Fehmi Keçe

Cerebral Embolism during Atrial Fibrillation Ablation 77 3 was common practice in our center already with the first generation PVAC, as in general we often failed to isolate pulmonary veins with 4:1 energy mode. No touch-ups with a single- tip catheter were performed. Thermocool catheter: A point-by-point ablation around both ipsilateral veins was performed until PVI was achieved. RF power was set at 30-35 Wwith a flow rate of 17-20 ml/min and a maximum temperature of 43°C. Post-ablation Phase: After a waiting period of 30 min, PVI was confirmed and 5000 IU protamine was administered before sheath removal. In this study no additional measures (e.g. adenosine testing) were taken to ensure lesion durability. 3.2.3 Cerebral MRI A cerebral MRI (1.5 Tesla, Philips Medical Systems, Best, The Netherlands) was performed on the days before and after ablation. Hyperintensities on the diffusion-weighted image were identified and the corresponding apparent diffusion coefficient maps were calculated. In addition, turbo fluid attenuated inversion recovery (FLAIR) and T2-weighted turbo spin echo sequences were performed. Technical details of the MRI sequences are described in Supplementary Table A . White matter lesions were categorized with the modified Fazekas scale (11). ACE was defined as a new diffusion abnormality on the diffusion-weighted image sequence with an apparent diffusion coefficient reduced map. Cerebral infarcts were defined as positive ACE with a positive FLAIR. Patients with ACE or cerebral infarcts underwent follow-up MRI using the same protocol 3 months later. MRI results were confirmed by 2 independent radiologists. 3.2.4 Transcranial Doppler Ultrasonography 2 MHz transcranial Doppler ultrasonography (DWL Multi Dop P, DWL Sipplingen, Germany) of the right middle cerebral artery was continuously performed from venous access to catheter removal. Raw Doppler signals were recorded as MP3 (Eridol R-09, Roland Corporation Nakagawa, Japan) for off-line analysis. Micro-embolic signals (MES) were automatically detected and discriminated from artefacts using a locally developed MATLAB algorithm (MATLAB R2007b, The MathWorks Inc., Natrick, MA, USA)(12). Number and concentration of MES (MES per unit of time) were calculated for the entire procedure and per ablation phase: pre-ablation, ablation (10s before first RF until 60s after last RF) and post-ablation phase.