Extended Reality Simulator for Dynamic Visualization and Evaluation of Stereotactic Arrhythmia Radioablation (STAR) Treatment Plans within RAVENTA Trial

Figure 1. Challenges in STAR treatments. Panel 1, complex planning workflow: STAR requires integration of multimodal data from different clinical teams. The cardiology team defines the ablation target based on EAM and structural information from cardiac CT, while the radiation oncology team relies on anatomical imaging from the planning CT to design the RT dose plan. Integration of these datasets is complicated by differences in imaging modalities and spatial reference frames. Panel 2, cardiorespiratory motion: intrafractional respiratory and cardiac motion can displace both the target and nearby organs at risk (OARs) during treatment delivery, potentially compromising target coverage and increasing exposure to surrounding structures. Icons from flaticon (https://www.flaticon.com)

Figure 2. Overview of the six steps data processing workflow used to generate motion-aware simulations for STAR. (1) EAM to pCT mapping: EAM, cCT, and pCT are rigidly registered to align anatomical structures. (2) Hollow cardiac model generation: logical operations are applied to subtract the endocardial blood pools (segmented from the contrast-enhanced CT) from the overall heart volume segmented in the pCT, resulting in a hollow cardiac anatomy model. (3) Structures warping: deformable B-spline registration aligns anatomical structures segmented on the pCT to motion-resolved 4DCT phases. (4) Phase-specific dose recalculation: RT dose is recomputed per 4DCT phase using the open-source matRad platform. (5) Cardiac intramural dose masking: organ specific dose accumulation is computed across the respiratory cycle. (6) Data integration and simulation: registered anatomical structures, dose volumes, and EAM data are combined into an interactive XR environment.

Figure 3. (A) Recomputed clinical dose volume overlaid onto the pCT with the considered targets: Full healthy heart, Hollow healthy heart, Full PTV, Myocardial PTV. (B) Multidata integration: Masked intramural cardiac dose, EAM and target. (c.1) Dose distribution within the inner wall of myocardium in respiratory phase 1, FI, and 4, FE, for patient 1. (c.2) Dose distribution within the inner wall of myocardium in respiratory phase 1, FI, and 4, FE, for patient 2. (c.3) Dose distribution within the inner wall of myocardium in respiratory phase 1, FI, and 6, FE, for patient 3.

Figure 4. Respiratory-phase analysis of conformity metrics and OAR dose for the three patients included in the study. Panels (A–C) report the CN for the full PTV and the myocardial PTV, comparing the planning dose conformity with all respiratory phases. Panels (D–F) show the variation of D2 for the full healthy heart and hollow healthy heart, stomach, and oesophagus across phases. Panels (G–I) display ED of the centroid of the full PTV across the respiratory phases with respect to the first phase centroid of the 4DCT. Shaded regions highlight the Full Inhale (FI) and the Full Exhale (FE) intervals.

Figure 5. Experts’ evaluation of the XR simulator for STAR planning. Bars show the proportion of high-agreement responses for each questionnaire item, defined as scores of 6 or 7 on a 7-point Likert scale. Percentages indicate the proportion of respondents reporting high agreement. Given the sample size of six respondents, 83%, 67%, and 100% correspond to 5/6, 4/6, and 6/6 respondents, respectively.