Islands of Fractionation for Automated Fractionation Mapping in Epicardial Brugada Syndrome Substrate Ablation
Cheng‐Han Chan, Ting Yung Chang, Shih Lin Chang, Chin‐Yu Lin, Ling Kuo, Cheng I. Wu, Chih Min Liu, Lo Chieh Ling, Ding Jun Juan, Danica Ann Leycano, Yat Yu Cheung, Arlene Melissa T. Dychiching, Khuong Tran Dang, Li Wei LoABSTRACT
Introduction
Brugada syndrome epicardial RVOT ablation reduces VF, but procedural endpoints remain debated. Automated fractionation maps are often interpreted as scattered points, which may reflect artifacts.
Objectives
We propose Islands of Fractionation (IOF), a framework comparing EnSite X fractionation count settings against voltage‐defined lesion‐effect reference region.
Methods
This case series included six Brugada syndrome patients undergoing epicardial substrate ablation. Mapping was performed in sinus rhythm using EnSite X with an Advisor HD Grid catheter. Automated fractionation maps were generated with the EnSite X CFE‐count algorithm using refractory settings of 14 and 20 ms. Abnormal points were defined as fractionation count ≥3; lesion‐effect reference region was defined by voltage change (pre >0.5 mV, post <0.3 mV). Island overlap was quantified using DBSCAN clustering and mesh‐based area estimation.
Results
IOF revealed a consistent coverage–parsimony trade‐off between refractory settings. The 14‐ms setting produced broader islands with higher reference‐region coverage (median overlap 84.6% [IQR 80.8–93.7]) but larger extraneous mapped area (up to 72.9 cm 2 ). The 20‐ms setting produced more compact islands with markedly reduced extraneous area (0.0–34.8 cm 2 ) but lower coverage (59.9% [IQR 37.2–74.7]). In parallel, fractionation burden within the ablated region decreased substantially after ablation in both settings (median total burden reduction 86% vs 93%), supporting lesion‐effect concordance of automated fractionation metrics.
Conclusion
IOF provides an island‐based framework for interpreting automated fractionation mapping and describing its spatial concordance with lesion‐effect regions. In this retrospective series, 14 ms favored broader coverage, whereas 20 ms favored greater parsimony.