, the initial virtual population is randomly sampled and re-screened to obtain the target virtual population; Based on the target virtual population, cardiac electrophysiological simulation was performed.

2. The method according to claim 1, characterized in that, The process of simulating the electrophysiological performance of each initial virtual individual model and performing multi-level screening based on the electrophysiological performance to obtain an initial virtual population includes: Under preset conditions, the electrophysiological performance of the human body under resting, stress and rhythm change states is simulated respectively, and the initial virtual individual model that can maintain physiological homeostasis is screened to obtain the screened virtual individual model. The preset conditions include sympathetic nerve excitation, shortened cardiac cycle and prolonged cardiac cycle.

3. The method according to claim 2, characterized in that, The screening of the initial virtual individual models capable of maintaining physiological homeostasis includes: For the initial virtual individual model, at least one of the following should be performed: cell-level screening, tissue-level screening, and pseudo-ECG-level screening. The cell-level screening includes detecting whether the action potential of the discrete cardiomyocyte units and the calcium ion concentration within the discrete cardiomyocyte units meet the corresponding screening criteria. The tissue-level screening includes applying standard stimulation to the myocardial tissue model and detecting whether action potential propagation blockage or repolarization failure occurs. The pseudo-ECG screening includes detecting whether the QRS waveform or QT interval in the generated pseudo-ECG signal exceeds the normal physiological range.

4. The method according to claim 3, characterized in that, The calculation of the pseudo-ECG signal includes: ， in, x is the location of a discrete cardiomyocyte unit. and The location of the electrode used to measure the ECG signal, where D is the conductivity. Let be the gradient of the potential field.

5. The method according to claim 1, characterized in that, The discrete cardiomyocyte units are constructed based on an ion channel dynamics model, and the discrete cardiomyocyte units are electrically coupled through a diffusion coefficient D. The transmembrane currents in the myocardial tissue model include at least a fast sodium current, an L-type calcium current, a delayed rectified potassium current, an inward rectified potassium current, a sodium-calcium exchange current, and a sodium-potassium pump current. The acquisition of transmembrane voltage in the myocardial tissue model includes: ,in, For transmembrane current, C represents the stimulation current that controls the rhythm. m It is the capacitance of the myocardial fiber membrane.

6. The method according to claim 1, characterized in that, The step of forming an initial virtual individual model based on the physiological value range of preset parameters in the discrete cardiomyocyte units includes: The physiological range of values ​​for the ion channel parameters and the transmembrane transporter parameters is determined using database data, wherein the database data includes publicly available human electrophysiological experimental data and literature data. Based on the physiological range, the parameter data of the endocardium are randomly generated; The generated endocardial parameter data is used to generate mid-layer parameter data and epicardial parameter data separately according to an independent scaling factor; An initial virtual individual model is formed based on the endocardial parameter data, the mid-layer parameter data, and the epicardial parameter data.

7. The method according to claim 1, characterized in that, The step of randomly sampling and re-screening the initial virtual population based on the corrected QT to obtain the target virtual population includes: Obtain the standard corrected QT distribution as the target reference distribution; By using a sampling strategy, a set of virtual individuals that are consistent with the target reference distribution is selected from the initial virtual population to obtain the target virtual population.

8. A cardiac electrophysiology simulation device, characterized in that, The device includes: A construction module is used to construct a myocardial tissue model, wherein the myocardial tissue model includes an endocardium, a media, and an epicardium, and the endocardium, the media, and the epicardium are all composed of multiple discrete myocardial cell units; A forming module is used to form an initial virtual individual model based on the physiological value range of preset parameters in the discrete cardiomyocyte unit, wherein the preset parameters include ion channel parameters and transmembrane transporter parameters; A screening module is used to simulate the electrophysiological performance of each initial virtual individual model and perform multi-level screening based on the electrophysiological performance to obtain an initial virtual population. The sampling module is used to randomly sample and re-screen the initial virtual population based on the calibration QT to obtain the target virtual population; The simulation module is used to perform cardiac electrophysiological simulations based on the target virtual population.

9. An electronic device, characterized in that, include: A memory and a processor are communicatively connected, the memory stores computer instructions, and the processor executes the cardiac electrophysiological simulation method according to any one of claims 1 to 7 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the cardiac electrophysiological simulation method according to any one of claims 1 to 7.

Images