Olaf Dössel’s research while affiliated with Karlsruhe Institute of Technology and other places

Background and Aims The effective refractory period (ERP) is one of the main electrophysiological properties governing arrhythmia, yet ERP personalisation is rarely performed when creating patient-specific computer models of the atria to inform clinical decision-making. This study evaluates the impact of integrating clinical ERP measurements into personalised in silico models on arrhythmia vulnerability. Methods Clinical ERP measurements were obtained in seven patients from multiple locations in the atria. Atrial geometries from the electroanatomical mapping system were used to generate personalised anatomical atrial models. The Courtemanche cellular model was adjusted to reproduce patientspecific ERP. Four modelling approaches were compared: homogeneous (A), heterogeneous (B), regional (C), and continuous (D) ERP distributions. Non-personalised approaches (A, B) were based on literature data, while personalised approaches (C, D) were based on patient measurements. Modelling effects were assessed on arrhythmia vulnerability and tachycardia cycle length, with sensitivity analysis on ERP measurement uncertainty. Results Mean vulnerability was 3.4±4.0%, 7.7±3.4%, 9.0±5.1%, 7.0±3.6% for scenarios A to D, respectively. Mean tachycardia cycle length was 167.1±12.6 ms, 158.4±27.5 ms, 265.2±39.9 ms, and 285.9±77.3 ms for scenarios A to D, respectively. Incorporating perturbations to the measured ERP in the range of 2, 5, 10, 20, and 50ms changed the vulnerability of the model to 5.8±2.7%, 6.1±3.5%, 6.9±3.7%, 5.2±3.5%, 9.7±10.0% respectively. Conclusion Increased ERP dispersion had a greater effect on reentry dynamics than on vulnerability. Inducibility was higher in personalised scenarios compared to scenarios with uniformly reduced ERP; however, this effect was reversed when incorporating fibrosis informed by low voltage areas. ERP measurement uncertainty up to 20 ms slightly influenced vulnerability. Electrophysiological personalisation of atrial in silico models appears essential and requires confirmation in larger cohorts.

Background and Aims: The effective refractory period is one of the main electrophysiological properties governing arrhythmia maintenance, yet effective refractory period personalisation is rarely performed when creating patient-specific computer models of the atria to inform clinical decision-making. The aim of this study is to evaluate the impact of incorporating clinical effective refractory period measurements when creating in silico personalised models on arrhythmia vulnerability. Methods: Clinical effective refractory period measurements were obtained in seven patients from multiple locations in the atria. The atrial geometries from the electroanatomical mapping system were used to generate personalised anatomical atrial models. To reproduce patient-specific refractory period measurements, the Courtemanche cellular model was gradually reparameterised from control conditions to a setup representing atrial fibrillation-induced remodelling. Four different modelling approaches were compared: homogeneous (A), heterogeneous (B), regional (C), and continuous (D) distribution of effective refractory period. The first two configurations were non-personalised based on literature data, the latter two were personalised based on patient measurements. We evaluated the effect of each modelling approach by quantifying arrhythmia vulnerability and tachycardia cycle length. We performed a sensitivity analysis to assess the influence of effective refractory period measurement uncertainty on arrhythmia vulnerability. Results: The mean vulnerability was 3.4±4.0 %, 7.7±3.4 %, 9.0±5.1 %, 7.0±3.6 % for scenarios A to D, respectively. The mean tachycardia cycle length was 167.1±12.6ms, 158.4±27.5ms, 265.2±39.9ms, and 285.9±77.3ms for scenarios A to D, respectively. Incorporating perturbations to the measured effective refractory period in the range of 2, 5, 10 and 20ms, had an impact on the vulnerability of the model of 5.8±2.7 %, 6.1±3.5 %, 6.9±3.7 %, 5.2±3.5 %, respectively. Conclusion: Increased dispersion of the effective refractory period had a greater effect on reentry dynamics than on mean vulnerability values. The incorporation of personalised effective refractory period in the form of gradients had a greater impact on vulnerability than had a homogeneously reduced effective refractory period. Effective refractory period measurement uncertainty up to 20ms slightly influences arrhythmia vulnerability. Electrophysiological personalisation of atrial in silico models appears essential and warrants confirmation in larger cohorts.

Feature importance methods promise to provide a ranking of features according to importance for a given classification task. A wide range of methods exist but their rankings often disagree and they are inherently difficult to evaluate due to a lack of ground truth beyond synthetic datasets. In this work, we put feature importance methods to the test on real-world data in the domain of cardiology, where we try to distinguish three specific pathologies from healthy subjects based on ECG features comparing to features used in cardiologists' decision rules as ground truth. We found that the SHAP and LIME methods and Chi-squared test all worked well together with the native Random forest and Logistic regression feature rankings. Some methods gave inconsistent results, which included the Maximum Relevance Minimum Redundancy and Neighbourhood Component Analysis methods. The permutation-based methods generally performed quite poorly. A surprising result was found in the case of left bundle branch block, where T-wave morphology features were consistently identified as being important for diagnosis, but are not used by clinicians.

Introduction The role of the right atrium (RA) in atrial fibrillation (AF) has long been overlooked. Computer models of the atria can aid in assessing how the RA influences arrhythmia vulnerability and in studying the role of RA drivers in the induction of AF, both aspects challenging to assess in living patients. It remains unclear if incorporating the RA influences the reentry inducibility of the model. As personalized ablation strategies rely on non-inducibility criteria, the adequacy of left atrium (LA)-only models for developing such ablation tools is uncertain. Aim To evaluate the effect of incorporating the RA in 3D patient-specific computer models on arrhythmia vulnerability. Methods Imaging data from 8 subjects were obtained to generate patient-specific computer models. We created 2 models for each subject: a monoatrial with only the LA and a biatrial with both the RA and LA. We considered 3 different states of substrate remodeling: healthy (H), mild (M), and severe (S). The Courte-manche et al. cellular model was modified from control conditions to a setup representing AF-induced remodeling with 0 %, 50 %, and 100 % changes for H, M, and S, respectively. Conduction velocity was set to 1.2, 1.0, and 0.8 m/s for each remodeling state. Fibrosis extent corresponded to Utah 2 (5-20 %) and Utah 4 ( > 35 %) stages for M and S, while the H state was modeled without fibrosis. Arrhythmia vulnerability was assessed by virtual S1S2 pacing from different points separated by 2cm using openCARP. A point was classified as inducing arrhythmia if reentry was maintained for at least 1 s. The vulnerability ratio was defined as the number of inducing points divided by the number of stimulation points. The mean tachycardia cycle length (TCL) was assessed at the stimulation site. We compared LA vulnerability ratios in monoatrial and biatrial models. Results Incorporating the RA increased the mean LA vulnerability ratio by 115.8 % (0.19 ± 0.13 to 0.41 ± 0.22, p = 0.033) in state M and 29.0 % in state S (0.31 ± 0.14 to 0.40 ± 0.15, p = 0.219). No arrhythmia was induced in the H models. RA inclusion increased the TCL of LA reentries by 5.5 % (186.9 ± 13.3 ms to 197.2 ± 18.3 ms, p = 0.006) in scenario M and decreased it by 7.2 % (224.3 ± 27.6 ms to 208.2 ± 34.8 ms , p = 0.010) in scenario S. RA inclusion increased LA inducibility revealing 5.5 ± 3.0 new points per patient in the LA for the biatrial model, which did not induce reentry in the monoatrial model. Conclusions LA reentry vulnerability in a biatrial model is higher than in a monoatrial model. Incorporating the RA in patient-specific computational models unmasked potential inducing points in the LA. The RA had a substrate-dependent effect on reentry dynamics, altering the TCL of LA-induced reentries. Our results provide evidence for an important role of the RA in the maintenance and induction of arrhythmia in patient-specific computational models, thus suggesting the use of biatrial models.

Atrial fibrillation (AF) is the most common sustained arrhythmia posing a significant burden to patients and leading to an increased risk of stroke and heart failure. Additional ablation of areas of arrhythmogenic substrate in the atrial body detected by either late gadolinium enhancement magnetic resonance imaging (LGE-MRI) or electroanatomical mapping (EAM) may increase the success rate of restoring and maintaining sinus rhythm compared to the standard treatment procedure of pulmonary vein isolation (PVI). To evaluate if LGE-MRI and EAM identify equivalent substrate as potential ablation targets, we divided the left atrium (LA) into six clinically important regions in ten patients. Then, we computed the correlation between both modalities by analyzing the regional extents of identified pathological tissue. In this regional analysis, we observed no correlation between late gadolinium enhancement (LGE) and low voltage areas (LVA), neither in any region nor with regard to the entire atrial surface (-0.3<r<0.3). Instead, the regional extents identified as pathological tissue varied significantly between both modalities. An increased extent of LVA compared to LGE was observed in the septal wall of the LA (asept.,LVA= 19.63% and asept.,LGE= 3.94%, with = median of the extent of pathological tissue in the corresponding region). In contrast, in the inferior and lateral wall, the extent of LGE was higher than the extent of LVA for most geometries (ainf.,LGE= 27.22% and alat.,LGE= 32.70% compared to ainf.,LVA= 9.21% and alat.,LVA= 6.69%). Since both modalities provided discrepant results regarding the detection of arrhythmogenic substrate using clinically established thresholds, further investigations regarding their constraints need to be performed in order to use these modalities for patient stratification and treatment planning.

Atrial fibrillation (AF) is one of the most common cardiac diseases. However, a complete understanding of how to treat patients suffering from AF is still not achieved. As the isolation of the pulmonary veins in the left atrium (LA) is the standard treatment for AF, the role of the right atrium (RA) in AF is rarely considered. We investigated the impact of including the RA on arrhythmia vulnerability in silico. We generated a dataset of five mono-atrial (LA) and five bi-atrial models with three different electrophysiological (EP) setups each, regarding different states of AF-induced remodelling. For every model, a pacing protocol was run to induce reentries from a set of stimulation points. The average share of inducing points across all EP setups was 0.0, 0.8 and 6.7% for the mono-atrial scenario, 0.5, 27.3 and 37.9% for the biatrial scenario. The increase in inducibility of LA stimulation points from mono- to bi-atrial scenario was 0.91 ± 2.03%, 34.55 ± 14.9% and 44.2 ± 14.9 %, respectively. In this study, the RA had a marked impact on the results of the vulnerability assessment that needs to be further investigated.