Mapping of computational cardiac depolarization and cardiac repolarization simulation libraries for non-invasive arrhythmia risk stratification.

Abstract

A non-invasive method for cardiac arrhythmia risk stratification can include identifying, based at least on the patient's electrical recording, a cardiac depolarization simulation and a cardiac repolarization simulation corresponding to the patient's electrical recording. Based on the cardiac depolarization simulation and the cardiac repolarization simulation, one or more regions of an increased spatial repolarization gradient can be identified, in which a first region of the patient's myocardium exhibits a first repolarization rate that, when divided by the spatial distance between the two regions, differs from a second repolarization rate of a second region of the myocardium by a predetermined amount, i.e., a predetermined threshold. Based on the magnitude of the increased spatial repolarization gradient, a risk of cardiac arrhythmia for the patient can be identified. Furthermore, a treatment plan for the patient can be determined based on the magnitude and / or location of the increased spatial repolarization gradient.

Description

Technical Field

[0001] Related Applications This application claims priority to U.S. Provisional Application No. 63 / 104,930, filed October 23, 2020, entitled "COMPUTATIONAL CARDIAC DEPOLARIZATION AND REPOLARIZATION SIMULATION LIBRARY MAPPING FOR NON-INVASIVE ARRHYTHMIA RISK STRATIFICATION", the entire disclosure of which is hereby incorporated by reference.

[0002] [[ID=*11]]Technical Field The subject matter described herein generally relates to computational modeling and simulation, and more specifically, to mapping of simulation libraries for non-invasive arrhythmia mapping and risk stratification.

[0003] Background Cardiac arrhythmia is a common disease in which the heart contracts in a suboptimal state due to abnormal electrical signals in the heart. The resulting abnormal heartbeat or arrhythmia can occur in the atria of the heart (e.g., atrial fibrillation (AF)) and / or the ventricles of the heart (e.g., ventricular tachycardia (VT) or ventricular fibrillation (VF)). Treatments for cardiac arrhythmia attempt to address the mechanisms that cause persistent and / or clinically significant symptoms, such mechanisms including, for example, stable electrical rotors, recurring electrical focal sources (where excitation repeatedly appears from a single point and spreads radially outwards), reentry circuits, and / or the like. If left untreated, cardiac arrhythmia can cause serious health complications such as morbidity (e.g., fainting, stroke, and / or the like) and death (e.g., sudden cardiac death (SCD)).

[0004] Summary Note: There seems to be a duplicate "Technical Field" in the original text. I've translated both occurrences as per the instructions. If this is an error in the original, you may want to correct it in the source material.A system, method, and product are provided, which includes a computer program product for mapping a library of computational cardiac depolarization and cardiac repolarization simulations for non-invasive arrhythmia risk stratification. In one embodiment, a system for non-invasive arrhythmia risk stratification is provided. The system may include at least one processor and at least one memory storing instructions that, when executed by the at least one processor, cause the following operations: the operation of identifying cardiac depolarization and cardiac repolarization simulations in a computation library that correspond to the patient's electrical record, and identifying one or more regions of increased spatial repolarization gradients based at least on the cardiac depolarization and cardiac repolarization simulations, wherein the region is a first region of the patient's myocardium. (area) This shows the first repolarization rate, and the first repolarization rate corresponds to the second region of the myocardium. (area) The second repolarization rate is the two regions (area) This may include identifying one or more regions that, when divided by the spatial distance between them, differ by a predetermined amount, i.e., by a predetermined threshold, and identifying the risk of cardiac arrhythmia or sudden cardiac death (SCD) for the patient, at least based on the magnitude of the increased spatial repolarization gradient.

[0005] In some variations, one or more features, including those disclosed herein, can be optionally included in any viable combination. The operation may further include determining a treatment plan for the patient, at least based on the magnitude of the increased spatial repolarization gradient.

[0006] In some variant forms, the treatment plan can be determined to include defibrillator implantation or invasive electrophysiological testing and ablation, at least based on the magnitude of the increased spatial repolarization gradient.

[0007] In some variant forms, the treatment plan may include identifying locations for targeted therapies such as radiofrequency catheter ablation, cryoablation, radiofrequency ultrasound ablation, laser therapy, or pulsed field ablation, based at least on the location of one or more regions of the amplified spatial repolarization gradient.

[0008] In some variant forms, targeted therapy may include catheter ablation and / or stereotactic radiotherapy (SAbR).

[0009] In some variants, cardiac depolarization simulations may include ventricular excitation simulations, while cardiac repolarization simulations may include ventricular recovery simulations.

[0010] In some variant forms, the operation may further include generating a computational library containing multiple cardiac depolarization simulations and multiple cardiac repolarization simulations, where the multiple cardiac depolarization simulations and multiple cardiac repolarization simulations correspond to various cardiac geometries, cardiac orientation, scar structure, degree of cardiac fibrosis and scarring, depolarization patterns, and / or types of excitation, and identifying within the computational library the cardiac depolarization simulations and cardiac repolarization simulations corresponding to the patient's electrical records.

[0011] In some variants, the computational library can be supplemented with clinical patient samples in which the location of the source of the arrhythmic substrate is known, providing additional data for comparison with the patient's electrical recordings.

[0012] In some variant forms, the computational library may include clinical patient samples with known sources of arrhythmic substrates to form reference data for comparison with the patient's electrical recordings.

[0013] In some variations, the operation may further include identifying a subset of simulations corresponding to the patient's anatomical structure from a computational library, based at least on patient-related clinical data, and within that subset of simulations corresponding to the patient's anatomical structure, identifying cardiac depolarization and cardiac repolarization simulations corresponding to the patient's electrical recordings.

[0014] In some variant forms, clinical data can include patient demographics.

[0015] In some variant forms, clinical data may include cardiac imaging data, which may show the location of one or more scar tissues, borderline tissues, and normal tissues, cardiac chamber sizes, the presence of hypertrophy or enlargement, the location of fibrosis, areas of normal and abnormal contractility, and / or areas of wall thinning.

[0016] In some variant forms, the operation may further include generating a custom computational library that includes one or more cardiac depolarization and / or cardiac repolarization simulations specific to the patient's anatomical structure, based at least on the patient's clinical data, in response to the inability to identify a subset of simulations corresponding to the patient's anatomical structure.

[0017] In some variations, the operation may further include applying a machine learning model trained to determine whether cardiac depolarization and cardiac repolarization simulations match the patient's electrical recordings.

[0018] In some variations, machine learning models can include neural networks, regression models, instance-based models, regularization models, decision trees, random forests, Bayesian models, clustering models, associative models, dimensionality reduction models, and / or ensemble models.

[0019] In some variations, the operation may further include applying one or more signal processing techniques to the patient's electrical recording.

[0020] In some variations, one or more signal processing techniques may include recording, filtering, digitization, transformation, and / or spatial analysis.

[0021] In some variations, the electrical recording may include one or more of the following: electrophoresis, vector diagrams, electrocardiograms, electroencephalograms, or vector electrocardiograms.

[0022] In some variations, the electrical recording may further include one or more surface potential recordings.

[0023] In some variations, the electrical recording may include electrocardiogram imaging (ECGi) data, which may include one or more surface potential recordings.

[0024] In some variations, the operation may further include identifying one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block, based at least on cardiac depolarization and cardiac repolarization simulations, and identifying the risk of cardiac arrhythmia to the patient, based at least on the presence and / or absence of one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, and / or conduction block.

[0025] In some variations, the operation may further include determining a treatment plan for the patient based on at least one or more areas of early excitation, slow conduction, independent excitation pathways, delayed excitation, and / or conduction block.

[0026] In some variations, the treatment plan can target one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, and / or conduction block individually or in groups.

[0027] In some variations, the treatment plan can be determined to include one or more drug therapies based at least on the presence and / or absence of one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block.

[0028] In another aspect, a method for non-invasive arrhythmia risk stratification is provided. The method includes identifying cardiac depolarization simulations and cardiac repolarization simulations corresponding to a patient's electrical recordings within a computational library, identifying one or more regions of increased spatial repolarization gradient based at least on the cardiac depolarization simulations and the cardiac repolarization simulations, wherein in the region, a first region of the patient's myocardium exhibits a first repolarization rate, and the first repolarization rate is different by a predetermined amount, i.e., a predetermined threshold, when divided by the spatial distance between the two regions, from a second repolarization rate of a second region of the myocardium, identifying one or more regions, and identifying the risk of cardiac arrhythmia for the patient based at least on the magnitude of the increased spatial repolarization gradient.

[0029] In some variations, one or more features including the following features disclosed herein can be optionally included in any executable combination. The operation can further include determining a treatment plan for the patient based at least on the magnitude of the increased spatial repolarization gradient.

[0030] In some variations, the treatment plan can be determined to include implanting a defibrillator or invasive electrophysiological examination and ablation based at least on the magnitude of the increased spatial repolarization gradient.

[0031] In some variants, the treatment plan may include identifying locations for targeted therapy based on the location of at least one or more regions of the amplified spatial repolarization gradient.

[0032] In some variant forms, targeted therapy may include catheter ablation and / or stereotactic radiotherapy (SAbR).

[0033] In some variants, cardiac depolarization simulations may include ventricular excitation simulations, while cardiac repolarization simulations may include ventricular recovery simulations.

[0034] In some variations, the method may further include generating a computational library containing multiple cardiac depolarization simulations and multiple cardiac repolarization simulations, where the multiple cardiac depolarization simulations and multiple cardiac repolarization simulations correspond to various cardiac geometries, cardiac orientations, scar structures, degrees of cardiac fibrosis and scarring, depolarization patterns, and / or types of excitation, and identifying within the computational library the cardiac repolarization simulations and cardiac depolarization simulations corresponding to the patient's electrical recordings.

[0035] In some variations, the method may further include identifying a subset of simulations corresponding to the patient's anatomical structure from a computational library, at least based on patient-related clinical data, and within that subset of simulations corresponding to the patient's anatomical structure, identifying cardiac repolarization and cardiac depolarization simulations corresponding to the patient's electrical recordings.

[0036] In some variant forms, clinical data can include patient demographics.

[0037] In some variant forms, clinical data may include cardiac imaging data, which may show the location of one or more scar tissues, borderline tissues, and normal tissues, cardiac chamber sizes, the presence of hypertrophy or enlargement, the location of fibrosis, areas of normal and abnormal contractility, and / or areas of wall thinning.

[0038] In some variant forms, the method may further include generating a custom computational library that includes one or more cardiac depolarization and / or cardiac repolarization simulations specific to the patient's anatomical structure, based at least on the patient's clinical data, in response to the inability to identify a subset of simulations corresponding to the patient's anatomical structure.

[0039] In some variations, the method may further include applying a machine learning model trained to determine whether cardiac depolarization and cardiac repolarization simulations match the patient's electrical recordings.

[0040] In some variations, machine learning models can include neural networks, regression models, instance-based models, regularization models, decision trees, random forests, Bayesian models, clustering models, associative models, dimensionality reduction models, and / or ensemble models.

[0041] In some variations, the method may further include applying one or more signal processing techniques to the patient's electrical recording.

[0042] In some variations, one or more signal processing techniques may include recording, filtering, digitization, transformation, and / or spatial analysis.

[0043] In some variations, the electrical recording may include one or more of the following: electrophoresis, vector diagrams, electrocardiograms, electroencephalograms, or vector electrocardiograms.

[0044] In some variations, the electrical recording may further include one or more surface potential recordings.

[0045] In some variations, the electrical recording may include electrocardiogram imaging (ECGi) which includes one or more surface potential recordings.

[0046] In some variations, the method may further include identifying one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block, based at least on cardiac depolarization and cardiac repolarization simulations, and identifying the risk of cardiac arrhythmia to a patient, based at least on the presence and / or absence of one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block.

[0047] In some variations, the method may further include determining a treatment plan for a patient based on at least one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block.

[0048] In some variant forms, the treatment plan may target one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction blockage, individually or in groups.

[0049] In some variant forms, the treatment plan can be determined to include one or more pharmacotherapy therapies based at least on the presence and / or absence of one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block.

[0050] In another embodiment, a non-temporary computer-readable medium is provided that stores instructions that, when executed by at least one data processor, cause the following operations: The operations include identifying within a computational library a cardiac depolarization simulation and a cardiac repolarization simulation corresponding to a patient's electrical record; identifying one or more regions of an increased spatial repolarization gradient, where a first region of the patient's myocardium exhibits a first repolarization rate, and the first repolarization rate differs from a second repolarization rate of a second region of the myocardium by a predetermined amount, i.e., a predetermined threshold, when divided by the spatial distance between the two regions; and identifying the risk of cardiac arrhythmia to the patient, at least based on the magnitude of the increased spatial repolarization gradient.

[0051] In another embodiment, a device for non-invasive arrhythmia risk stratification is provided. The device may include means for identifying cardiac depolarization and cardiac repolarization simulations corresponding to a patient's electrical records within a computational library; means for identifying one or more regions of an increased spatial repolarization gradient, based at least on the cardiac depolarization and cardiac repolarization simulations, wherein a first region of the patient's myocardium exhibits a first repolarization rate, the first repolarization rate differs from a second repolarization rate of a second region of the myocardium by a predetermined threshold, and is spatially adjacent to the second repolarization rate of the second region of the myocardium by a predetermined threshold; and means for identifying the risk of cardiac arrhythmia to the patient, based at least on the magnitude of the increased spatial repolarization gradient.

[0052] Implementations of the subject matter may include systems and methods matching one or more of the features described herein, and articles comprising tangibly embodied machine-readable media capable of operating to result in the operations described herein by one or more machines (e.g., a computer). Similarly, computer systems are also described that may include one or more processors and one or more memories coupled to one or more processors. Memories, which may include computer-readable storage media, may contain, encode, store, or similarly contain one or more programs causing one or more processors to perform one or more of the operations described herein. Computer implementations matching one or more implementations of the subject matter may be implemented by one or more data processors residing in a single computing system or a group of computing systems. Such a group of computing systems may be interconnected and exchange data and / or instructions or other directives or similar through one or more connections, such connections may include, for example, connections via a network (e.g., the Internet, a wireless wide area network, a local area network, a wide area network, a wired network, or similar), direct connections between one or more of the group of computing systems, and / or similar.

[0053] Details of one or more variations of the subject matter described herein are illustrated in the accompanying drawings and the following description. Other features and advantages of the subject matter described herein may be apparent from the specification and drawings and from the claims. Certain features of the subject matter disclosed herein are described for illustrative purposes with respect to the mapping of computational cardiac depolarization and cardiac repolarization simulation libraries for non-invasive arrhythmia risk stratification, but it should be readily understood that such features are not intended to be limiting. The claims following this disclosure are intended to define the scope of the protected subject matter.

[0054] Description of the drawing The accompanying drawings incorporated herein and constituting part of this specification illustrate specific embodiments of the subject matter disclosed herein and, together with the description thereof, are useful for illustrating some of the principles relating to the disclosed implementations. [Brief explanation of the drawing]

[0055] [Figure 1] This is a system diagram illustrating an example of a risk stratification system based on several exemplary embodiments. [Figure 2] This figure shows examples of non-invasive 12-lead electrocardiograms (ECGs) and supplemental surface potential maps for calculating spatial repolarization gradients and the risk of cardiac arrhythmias, according to several exemplary embodiments. [Figure 3] This figure shows an example of a computational library of cardiac geometry that can be used to generate simulation libraries for cardiac depolarization and cardiac repolarization. According to several exemplary embodiments, within each geometry, various scar configurations of cardiac arrhythmia simulations, the degree of cardiac fibrosis and scarring, conduction velocity, and other variables can be modified. [Figure 4] This figure shows an example of a computational library for cardiac depolarization and cardiac repolarization simulations, based on several exemplary embodiments. [Figure 5]This figure shows exemplary imaging modalities for localizing scar tissue, borderline tissue, normal tissue, cardiac chamber size, and myocardial contractility, according to several exemplary embodiments. [Figure 6] This figure shows an example of a process for calculating the size and location of an increased spatial repolarization gradient region using non-invasive 12-lead electrocardiogram (ECG) data, with or without surface potential recording, according to several exemplary embodiments. [Figure 7] This figure shows an example of a biventricular implantable cardioverter-defibrillator (ICD) according to several exemplary embodiments. [Figure 8] This figure shows an example of targeted therapy for cardiac arrhythmias, according to several exemplary embodiments. [Figure 9] This flowchart shows an example of a process for non-invasive arrhythmia risk stratification, based on several exemplary embodiments. [Figure 10] This block diagram shows a computing system in several exemplary embodiments. [Figure 11] This figure shows an example of an experimental setup for capturing patient electrical recordings, according to several exemplary embodiments. [Figure 12] This figure shows an example of data tracing according to several exemplary embodiments. [Figure 13] This figure shows an example of a three-dimensional plot of cardiac depolarization and cardiac repolarization according to several exemplary embodiments. [Figure 14] This figure shows an example of a plot comparing the spatial heterogeneity of cardiac repolarization in patients with ventricular arrhythmias with that of control patients, according to several exemplary embodiments.

[0056] Similar reference symbols, when useful, refer to similar structures, features, or elements.

[0057] Detailed explanation Sudden cardiac death (SCD) affects approximately 400,000 people annually in the United States and about 3 million worldwide. Currently, techniques that measure left ventricular ejection fraction (LFC), i.e., the volume fraction of blood ejected from the left ventricle, are used to assess a patient's risk of ventricular arrhythmias and sudden cardiac death. In many cases, LFC can also be used to determine whether a patient is a suitable candidate for implantable cardioverter-defibrillator (ICD) implantation or for invasive electrophysiological testing and catheter ablation. However, LFC has suboptimal sensitivity and specificity for ventricular arrhythmias and sudden cardiac death. Furthermore, LFC cannot predict death due to worsening heart failure.

[0058] Cardiac arrhythmias (e.g., atrial fibrillation, ventricular tachycardia, ventricular fibrillation) can be treated by targeting mechanisms that cause persistent and / or clinically significant symptom development, such mechanisms including, for example, stable electrical rotors, recurring electrical focal sources, reentry circuits, and / or similar phenomena. Ablation is one exemplary treatment for cardiac arrhythmias, in which radiofrequency, cryogenic, ultrasound, laser energy, pulsed field ablation, and / or radiation (e.g., stereotactic radiotherapy (SAbR)) can be applied to the cardiac arrhythmia source. The resulting damage can alleviate the cardiac arrhythmia by interrupting and / or removing the unstable electrical signals that cause the abnormal excitation of the heart.

[0059] A major challenge in managing arrhythmias is the difficulty in predicting individual risk for ventricular tachycardia (VT) and ventricular fibrillation (VF). Currently, the majority of ventricular arrhythmic events occur in “low-risk” populations. Therefore, the majority of patients (e.g., 55% of men and 65% of women) experience sudden cardiac death (SCD) as the first expression of their increased arrhythmia risk. Furthermore, risk stratification in patients with hereditary cardiomyopathy remains a challenge. The risk of sudden cardiac death in patients with hereditary familial cardiomyopathy is influenced by numerous factors, including cardiomyopathy-specific electrophysiological changes (e.g., conduction slowing, action potential morphology and recovery, intracellular calcium processing, and mechoelectric feedback) that may act as triggers for arrhythmias and provide myocardial structural remodeling (e.g., hypertrophy, fibrosis, fibrolipid deposition) that perpetuate arrhythmias. The varying penetrance of many cardiomyopathy subtypes, and / or unknown severity, further complicates the assessment of arrhythmia risk. Traditionally, the risk of sudden cardiac death in patients with hereditary cardiomyopathy has been assessed using left ventricular ejection fraction or score criteria. However, in practice, the sensitivity and specificity of these conventional methods are suboptimal, leaving patients uncertain about their individual arrhythmia risk and appropriate treatment options (e.g., implantable cardioverter-defibrillator (ICD) therapy or catheter ablation).

[0060] Furthermore, monitoring the initiation and maintenance of QT-prolonging drug therapy, including antiarrhythmic drugs such as dofetilide and sotalol, remains challenging, given the individual patient responses to sodium or potassium channel blockers. Such drug therapies, which delay ventricular recovery (or ventricular repolarization), can prolong the patient's QT interval. Therefore, previous attempts have focused on prolonging the QT interval or corrected QT (QTc) interval on the patient's surface electrocardiogram. While these measurements assess important physiological changes, their sensitivity and specificity for accurately monitoring QT-prolonging drug therapy remain suboptimal.

[0061] In light of the suboptimal sensitivity and specificity of conventional methodologies for assessing the risk of ventricular arrhythmias and conventional methods for identifying the location of arrhythmias that initiate or sustain ventricular arrhythmias, the various implementations of this disclosure include rapid, non-invasive tools for providing a more accurate assessment of ventricular arrhythmia risk across diverse patient populations and for localizing potentially arrhythmia-inducing regions. According to some exemplary embodiments, the assessment engine can be configured to provide a patient-specific assessment of arrhythmia-inducing factors. In particular, the assessment engine can stratify the risk of cardiac arrhythmias based on a spatial repolarization gradient, which corresponds to a magnitude such that the refractory period of adjacent regions within the myocardium changes by this magnitude as a function of distance across cardiac tissue. For example, the evaluation engine can identify potential arrhythmogenic conditions that are likely to result in excitation wave splitting, re-entry, and life-threatening arrhythmias, based at least on adjacent myocardial regions that show a difference above a threshold with respect to their respective cardiac repolarization rates (e.g., a first region of long repolarization directly adjacent to a second region of short repolarization).

[0062] Figure 1 shows a system diagram illustrating an example of a risk stratification system 100 according to several exemplary embodiments. Referring to Figure 1, the risk stratification system 100 may include an evaluation engine 110, a data store 120 storing one or more computing libraries 125, and client devices 130. As shown in Figure 1, the evaluation engine 110, the data store 120, and the client devices 130 can be communicably coupled via a network 140. The data store 120 may be a database including, for example, a graph database, an in-memory database, a relational database, a non-SQL (NoSQL) database, and / or similar. The client devices 130 may be processor-based devices including, for example, a mobile phone, a smartphone, a tablet computer, a laptop computer, a desktop, a workstation, and / or similar. The network 140 may be a wired network and / or a wireless network including, for example, a wide area network (WAN), a local area network (LAN), a virtual local area network (VLAN), a public land mobile network (PLMN), the internet, and / or similar.

[0063] In some exemplary embodiments, the evaluation engine 110 can be configured to identify a patient's risk of cardiac arrhythmias based on at least one region of an increased spatial repolarization gradient associated with the patient. As used herein, the term “spatial repolarization gradient” can refer to a magnitude such that the refractory period of adjacent regions within the myocardium changes by this magnitude as a function of distance across the cardiac tissue. For example, the evaluation engine 110 may determine that a patient exhibits a potentially arrhythmic-promoting condition that is likely to result in excitation wave splitting, re-entry, and life-threatening arrhythmias, based at least on the presence of adjacent myocardial regions where the respective cardiac repolarization rates differ by a threshold (e.g., a first region of long repolarization directly adjacent to a second region of short repolarization).

[0064] In some exemplary embodiments, the computation engine 110 can identify one or more regions of a patient's increased spatial repolarization gradient based on at least one or more computational simulations of cardiac depolarization (e.g., ventricular excitation) and / or cardiac repolarization (e.g., ventricular recovery). One or more computational simulations can be selected from a computation library 125, which is stored, for example, in a data store 120. For example, the computation library 125 may include cardiac depolarization simulations (e.g., ventricular excitation simulations) relating to various cardiac geometries, cardiac orientation, scar structure, degree of cardiac fibrosis and scarring, depolarization patterns, types of excitation (e.g., left bundle branch block, right bundle branch block, left anterior fascicular block, left posterior fascicular block, ventricular premature contractions, ventricular tachycardia, and ventricular fibrillation), and combinations thereof. Alternatively and / or additionally, the computational library 125 may include cardiac repolarization simulations (e.g., ventricular recovery simulations) for various cardiac geometries, cardiac orientations, scar structures, degrees of cardiac fibrosis and scarring, depolarization patterns, types of excitation (e.g., left bundle branch block, right bundle branch block, left anterior fascicular block, left posterior fascicular block, ventricular premature contractions, ventricular tachycardia, and ventricular fibrillation), and combinations thereof. In some cases, the computational library 125 may include a first computational library for cardiac depolarization simulations (e.g., ventricular excitation simulations) and a second computational library for cardiac repolarization simulations (e.g., ventricular recovery simulations).

[0065] To further illustrate, Figure 3 shows an example of various cardiac geometries used to generate the computational library 125, which has cardiac depolarization simulations (e.g., ventricular depolarization simulations) with respect to various scar structures, conduction velocities, cardiac orientation, etc. Figure 4 shows another example of the computational library 125, which has simulations of cardiac depolarization (e.g., leading wavefronts in bright regions) and cardiac repolarization (e.g., leading wavefronts in dark regions).

[0066] In some exemplary embodiments, the evaluation engine 110 can identify one or more simulations corresponding to a patient's electrical recording within the computational library 125. Examples of electrical recordings include electrophoresis, surface potential recordings (e.g., electrocardiogram imaging (ECGi)), vector diagrams, electrocardiograms, electroencephalograms, and vector electrocardiograms. Figure 2(a) shows a 12-lead electrocardiogram (ECG) recording with optional supplemental electrophoresis recordings distributed around the torso, abdomen, neck, arms, and lower extremities. The resulting electrophoresis is shown in Figure 2(b). Figure 13 shows an example of a vector electrocardiogram (VCG) plot, which shows the magnitude and direction of the electrical forces generated by the patient's heart during depolarization and repolarization as a series of consecutive vectors. Based at least on the patient's depolarization (e.g., QRS complex waves observed in the patient's electrical recording), the evaluation engine 110 can identify one or more cardiac depolarization simulations (e.g., ventricular excitation simulations) that match the patient's electrical recording. Alternatively and / or additionally, the evaluation engine 110 may identify one or more cardiac repolarization simulations (e.g., ventricular recovery simulations) that match the patient's electrical recording, based at least on the patient's repolarization (e.g., T waves observed in the patient's electrical recording). In some cases, the patient's electrical recording may undergo one or more signal processing techniques before being compared with simulations in the computational library. Examples of applicable signal processing techniques include recording, filtering, digitization, transformation, and spatial analysis.

[0067] Figure 11 shows an example of a 30-lead electrocardiogram (ECG) setup for capturing an electrical record of a patient, according to several exemplary embodiments. As shown in Figure 11, a 30-lead ECG setup may include four portable harnesses having 24 unipolar anterior chest leads and 6 unipolar limb leads. Referring to Figure 11(a), a 30-lead ECG setup may include a first anterior row (electrodes) positioned across the manubrium angle of the sternum, a second anterior row positioned across the xiphoid process, and a third anterior row positioned equidistant between the first and second anterior rows. Furthermore, Figure 11(b) shows that a 30-lead ECG setup may include a first posterior row positioned across the scapulospinal process, a second posterior row positioned across the inferior border of the scapula, and a third posterior row positioned equidistant between the first and second posterior rows.

[0068] Figure 12 shows an example of data traces captured by a 30-lead electrocardiogram (ECG) setup, for example, as shown in Figure 11. For example, Figure 12 shows multiple data traces for the same lead (e.g., lead 3), but at different sampling rates (e.g., 500 Hz and 1000 Hz). In some exemplary embodiments, the evaluation engine 110 can process the data from each lead individually. In this way, the evaluation engine 110 can generate corresponding graphs and identify one or more peaks, onsets, and offsets present in the data. Furthermore, the evaluation engine 110 can extract one or more peaks, onsets, and / or offsets to identify measurements such as QT interval, QT variance, and / or similar.

[0069] In some exemplary embodiments, before matching a patient's electrocardiogram with one or more simulations in the computational library 125, the evaluation engine 110 can first identify a subset of simulations corresponding to the patient's anatomical structure, so that subsequent matching is performed within this subset of simulations rather than the entire computational library 125. For example, the evaluation engine 110 can identify a subset of simulations corresponding to a patient's anatomical structure based on patient-specific clinical data such as patient demographics and cardiac imaging. Examples of relevant imaging modalities include cardiac computed tomography, cardiac magnetic resonance imaging (MRI), sestamibi imaging (nuclear scintigraphy), cardiac positron emission tomography and computed tomography scanning (cardiac PET / CT), two-dimensional and three-dimensional echocardiography, three-dimensional electroanatomical mapping incorporating voltage mapping, electromorphological mapping, excitation mapping, entrainment mapping, isochronous late activation mapping, and impedance mapping.

[0070] Cardiac imaging data can be used to identify the locations of scar tissue, borderline tissue, and normal tissue within a patient. For example, Figure 5 shows a 3D volume rendering from 4D computed tomography (CT) with a region corresponding to borderline tissue (e.g., between scar tissue and normal tissue) indicated by a dotted line. The region inside the circle indicates a relative lack of motion, which corresponds to dense scar tissue. Thus, the evaluation engine 110 can use cardiac imaging data in combination with patient demographics to eliminate simulations that are related anatomical structures that do not sufficiently resemble the patient's anatomical structure, and therefore simulations that are unlikely to match the patient's electrical recording. In this way, the computational speed and efficiency of subsequent matching to identify simulations that match the patient's electrical recording can be improved.

[0071] In some cases where the patient's anatomical structure is not adequately represented in the computational library 125, the evaluation engine 110 can generate a custom computational library for the patient, including one or more patient-specific cardiac repolarization simulations (e.g., ventricular excitation simulations) and cardiac depolarization simulations (e.g., ventricular recovery simulations). For example, a custom computational library can be generated if the clinical data associated with the simulations included in the computational library 125 (e.g., demographics, cardiac imaging, and / or similar) is not sufficiently similar to the clinical data relevant to the patient. Thus, a custom computational library for a patient can be generated based on patient-specific clinical data such as patient demographics and cardiac imaging.

[0072] In some exemplary embodiments, a variety of techniques can be applied to match one or more cardiac depolarization simulations (e.g., ventricular excitation simulations) and cardiac repolarization simulations (e.g., ventricular recovery simulations) with the patient's electrical recordings. For example, the evaluation engine 110 can apply one or more machine learning models trained to identify one or more cardiac depolarization simulations (e.g., ventricular excitation simulations) and cardiac repolarization simulations (e.g., ventricular recovery simulations) that match the patient's electrical recordings. Examples of suitable machine learning models include neural networks, regression models, instance-based models, regularization models, decision trees, random forests, Bayesian models, clustering models, associative models, dimensionality reduction models, ensemble models, and / or similar.

[0073] In this way, the evaluation engine 110 can compare spatial data and / or temporal data. For example, a 12-lead electrocardiogram (ECG) recording may include time-series data in which voltages measured by electrodes are recorded at regular time intervals (e.g., every millisecond; 1000 Hz). To identify simulations that match the patient's 12-lead ECG recording, the evaluation engine 110 can apply a recurrent neural network (e.g., a long-term short-term memory (LSTM) network and / or similar) that can recognize patterns present across two or more measurement sequences. The recurrent neural network (RNN) can be trained, for example, to detect QRS complexes in one or more cardiac depolarization simulations (e.g., ventricular excitation simulations) that match QRS complexes present in the patient's 12-lead ECG recording. Alternatively and / or additionally, a recurrent neural network can be trained to detect T waves in one or more cardiac repolarization simulations (e.g., ventricular recovery simulations) that match T waves present in a patient's 12-lead electrocardiogram (ECG) recording.

[0074] Once the evaluation engine 110 identifies (or generates) cardiac depolarization simulations (e.g., ventricular excitation simulations) and cardiac repolarization simulations (e.g., ventricular recovery simulations) that match the patient's electrical recordings, it can apply the simulations to generate a three-dimensional evaluation for the patient. For example, based at least on cardiac depolarization simulations (e.g., ventricular excitation simulations) and cardiac repolarization simulations (e.g., ventricular recovery simulations) that match the patient's electrical recordings, the evaluation engine 110 can identify the size and / or location of areas of increased spatial repolarization gradient, slow conduction, conduction block, or premature excitation. As shown in Figure 14, the cardiac repolarization phase in a patient with ventricular arrhythmias may exhibit one or more abnormalities not present in the cardiac repolarization of a control patient. For example, the presence of adjacent myocardial regions where the respective cardiac repolarization rates differ by a threshold (e.g., a first region of long repolarization directly adjacent to a second region of short repolarization) may be a potential arrhythmogenic state. It is also known that sites of slow conduction can perpetuate certain types of arrhythmias. Furthermore, the site of early excitation may be the location of arrhythmia-inducing tissue. Therefore, in some exemplary embodiments, the evaluation engine 110 can determine an estimate of the patient's risk for cardiac arrhythmias (e.g., ventricular arrhythmias) based at least on the magnitude of the increased spatial repolarization gradient.

[0075] Furthermore, the evaluation engine 110 can determine a treatment plan for the patient based at least on the magnitude and / or location of the increased spatial repolarization gradient. For example, the evaluation engine 110 can determine, at least on the magnitude of the increased spatial repolarization gradient, whether the patient could benefit from an implantable cardioverter-defibrillator or from invasive electrophysiological examinations and ablations. Figure 7 shows a biventricular implantable cardioverter-defibrillator (ICD) 700 placed in a patient with cardiomyopathy and a left ventricular ejection fraction of less than 35%. Alternatively and / or additionally, the location of the increased spatial repolarization gradient, slow conduction region, protected conduction isthmus, or early excitation site can be used to identify locations for targeted therapies, including, for example, radiofrequency catheter ablation, cryoablation, stereotactic radiotherapy (SAbR) (or stereotactic body radiotherapy (SBRT)) and / or similar. Figure 8(a) shows an example of an electroanatomical map of ventricular tachycardia ablation with a circular marker indicating the ablation site located in the middle-lower left ventricle. Figure 8(b) shows a patient with refractory ventricular arrhythmia undergoing stereotactic radiotherapy (SAbR) targeting the anterior mitral annulus of the left ventricle.

[0076] Figure 6 shows an example of a process 600 for calculating the size and location of an increased spatial repolarization gradient region using non-invasive 12-lead electrocardiogram (ECG) data, with or without body surface potential recordings, according to several exemplary embodiments. As shown in Figure 6(A), the 12-lead ECG recording can be performed on the patient with or without supplemental body surface potential recordings (e.g., around the patient's torso, abdomen, neck, arms, and / or lower extremities). In Figures 6(B) and 6(C), the resulting electrographs can be subjected to one or more signal processing techniques, including, for example, recording, filtering, digitization, transformation, spatial analysis, and / or similar. In Figure 6(D), a subset of simulations corresponding to the patient's anatomical structure can be identified from the computational library 125, so that subsequent matching is performed within this subset of simulations rather than the entire computational library 125. In Figure 6(E), cardiac depolarization simulations (e.g., ventricular excitation simulations) and cardiac repolarization simulations (e.g., ventricular recovery simulations) can be identified that match the patient's 12-lead electrocardiogram (ECG) recording. In Figure 6(F), the magnitude and / or location of the amplified spatial repolarization gradient can be calculated based on the cardiac depolarization simulations (e.g., ventricular excitation simulations) and cardiac repolarization simulations (e.g., ventricular recovery simulations) that match the patient's 12-lead electrocardiogram (ECG) recording. As previously mentioned, the presence of adjacent myocardial regions where the respective cardiac repolarization rates differ by a threshold (e.g., a first region of long repolarization directly adjacent to a second region of short repolarization) may indicate a potential arrhythmogenic state. Therefore, the patient's risk of cardiac arrhythmias (e.g., ventricular arrhythmias) can be identified, at least based on the magnitude of the amplified spatial repolarization gradient. Furthermore, one or more appropriate treatments for the patient can be identified, at least based on the magnitude and / or location of the amplified spatial repolarization gradient.

[0077] In some cases, the evaluation engine 110 can further identify the presence and / or absence of one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block, based at least on one or more cardiac depolarization and cardiac repolarization simulations corresponding to the patient's electrical recording. The treatment plan for the patient can further take these regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, and / or conduction block into consideration. For example, the evaluation engine 110 can generate a treatment plan that targets one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, and / or conduction block individually or in groups. Alternatively and / or additionally, the evaluation engine 110 can determine whether the patient is a candidate for one or more drug therapies, based at least on the presence and / or absence of one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, and / or conduction block.

[0078] Figure 9 shows a flowchart illustrating an example of a process 900 for non-invasive cardiac arrhythmia risk stratification by several exemplary embodiments. Referring to Figures 1 and 9, the process 900 can be performed by the assessment engine 110 to identify a patient's risk for cardiac arrhythmias.

[0079] In 902, the evaluation engine 110 can generate one or more computational libraries of cardiac depolarization simulations and / or cardiac repolarization simulations. For example, in some exemplary embodiments, the evaluation engine 110 can generate computational libraries 125 that include cardiac depolarization simulations (e.g., ventricular excitation simulations) for various cardiac geometries, cardiac orientations, scar structures, degrees of cardiac fibrosis and scarring, depolarization patterns, types of excitation (e.g., left bundle branch block, right bundle branch block, left anterior fascicular block, left posterior fascicular block, ventricular premature contractions, ventricular tachycardia, and ventricular fibrillation), and combinations thereof. Alternatively and / or additionally, the evaluation engine 110 can generate a computational library 125 that includes cardiac repolarization simulations (e.g., ventricular recovery simulations) for various cardiac geometries, cardiac orientations, scar structures, degrees of cardiac fibrosis and scarring, depolarization patterns, types of excitation (e.g., left bundle branch block, right bundle branch block, left anterior fascicular block, left posterior fascicular block, ventricular premature contractions, ventricular tachycardia, and ventricular fibrillation), and combinations thereof. In some cases, the computational library 125 may include a first computational library for cardiac depolarization simulations (e.g., ventricular excitation simulations) and a second computational library for cardiac repolarization simulations (e.g., ventricular recovery simulations).

[0080] In 904, the evaluation engine 110 can acquire patient-related cardiac imaging data. For example, the evaluation engine 110 can acquire cardiac imaging data in a variety of imaging modalities, including, for example, cardiac computed tomography (e.g., 3D volume rendering from 4D computed tomography shown in Figure 5), cardiac magnetic resonance imaging (MRI), sestamibi imaging (nuclear scintigraphy), cardiac positron emission tomography and computed tomography scanning (cardiac PET / CT), 2D and 3D echocardiography, 3D electroanatomical mapping incorporating voltage mapping, electromorphological mapping, excitation mapping, entrainment mapping, isochronous late activation mapping, and impedance mapping. In some exemplary embodiments, patient-related cardiac imaging data can enable the evaluation engine 110 to identify the location of scar tissue, borderline tissue, and normal tissue within the patient.

[0081] In 906, the evaluation engine 110 can acquire electrical records related to the patient. In some exemplary embodiments, the evaluation engine 110 can acquire a variety of electrical records about the patient, including, for example, electropotential diagrams, vector diagrams, electrocardiograms, electroencephalograms, vector electrocardiograms, and / or similar. Figure 2 shows a 12-lead electrocardiogram (ECG) recording, which is an example of an electrical record that can be acquired by the evaluation engine 110. As shown in Figure 2, a 12-lead electrocardiogram (ECG) recording can be performed with (or without) supplemental surface potential recordings distributed around the patient's torso, abdomen, neck, arms, and / or lower extremities, for example.

[0082] In 908, the evaluation engine 110 may apply one or more signal processing techniques to the electrical record. In some exemplary embodiments, the evaluation engine 110 may apply a variety of signal processing techniques to the patient's electrical record (e.g., a 12-lead electrocardiogram (ECG) record), including, for example, recording, filtering, digitization, conversion, spatial analysis, and / or similar. Processing of the electrical record may be optional, and matching of the electrical record may be performed without any signal processing.

[0083] In 910, the evaluation engine 110 can identify a subset of simulations corresponding to the patient's anatomical structure from one or more computational libraries, based at least on clinical data including patient demographics and patient cardiac imaging data. In some exemplary embodiments, the evaluation engine 110 can first identify a subset of simulations corresponding to the patient's anatomical structure before matching the patient's electrocardiogram with one or more simulations in the computational library 125. Subsequent matching to identify the simulation to match can then be performed within this subset of simulations rather than the entire computational library 125. The evaluation engine 110 can identify a subset of simulations corresponding to the patient's anatomical structure based on patient-specific clinical data such as patient demographics and cardiac imaging. For example, based at least on patient demographics and cardiac imaging data, the evaluation engine 110 can eliminate simulations that are related anatomical structures not sufficiently similar to the patient's anatomical structure, and therefore simulations that are unlikely to match the patient's electrical recording. As previously described, the evaluation engine 110 can first identify a subset of simulations in order to improve the computational speed and efficiency of subsequent matching to identify simulations that match the patient's electrical records. However, it should be understood that selecting a subset of simulations corresponding to the patient's anatomical structure is an optional optimization, and instead, subsequent matching may be performed on the entire computational library 125 without first identifying a subset of simulations corresponding to the patient's anatomical structure.

[0084] If the evaluation engine 110 is unable to identify a subset of simulations corresponding to the patient's anatomical structure, in 912, the evaluation engine 110 may generate a custom simulation library for the patient. In some cases, the patient's anatomical structure may not be adequately represented in the computational library 125. In such scenarios, the evaluation engine 110 may generate a custom computational library for the patient, including one or more patient-specific cardiac repolarization simulations (e.g., ventricular excitation simulations) and cardiac depolarization simulations (e.g., ventricular recovery simulations). For example, a custom computational library may be generated if the clinical data associated with the simulations included in the computational library 125 (e.g., demographics, cardiac imaging, and / or similar) is not sufficiently similar to the clinical data relevant to the patient. Thus, the evaluation engine 110 may generate a custom computational library for the patient based on patient-specific clinical data such as patient demographics and cardiac imaging.

[0085] In 914, the evaluation engine 110 can identify one or more cardiac depolarization and cardiac repolarization simulations corresponding to the patient's electrical recording. For example, the evaluation engine 110 can identify one or more cardiac depolarization simulations (e.g., ventricular excitation simulations) that match the patient's electrical recording, based at least on the patient's depolarization (e.g., QRS complex waves observed in the patient's electrical recording). Alternatively and / or additionally, the evaluation engine 110 can identify one or more cardiac repolarization simulations (e.g., ventricular recovery simulations) that match the patient's electrical recording, based at least on the patient's repolarization (e.g., T waves observed in the patient's electrical recording). This matching can be performed by comparing the patient's electrical recording with at least a subset of simulations included in the computational library 125, or optionally by comparing it with a subset of simulations determined to correspond to the patient's anatomical structure. The evaluation engine 110 can apply a variety of techniques to identify one or more cardiac depolarization simulations (e.g., ventricular excitation simulations) and cardiac repolarization simulations (e.g., ventricular recovery simulations) that match the patient's electrical recordings. For example, the evaluation engine 110 can apply one or more machine learning models trained to identify one or more cardiac depolarization simulations (e.g., ventricular excitation simulations) and cardiac repolarization simulations (e.g., ventricular recovery simulations) that match the patient's electrical recordings.

[0086] In 918, the evaluation engine 110 can determine the magnitude and / or location of an increased spatial repolarization gradient for the patient, based on at least one or more cardiac depolarization and cardiac repolarization simulations corresponding to the patient's electrical recording. A spatial repolarization gradient may exist if adjacent myocardial regions show a difference exceeding a threshold with respect to their respective cardiac repolarization rates. Thus, the evaluation engine 110 can identify a first myocardial region with long repolarization that is directly adjacent to a second myocardial region with short repolarization, based on at least one or more cardiac depolarization and cardiac repolarization simulations corresponding to the patient's electrical recording. Furthermore, the evaluation engine 110 can determine the difference between the first repolarization rate of the first region and the second repolarization rate of the second region, based on at least one or more cardiac depolarization and cardiac repolarization simulations corresponding to the patient's electrical recording.

[0087] In 920, the evaluation engine 110 can determine the patient's risk of cardiac arrhythmias and / or a treatment plan for the patient, at least based on the magnitude and / or location of the increased spatial repolarization gradient. For example, the evaluation engine 110 can identify potential arrhythmic-promoting conditions that are likely to result in excitation wave splitting, re-entry, and life-threatening arrhythmias, at least based on adjacent myocardial regions where the respective cardiac repolarization rates differ by a threshold (e.g., a first region of long repolarization directly adjacent to a second region of short repolarization). Thus, the evaluation engine 110 can identify the patient's risk to cardiac arrhythmias (e.g., ventricular arrhythmias) at least based on the magnitude of the increased spatial repolarization gradient. In some cases, the evaluation engine 110 can further identify one or more appropriate treatments for the patient based on the magnitude and / or location of the increased spatial repolarization gradient. For example, an increased spatial repolarization gradient, the extent of slow conduction, or the magnitude of early excitation can enable the evaluation engine 110 to determine whether the patient would benefit from defibrillator implantation or from invasive electrophysiological examinations and ablations. Alternatively and / or additionally, the location of the increased spatial repolarization gradient can be used to identify the location of targeted therapies, including, for example, catheter ablation, stereotactic radiotherapy (SAbR) (or stereotactic body radiotherapy (SBRT)), and / or similar procedures.

[0088] Figure 11 shows a block diagram illustrating a computing system 1100 in several exemplary embodiments. Referring to Figures 1 and 11, the computing system 1100 can be used to implement the evaluation engine 110 and / or any components within the evaluation engine 110.

[0089] As shown in Figure 11, the computing system 1100 may include a processor 1110, memory 1120, a storage device 1130, and an input / output device 1140. The processor 1110, memory 1120, storage device 1130, and input / output device 1140 can be interconnected via a system bus 1150. The processor 1110 can process instructions to be executed within the computing system 1100. Such instructions to be executed may, for example, implement one or more components of the evaluation engine 110. In some implementations of this subject, the processor 1110 may be a single-threaded processor. Alternatively, the processor 1110 may be a multi-threaded processor. The processor 1110 can process instructions stored in memory 1120 and / or on the storage device 1130 to display graphic information for a user interface provided via the input / output device 1140.

[0090] Memory 1120 is a computer-readable medium, such as volatile or non-volatile, that stores information within the computing system 1100. Memory 1120 can store, for example, data structures representing a configuration object database. Storage device 1130 can provide persistent storage for the computing system 1100. Storage device 1130 may be a floppy disk device, a hard disk device, an optical disk device, or a tape device, or other suitable persistent storage means. Input / output device 1140 provides input / output operations for the computing system 1100. In some implementations of this subject, input / output device 1140 includes a keyboard and / or a pointing device. In various other implementations, input / output device 1140 includes a display unit for displaying a graphical user interface.

[0091] In some implementations of this subject, the input / output device 1140 can provide input / output operations for network devices. For example, the input / output device 1140 may include an Ethernet port or other network port for communicating with one or more wired and / or wireless networks (e.g., a local area network (LAN), a wide area network (WAN), or the Internet).

[0092] In some implementations of this subject matter, the computing system 1100 can be used to run various interactive computer software applications, which can be used to organize, analyze, and / or store data in various formats (e.g., tabular format). Alternatively, the computing system 1100 can be used to run any type of software application. These applications can be used to perform various functions, such as planning functions (e.g., generating, managing, and editing spreadsheet documents, word processing documents, and / or any other objects), calculation functions, communication functions, and / or similar functions. These applications may include various add-in functions or may be standalone computing products and / or functions. These functions, when activated within an application, can be used to generate a user interface provided via the input / output device 1140. The user interface can be generated by the computing system 1100 and presented to the user (e.g., on a computer screen monitor).

[0093] One or more embodiments or features of the subject matter described herein can be realized in the form of digital electronic circuits, integrated circuits, specially designed application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments or features may include implementations in the form of one or more computer programs executable and / or interpretable in a programmable system comprising at least one programmable processor, which may be dedicated or general-purpose, coupled to receive data and instructions from a storage system, at least one input device, and at least one output device, and to transmit data and instructions to the storage system, at least one input device, and at least one output device. The programmable system or computing system may include clients and servers. Clients and servers are generally located remotely from each other and typically interact via a communication network. The client-server relationship is created by computer programs running on individual computers and having a client-server relationship with each other.

[0094] These computer programs, which may also be referred to as programs, software, software applications, applications, components, or code, contain machine instructions for a programmable processor and can be implemented in high-level procedural programming languages ​​and / or object-oriented programming languages ​​and / or assembly language / machine language. As used herein, the term “machine-readable medium” refers to any computer program product, apparatus, and / or device, such as magnetic disks, optical disks, memory, and programmable logic devices (PLDs), that are used to provide machine instructions and / or data to a programmable processor, which contains a machine-readable medium that receives machine instructions as machine-readable signals. The term “machine-readable signals” refers to any signals used to provide machine instructions and / or data to a programmable processor. A machine-readable medium can store such machine instructions non-temporarily, as does, for example, non-temporarily solid-state memory or a magnetic hard drive or any equivalent storage medium. Alternatively or additionally, a machine-readable medium can store such machine instructions temporarily, as does, for example, a processor cache or other random-access memory associated with one or more physical processor cores.

[0095] The subject matter described herein can be embodied in the form of systems, apparatus, methods, and / or articles, depending on the desired configuration. The implementations described above do not represent all implementations that correspond to the subject matter described herein. Rather, the implementations described above are merely examples that correspond to aspects related to the subject matter described herein. While several variations have been described in detail above, other modifications or additions are possible. In particular, further features and / or variations can be provided in addition to those described herein. For example, the implementations described above may cover various combinations and components of the disclosed features, and / or combinations and components of several further features disclosed above. Furthermore, the logical flows illustrated in the accompanying drawings and / or described herein do not necessarily require a specific order or sequence presented to achieve the desired result. Other implementations may also be included in the following claims.

Claims

1. A system, wherein the system is At least one processor, At least one memory location containing program code and Includes, The program code, when executed by the at least one processor, provides the following operations, Within the computational library, identify cardiac depolarization and cardiac repolarization simulations corresponding to the patient's electrical records, Identifying one or more regions exhibiting an increased spatial repolarization gradient, based at least on the cardiac repolarization simulation, or the cardiac depolarization simulation and the cardiac repolarization simulation, wherein in such regions, the ratio of the difference between the first repolarization rate of a first area of ​​the patient's myocardium and the second repolarization rate of a second area of ​​the myocardium to the spatial distance between the first area and the second area exceeds a threshold. Based at least on the magnitude of the increased spatial repolarization gradient, the risk of cardiac arrhythmia in the patient is identified. A system that includes this.

2. The system according to claim 1, wherein the operation further comprises determining a treatment plan for the patient based at least on the magnitude of the increased spatial repolarization gradient.

3. The system according to claim 2, wherein the treatment plan is determined to include implantation of a defibrillator or to include invasive electrophysiological examination and ablation, at least based on the magnitude of the increased spatial repolarization gradient.

4. The system according to claim 2 or 3, wherein the treatment plan includes identifying a location for targeted therapy based on the location of at least one of the regions of the increased spatial repolarization gradient.

5. The system according to claim 4, wherein the targeted therapy includes catheter ablation and / or stereotactic radiotherapy (SAbR).

6. The aforementioned cardiac depolarization simulation includes a ventricular excitation simulation. The aforementioned cardiac repolarization simulation includes a ventricular recovery simulation. The system according to any one of claims 1 to 5.

7. The system according to any one of claims 1 to 6, wherein the computational library is supplemented by clinical patient samples in which the location of the source of an arrhythmic substrate is known, for providing additional data for comparison with the patient's electrical recording.

8. The system according to any one of claims 1 to 7, wherein the calculation library includes a clinical sample in which the location of the source of an arrhythmic substrate is known, to serve as reference data for comparison with the electrical record of the patient.

9. The aforementioned operation is To generate the computational library to include multiple cardiac depolarization simulations and multiple cardiac repolarization simulations, wherein the multiple cardiac depolarization simulations and the multiple cardiac repolarization simulations correspond to various cardiac geometries, cardiac orientations, scar structures, degrees of cardiac fibrosis and scarring, depolarization patterns, and / or types of excitation, Within the calculation library, identify the cardiac depolarization simulation and the cardiac repolarization simulation corresponding to the patient's electrical record. The system according to any one of claims 1 to 8, further comprising:

10. The aforementioned operation is Based at least on clinical data related to the patient, identify a subset of simulations from the computational library that corresponds to the patient's anatomical structure, Within the subset of simulations corresponding to the anatomical structure of the patient, identify the cardiac depolarization simulation and the cardiac repolarization simulation corresponding to the electrical recording of the patient. The system according to claim 9, further comprising:

11. The system according to claim 10, wherein the clinical data includes patient demographics.

12. The aforementioned clinical data includes cardiac imaging data, The cardiac imaging data indicates the location of one or more scar tissues, border tissues, and normal tissues, cardiac chamber sizes, the presence of hypertrophy or enlargement, the location of fibrosis, areas of normal and abnormal contractility, or areas of wall thinning. The system according to claim 10 or 11.

13. The system according to any one of claims 10 to 12, wherein the operation further comprises generating a custom computational library, based at least on the patient's clinical data, which includes one or more cardiac depolarization and / or cardiac repolarization simulations specific to the patient's anatomical structure, in response to the inability to identify a subset of simulations corresponding to the patient's anatomical structure.

14. The system according to any one of claims 1 to 13, wherein the operation further comprises applying a machine learning model trained to determine whether the cardiac repolarization simulation and the cardiac depolarization simulation match the patient's electrical records.

15. The system according to claim 14, wherein the machine learning model includes a neural network, a regression model, an instance-based model, a regularization model, a decision tree, a random forest, a Bayesian model, a clustering model, an associative model, a dimensionality reduction model, and / or an ensemble model.

16. The system according to any one of claims 1 to 15, wherein the operation further comprises applying one or more signal processing techniques to the electrical record of the patient.

17. The system according to claim 16, wherein the one or more signal processing techniques include recording, filtering, digitization, conversion, and / or spatial analysis.

18. The system according to any one of claims 1 to 17, wherein the electrical record includes one or more of a potential diagram, a vector diagram, an electrocardiogram, an electroencephalogram, or a vector electrocardiogram.

19. The system according to claim 18, wherein the electrical recording further comprises one or more body surface potential recordings.

20. The system according to any one of claims 1 to 19, wherein the electrical recording includes an electrocardiogram imaging (ECGi) recording system that includes one or more surface potential recordings.

21. The aforementioned operation is Based at least on the cardiac depolarization simulation and the cardiac repolarization simulation, identify one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block, Identifying the risk of cardiac arrhythmia in the patient based at least on the presence and / or absence of one or more of the following regions: early excitation, slow conduction, independent excitation pathway, delayed excitation, protected conduction isthmus, and / or conduction block. The system according to any one of claims 1 to 20, further comprising:

22. The system according to claim 21, wherein the operation further comprises determining a treatment plan for the patient based on at least one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block.

23. The system according to claim 22, wherein the treatment plan targets one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block individually or in groups.

24. The system according to claim 22 or 23, wherein the treatment plan is determined to include one or more drug therapies based at least on the presence and / or absence of one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block.

25. A computer implementation method that is executed by a processor, Within the computational library, identify cardiac depolarization and cardiac repolarization simulations corresponding to the patient's electrical records, Identifying one or more regions exhibiting an increased spatial repolarization gradient, based at least on the cardiac repolarization simulation, or the cardiac depolarization simulation and the cardiac repolarization simulation, wherein in such regions, the ratio of the difference between the first repolarization rate of a first area of ​​the patient's myocardium and the second repolarization rate of a second area of ​​the myocardium to the spatial distance between the first area and the second area exceeds a threshold. Based at least on the magnitude of the increased spatial repolarization gradient, the risk of cardiac arrhythmia in the patient is identified. Computer implementation methods, including those mentioned above.

26. The method according to claim 25, further comprising determining a treatment plan for the patient based at least on the magnitude of the increased spatial repolarization gradient.

27. The method according to claim 26, wherein the treatment plan is determined to include implantation of a defibrillator or to include invasive electrophysiological examination and ablation, at least on the magnitude of the increased spatial repolarization gradient.

28. The method according to claim 26 or 27, wherein the treatment plan includes identifying a location for targeted therapy based on the location of at least one of the regions of the increased spatial repolarization gradient.

29. The method according to claim 28, wherein the targeted therapy includes catheter ablation and / or stereotactic radiotherapy (SAbR).

30. The aforementioned cardiac depolarization simulation includes a ventricular excitation simulation. The aforementioned cardiac repolarization simulation includes a ventricular recovery simulation. The method according to any one of claims 25 to 29.

31. The aforementioned method, To generate the computational library to include multiple cardiac depolarization simulations and multiple cardiac repolarization simulations, wherein the multiple cardiac depolarization simulations and the multiple cardiac repolarization simulations correspond to various cardiac geometries, cardiac orientations, scar structures, degrees of cardiac fibrosis and scarring, depolarization patterns, and / or types of excitation, Within the calculation library, identify the cardiac depolarization simulation and the cardiac repolarization simulation corresponding to the patient's electrical record. The method according to any one of claims 25 to 30, further comprising:

32. The aforementioned method, Based at least on clinical data related to the patient, identify a subset of simulations from the computational library that corresponds to the patient's anatomical structure, Within the subset of simulations corresponding to the anatomical structure of the patient, identify the cardiac depolarization simulation and the cardiac repolarization simulation corresponding to the electrical recording of the patient. The method according to claim 31, further comprising:

33. The method according to claim 32, wherein the clinical data includes patient demographics.

34. The aforementioned clinical data includes cardiac imaging data, The cardiac imaging data indicates the location of one or more scar tissues, border tissues, and normal tissues, cardiac chamber sizes, the presence of hypertrophy or enlargement, the location of fibrosis, areas of normal and abnormal contractility, and / or areas of wall thinning. The method according to claim 32 or 33.

35. The method according to any one of claims 32 to 34, further comprising generating a custom computational library, based at least on the patient's clinical data, that includes one or more cardiac depolarization and / or cardiac repolarization simulations specific to the patient's anatomical structure, in response to the inability to identify a subset of simulations corresponding to the patient's anatomical structure.

36. The method according to any one of claims 25 to 35, further comprising applying a machine learning model trained to determine whether the cardiac repolarization simulation and the cardiac depolarization simulation match the patient's electrical records.

37. The method according to claim 36, wherein the machine learning model includes a neural network, a regression model, an instance-based model, a regularization model, a decision tree, a random forest, a Bayesian model, a clustering model, an associative model, a dimensionality reduction model, and / or an ensemble model.

38. The method according to any one of claims 25 to 37, further comprising applying one or more signal processing techniques to the electrical record of the patient.

39. The method according to claim 38, wherein the one or more signal processing techniques include recording, filtering, digitization, conversion, and / or spatial analysis.

40. The method according to any one of claims 25 to 39, wherein the electrical record includes one or more of a potential diagram, a vector diagram, an electrocardiogram, an electroencephalogram, or a vector electrocardiogram.

41. The method according to claim 40, wherein the electrical recording further comprises one or more body surface potential recordings.

42. The method according to any one of claims 25 to 41, wherein the electrical recording includes electrocardiogram imaging (ECGi) which includes one or more surface potential recordings.

43. The aforementioned method, Based at least on the cardiac depolarization simulation and the cardiac repolarization simulation, identify one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block, Identifying the risk of cardiac arrhythmia in the patient based at least on the presence and / or absence of one or more of the following regions: early excitation, slow conduction, independent excitation pathway, delayed excitation, protected conduction isthmus, and / or conduction block. The method according to any one of claims 25 to 42, further comprising:

44. The method according to claim 43, further comprising determining a treatment plan for the patient based on at least one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block.

45. The method according to claim 44, wherein the treatment plan targets one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block individually or in groups.

46. The method according to claim 44 or 45, wherein the treatment plan is determined to include one or more drug therapies based at least on the presence and / or absence of one or more regions of early excitation, slow conduction, independent excitation pathways, delayed excitation, protected conduction isthmus, and / or conduction block.

47. A non-temporary computer-readable medium, the non-temporary computer-readable medium storing instructions that, when executed by at least one data processor, result in the following operations, the operations are: Within the computational library, identify cardiac depolarization and cardiac repolarization simulations corresponding to the patient's electrical records, Identifying one or more regions exhibiting an increased spatial repolarization gradient, based at least on the cardiac repolarization simulation, or the cardiac depolarization simulation and the cardiac repolarization simulation, wherein in such regions, the ratio of the difference between the first repolarization rate of a first area of ​​the patient's myocardium and the second repolarization rate of a second area of ​​the myocardium to the spatial distance between the first area and the second area exceeds a threshold. Based at least on the magnitude of the increased spatial repolarization gradient, the risk of cardiac arrhythmia in the patient is identified. Non-temporary computer-readable media, including [specific examples of such media].

48. It is a device, Within the computational library, means for identifying cardiac depolarization simulations and cardiac repolarization simulations corresponding to the patient's electrical records, Means for identifying one or more regions exhibiting an increased spatial repolarization gradient, based at least on the cardiac repolarization simulation, or the cardiac depolarization simulation and the cardiac repolarization simulation, wherein in such regions, the ratio of the difference between the first repolarization rate of a first area of ​​the patient's myocardium and the second repolarization rate of a second area of ​​the myocardium to the spatial distance between the first area and the second area exceeds a threshold. Means for identifying the risk of cardiac arrhythmia in the patient, based at least on the magnitude of the increased spatial repolarization gradient, A device including a device.

49. The apparatus according to claim 48, further comprising means for performing the operation of the method described in any one of claims 26 to 46.

Images