BACKGROUND:Substrate mapping and ablation for scar-related ventricular tachycardia (VT) are often performed with concomitant amiodarone (AMD), which can modify VT inducibility and alter electrogram (EGM) signatures of arrhythmogenic substrate. OBJECTIVE:This study aimed to evaluate how AMD influences VT inducibility and bipolar EGM markers using heart digital twins. METHODS:The effects of AMD at 2 therapeutic concentrations were characterized using heart digital twins incorporating patient-specific ischemic substrate distributions. VT inducibility was tested from multiple sites, and the reentrant circuits were characterized. Identified features of virtual electroanatomic maps included low-voltage, prolonged-duration, high-fractionation EGMs and isochronal crowding (IC) zones. Composite multiple-wavefront pacing maps were used to reduce electroanatomic map dependence on wavefront direction. RESULTS:AMD did not significantly change overall VT inducibility, although a nonsignificant trend toward fewer inducible sites emerged at higher doses. At the VT circuit level, several baseline circuits disappeared, new drug-associated pathways emerged, and persistent morphologies tended to exhibit longer cycle lengths. AMD broadened regions classified as abnormal by voltage, duration, and fractionation, expanding the apparent substrate and reducing mapping specificity. With a single pacing site, IC zones were within 9 mm of the critical isthmus, more targeted than abnormal signal features, and not appreciably expanded by AMD. Combining multiple pacing orientations markedly improved colocalization without loss of specificity. CONCLUSION:Concomitant AMD alters VT morphology and expands abnormal regions identified during substrate mapping, potentially obscuring critical ablation targets. Functional mapping strategies emphasizing IC and incorporating multiple-wavefront directions may improve long-term procedural success of scar-related VT ablation, particularly when AMD cannot be withdrawn.

Although pulmonary vein (PV) isolation (PVI) is an established atrial fibrillation (AF) treatment, its efficacy in persistent AF (PsAF) remains limited. Accordingly, post-PVI arrhythmia inducibility testing—mimicking non-PV ectopic activation—has often been employed during PsAF ablation, the value of non-inducibility as procedural endpoints or prognostic markers remains uncertain. Personalized atrial digital twins (DTs) can assess patient-specific arrhythmia inducibility; however, their clinical translation is hindered by the need for expert fibrosis characterization from MRI signals and substantial computational resources. To overcome these barriers, we leveraged deep learning (DL) to develop a practical framework for identifying potential non-PV trigger (NPVT) sites capable of inducing arrhythmia, without requiring patient-specific fibrosis thresholding or simulation. A comprehensive pacing-site set yielded significantly more patients’ DTs harboring potential NPVT sites than conventional catheter-based sites (85% vs. 54%). The DL model, InduceNet, trained on DT-derived results, accurately predicted potential NPVT sites (sensitivity 91%). This integrated DT–DL framework facilitates direct access, in clinical practice, to DT-derived insights into patient-specific potential NPVT sites, supporting more effective decision-making.

AIMS:Persistent atrial fibrillation (PsAF) is often refractory to pulmonary vein isolation, a well-established AF treatment, owing to the fibrosis-remodeled substrates sustaining reentries. Three-dimensional fibrosis distribution and atrial adiposity have been found to be essential contributors to PsAF arrhythmogenesis. We aimed to utilize late gadolinium enhancement (LGE)-MRI-derived personalized heart digital twins (DTs)-a recent promising tool for non-invasively assessing patient arrhythmogenesis-as well as contrast-enhanced cardiac computed tomography (CCT) images and electroanatomic maps (EAMs) to investigate relationships between fibrosis distribution, including endo-epi differences, adiposity infiltration, and electrophysiological abnormalities, and their influence on PsAF arrhythmogenic substrate. METHODS AND RESULTS:DTs incorporating fibrosis distribution were generated from consecutive PsAF patients' LGE-MRIs. Using rapid pacing, potential locations attracting reentries (LRs) were identified in DTs. CCTs were used to segment adipose tissue within pericardial sac and evaluate the distance from endocardial surface to closest adipose tissue (DEnCA). Bipolar- and unipolar-low-voltage area fractions (LVFs) were extracted from EAMs. Volumetric fibrosis fraction (FF) and surface FF on endocardial and epicardial surfaces at LRs and non-LRs were analyzed in relation to DEnCA and LVF. In 22 patients, adipose tissue volume correlated with BMI and CHA2DS2-VASc and, together with atrial-FF, predicted number of LRs. At LRs and non-LRs, while volumetric-FF was the sole determinant of LR classification, volumetric-FF correlated with endocardium-predominant fibrosis and bipolar-LVF. Endocardium-predominant fibrosis was independently linked to DEnCA. CONCLUSION:This study highlights the complex interactions between structural features, such as three-dimensional fibrosis distribution and atrial adiposity infiltration, electrophysiological features, and PsAF arrhythmogenesis.

Despite the established role of endo-epicardial dissociation in the substrate of persistent atrial fibrillation (PsAF), the underlying mechanisms within the 3D atrial wall that lead to this phenomenon are not fully understood. Using personalized digital twin (DT) technology, we investigated the association between 3D transmurally heterogenous fibrosis architecture and endo-epicardial dissociation in PsAF. LGE-MRI scans from 21 PsAF patients were used to generate bi-atrial DTs. Following virtual induction of reentry using programmed stimulation, a comparative analysis of the electrophysiological and structural properties of reentry-sustaining regions (RRs, n = 91) against controls (non-RRs, n = 73) was performed. RRs exhibited significantly greater endo-epicardial dissociation compared to non-RRs, evidenced by larger differences in median transmembrane potential (60.1 vs. 4.1 mV, P < 0.001) and local activation times (55 vs. 14.9 ms, P < 0.001). Lower dice similarity coefficient (DSC) was observed in RRs compared with non-RRs (median 0.731 (IQR 0.660-0.812) versus 0.880 (IQR 0.642-0.971), respectively; P = 0.0009). In multivariable logistic regression, fibrosis percentage, transmembrane voltage difference, and DSC were independently associated with RR status, whereas LAT difference was not. Our findings reveal that 3D transmurally heterogeneous fibrosis architecture is associated with endo-epicardial electrical dissociation in PsAF, with increased dissociation observed in regions harbouring reentrant circuits. Our findings advance the mechanistic understanding of PsAF arrhythmogenesis and provide a rationale for developing novel mapping and ablation strategies that account for the transmural complexity of the atrial substrate. KEY POINTS: Persistent atrial fibrillation (PsAF) remains a significant clinical challenge, often marked by high rates of post-ablation recurrence. While endo-epicardial dissociation is a known factor in the PsAF substrate, the underlying mechanisms within the 3D atrial wall that lead to this phenomenon remain incompletely understood. In this study, we utilized personalized digital twin technology derived from LGE-MRI scans of 21 patients to investigate the association between transmurally heterogeneous fibrosis and endo-epi electrical dissociation. Our findings reveal that 3D transmurally heterogeneous fibrosis architecture is associated with endo-epicardial electrical dissociation in PsAF. In multivariable analysis, fibrosis percentage, transmembrane voltage difference, and DSC were independently associated with reentrant activity. This work not only advances the mechanistic understanding of PsAF arrhythmogenesis but also suggests that successful clinical ablation strategies must account for the full 3D complexity of the atrial wall to target hidden intramural and epicardial pathways.

Aims Arrhythmogenic fibrotic substrates facilitate reentrant activity in the atria, contributing to the perpetuation of atrial fibrillation (AF). Catheter ablation may disrupt existing reentrant pathways but can also create new ones. This longitudinal study aimed to assess whether post-ablation AF recurrence is associated with incomplete elimination of native arrhythmogenic substrates or emergence of new arrhythmogenic substrates created by ablation lesions, addressing important questions in current AF management: why some patients experience recurrence post-ablation while others do not, and whether ablation lesions themselves contribute to post-ablation arrhythmogenesis.Methods and results Biatrial digital twins (DTs) derived from pre- and post-ablation contrast-enhanced magnetic resonance imaging were used to evaluate the arrhythmogenic propensity of the fibrotic substrate-quantified by potential reentry-sites (PRs) and a vulnerability index (VI) reflecting reentry inducibility. Pre- and post-ablation DT pairs were generated for 11 patients who experienced AF recurrence (R-DTs) and 11 who did not (N-DTs). In total, 58 pre-ablation PRs and 32 post-ablation PRs were detected, with a nearly even distribution of PRs between the left atrium (LA) and right atrium (RA) both pre- and post-ablation. Pre-ablation VI was similar between N-DTs and R-DTs; however, post-ablation VI was significantly higher in R-DTs (P = 0.015). N-DTs exhibited a marked reduction in PRs following ablation, whereas R-DTs did not (P = 0.017). Both groups had few residual PRs from pre- to post-ablation, but R-DTs had many newly emergent PRs. In R-DTs, emergent PRs in the RA were accompanied by a post-ablation increase in RA fibrotic burden. In the LA, where lesions were delivered, all post-ablation reentries anchored around ablation-induced scar (ScAReentries). ScAReentries were significantly more inducible than those occurring within fibrotic substrate and were nearly three times more prevalent in R-DTs, accounting for the elevated post-ablation VI.Conclusion In DTs, emergent PRs in both atria underlie AF recurrence post-ablation, with ablation itself creating some PRs of high arrhythmogenic propensity.

BACKGROUND Pulmonary vein isolation (PVI) is the cornerstone of atrial fibrillation (AF) therapies; however, in persistent AF (PsAF), characterized with fibrosis proliferation, its success is limited. Various extra-PVI strategies have also failed to improve outcome. Although PsAF entails inflammation-driven remodeling throughout the atria, the left atrium (LA) has been the main focus of ablative therapies. We hypothesized that the suboptimal outcomes of PsAF treatment are attributable to the arrhythmogenic role of the right atrium (RA). OBJECTIVE Using personalized digital twins (DTs), an emergent technology recently shown promise in non-invasively characterizing PsAF arrhythmogenesis, we aimed to mechanistically ascertain the difference in PsAF arrhythmogenesis between LA and RA post-PVI. METHODS From 61 patients with PsAF's late gadolinium enhancement magnetic resonance imaging, bi-atrial DTs incorporating fibrosis distribution were constructed. In each DT, following virtual PVI, PsAF arrhythmogenesis was assessed by identifying potential rotor-sustaining locations (LRs) following rapid pacing. Fibrotic and electrophysiological features of LRs were analyzed and compared between the LA and RA. RESULTS Of46 DTs with qualified images, 169 LRs were identified outside the PVI lesions, with substantial LR presence in the RA (66%). Fibrosis fraction within RA-LRs was higher than within LA-LRs. Fibrosis density and entropy and endocardial voltages at RA-LRs were comparable to LA-LRs. Age was found to be the independent determinant of PsAF arrhythmogenesis in RA. CONCLUSION DT analysis demonstrated that the RA contributed equally to PsAF arrhythmogenesis after PVI and exhibited fibrotic and electrophysiological features comparable to those of the LA post-PVI. Controlling RA arrhythmogenesis may be essential for PsAF management.

Background and purpose: Atrial fibrillation (AF), a common arrhythmia, is linked with atrial electrical and structural changes, notably low voltage areas (LVAs) which are associated with poor ablation outcomes and increased thromboembolic risk. This study aims to evaluate the efficacy of a deep learning model applied to 12-lead ECGs for non-invasively predicting the presence of LVAs, potentially guiding pre-ablation strategies and improving patient outcomes. Methods: A retrospective analysis was conducted on 204 AF patients, who underwent catheter ablation. Pre- procedural sinus rhythm ECGs and electroanatomical maps (EAM) were utilized alongside demographic data to train a deep learning model combining Long Short-Term Memory networks and Convolutional Neural Networks with a cross-attention layer. Model performance was evaluated using a 5-fold cross-validation strategy. Results: The model effectively identified the presence of LVA on the examined atrial walls, achieving accuracies of 78 % for both the anterior and posterior walls, and 82 % for the LA roof. Moreover, it accurately predicted the global left atrial (LA) average voltage <0.7 mV, with an accuracy of 88 %. Conclusion: The study showcases the potential of deep learning applied to 12-lead ECGs to effectively predict regional LVAs and global LA voltage in AF patients non-invasively. This model offers a promising tool for the pre- ablation assessment of atrial substrate, facilitating personalized therapeutic strategies and potentially enhancing ablation success rates.

Background: Pulmonary vein isolation (PVI), the gold-standard treatment for atrial fibrillation (AF), is often insufficient for persistent AF (PsAF). In this context, pulsed field ablation (PFA) has emerged as a key modality for PVI. However, PFA have been reported to create wider PVI lesions, dramatically narrowing left atrial (LA) posterior wall. Although the LA roofline lesion has been empirically added to prevent roof-dependent macroreentrant atrial tachycardia (RMAT) when the right and left PVI lines are in close proximity, it is intuitive that PFA lesions would increase the likelihood of RMAT occurrence. Recently, digital twins (DTs) have been utilized to test arrhythmia inducibility and assess the patient-specific substrates. Hypothesis: DTs can be utilized to preoperatively identify the situation under which RMATs would occur after PVI, thereby avoiding unnecessary ablation or redo procedure for RMATs post-PVI. Methods: From LGE-MRI scans of 51 consecutive PsAF patients, personalized biatrial DTs incorporating fibrosis distribution were generated. First, following virtual PVI with typical configuration of line lesions (7.5 mm), RMAT inducibility was tested via burst pacing from 12 sites at LA: roof, anterior wall (upper, middle and lower), posterior wall (upper, middle and lower), inferior wall, lateral wall, LA appendage ostium, and septum (upper and lower). Next, if no RMAT was induced, PVI was repeated, but this time wider lesions were executed by reducing the roof distance by 5 mm (2.5 mm to the right from the left PVI line and 2.5 mm to the left from the right PVI line). Then, the same inducibility test was performed. This process was repeated until the roof gap distance was reduced to below 5 mm. Finally, the fibrotic and electroanatomic features of DTs where RMAT was induced, as well as the occurrence timing, were investigated. Results: Among 40 DTs generated from qualified images, RMATs were induced only in 20 DTs (cycle length: 366±65 ms): 9 after typical PVI and 11 after wider PVI. In 34 DTs, PsAF arrhythmogenicity was also induced after PVI. The LA fibrosis burden is the only determinant for RMAT occurrence (p=0.008). Furthermore, LA fibrosis burden—rather than initial roof distance—was an independent predictor of earlier RMAT induction (p=0.006). Conclusions: LA fibrosis burden significantly contributed to RMAT occurrence post-PVI. DTs enable noninvasive prediction of the need for roofline ablation, diminishing unnecessary or redo ablation.

Pulmonary vein isolation (PVI), the standard-of-care for atrial fibrillation (AF), is effective even in some persistent AF (PsAF) patients despite atrial fibrosis proliferation, suggesting that PVI could not only be isolating triggers but diminishing arrhythmogenic substrates. Left atrial (LA) posterior wall isolation is the prevalent adjunctive strategy aiming to address PsAF arrhythmogenesis, however, its outcomes vary widely. To explore why current PsAF ablation treatments have limited success and under what circumstances each treatment is most effective, we utilized patient-specific heart digital twins of PsAF patients incorporating fibrosis distributions to virtually implement versions of PVI (individual ostial to wide antral) and posterior wall isolation. In most digital-twins (60%) PVI greatly decreased LA substrate arrhythmogenicity without the need of wider lesions or posterior wall isolation. Using digital-twin findings, a strategy was developed to stratify PsAF patients to an appropriate ablation option based on fibrosis features, thus potentially avoiding unnecessary heart damage.

Background: Accurate delineation of the atrial substrate is critical to tailoring therapy for atrial fibrillation (AF). Although bipolar voltage (BV) mapping is widely used, its functional meaning and correlation with local conduction velocity (CV) remain unclear. Unipolar voltage (UV) mapping, which captures deeper tissue, may complement bipolar data by better reflecting transmural structural changes. Objective: To determine whether combining UV and BV thresholds can classify atrial tissue into conduction-based phenotypes. Methods: High-density sinus-rhythm electro-anatomic maps from 63 patients with AF were retrospectively analysed. Local activation times in sinus rhythm were used to compute CV via a triangulation algorithm. Point-by-point CV values were plotted against corresponding unipolar and bipolar voltages. Using iterative threshold optimisation, regions exhibiting similar CV were clustered to define voltage-based tissue classes. Results: Plotting CV against UV–BV pairs revealed that CV rose steadily as both voltages increased (Figure 1). Six UV–BV ranges best separated conduction characteristics as shown in Figure 2: Cat 1: BV < 0.10 mV&UV < 0.50 mV—designated deep scar ; hence, no CV was assigned to this category. Cat 2: 0.10 ≤ BV < 0.50 mV&0.50 ≤ UV < 1.00 mV—median CV 0.53 m/s (IQR 0.27–1.02). Cat 3: 0.10 ≤ BV < 0.50 mV&1.00 ≤ UV < 2.50 mV—0.57 m/s (0.28–1.03). Cat 4: 0.50 ≤ BV < 1.70 mV&0.50 ≤ UV < 1.00 mV—0.62 m/s (0.33–1.13). Cat 5: 0.50 ≤ BV < 1.70 mV&1.00 ≤ UV < 2.50 mV—0.74 m/s (0.73–1.27). Cat 6: BV ≥ 1.70 mV&UV ≥ 2.50 mV—0.88 m/s (0.52-1.45). Pairwise comparisons between adjacent classes (Cat 6 vs 5, 5 vs 4, 4 vs 3, and 3 vs 2) showed progressively faster conduction with higher voltages (t-tests; all p < 0.001, Figure 3). The empirically derived thresholds align closely with histology-guided voltage ranges reported previously. Conclusion: The integration of unipolar and bipolar voltage thresholds effectively enables a conduction-based classification of atrial substrate, offering a promising approach for enhanced substrate characterization in AF.

Atrial fibrillation (AF), the most common heart rhythm disorder, may cause stroke and heart failure. For patients with persistent AF with fibrosis proliferation, the standard AF treatment-pulmonary vein isolation-has poor outcomes, necessitating redo procedures, owing to insufficient understanding of what constitutes good targets in fibrotic substrates. Here we present a prospective clinical and personalized digital twin study that characterizes the arrhythmogenic properties of persistent AF substrates and uncovers locations possessing rotor-attracting capabilities. Among these, a portion needs to be ablated to render the substrate not inducible for rotors, but the rest (37%) lose rotor-attracting capabilities when another location is ablated. Leveraging digital twin mechanistic insights, we suggest ablation targets that eliminate arrhythmia propensity with minimum lesions while also minimizing the risk of iatrogenic tachycardia and AF recurrence. Our findings provide further evidence regarding the appropriate substrate ablation targets in persistent AF, opening the door for effective strategies to mitigate patients' AF burden. Sakata et al. performed a prospective personalized mechanistic computational (digital twin) study focused on characterizing the arrhythmogenic properties of the atrial fibrotic substrate in patients with persistent atrial fibrillation, and they introduce here a novel mechanism-oriented strategy for optimal ablation.

Background Although targeting atrial fibrillation (AF) drivers and substrates has been used as an effective adjunctive ablation strategy for patients with persistent AF (PsAF), it can result in iatrogenic scar-related atrial tachycardia (iAT) requiring additional ablation. Personalized atrial digital twins (DTs) have been used preprocedurally to devise ablation targeting that eliminate the fibrotic substrate arrhythmogenic propensity and could potentially be used to predict and prevent postablation iAT. Objectives In this study, the authors sought to explore possible alternative configurations of ablation lesions that could prevent iAT occurrence with the use of biatrial DTs of prospectively enrolled PsAF patients. Methods Biatrial DTs were generated from late gadolinium enhancement-magnetic resonance images of 37 consecutive PsAF patients, and the fibrotic substrate locations in the DT capable of sustaining reentries were determined. These locations were ablated in DTs by representing a single compound region of ablation with normal power (SSA), and postablation iAT occurrence was determined. At locations of iAT, ablation at the same DT target was repeated, but applying multiple lesions of reduced-strength (MRA) instead of SSA. Results Eighty-three locations in the fibrotic substrates of 28 personalized biatrial DTs were capable of sustaining reentries and were thus targeted for SSA ablation. Of these ablations, 45 resulted in iAT. Repeating the ablation at these targets with MRA instead of SSA resulted in the prevention of iAT occurrence at 15 locations (18% reduction in the rate of iAT occurrence). Conclusions Personalized atrial DTs enable preprocedure prediction of iAT occurrence after ablation in the fibrotic substrate. It also suggests MRA could be a potential strategy for preventing postablation AT.

Predicting time-dependent dynamics of complex systems governed by non-linear partial differential equations (PDEs) with varying parameters and domains is a challenging task motivated by applications across various fields. We introduce a novel family of neural operators based on our Graph Fourier Neural Kernels, designed to learn solution generators for nonlinear PDEs in which the highest-order term is diffusive, across multiple domains and parameters. G-FuNK combines components that are parameter- and domain-adapted with others that are not. The domain-adapted components are constructed using a weighted graph on the discretized domain, where the graph Laplacian approximates the highest-order diffusive term, ensuring boundary condition compliance and capturing the parameter and domain-specific behavior. Meanwhile, the learned components transfer across domains and parameters via Fourier Neural Operators. This approach naturally embeds geometric and directional information, improving generalization to new test domains without need for retraining the network. To handle temporal dynamics, our method incorporates an integrated ODE solver to predict the evolution of the system. Experiments show G-FuNK's capability to accurately approximate heat, reaction diffusion, and cardiac electrophysiology equations across various geometries and anisotropic diffusivity fields. G-FuNK achieves low relative errors on unseen domains and fiber fields, significantly accelerating predictions compared to traditional finite-element solvers.