Article Text


050 Automated analysis of atrial delayed enhancement cardiac MRI correlates with voltage, AF recurrence post-ablation, and high stroke risk
  1. L Malcolme-Lawes1,
  2. C Juli1,
  3. R Karim1,
  4. W Bai1,
  5. R Quest1,
  6. P B Lim1,
  7. S Jamil-Copley1,
  8. P Kojodjojo1,
  9. B Ariff1,
  10. D W Davies1,
  11. D Rueckert1,
  12. R Hunter2,
  13. R Boubertakh2,
  14. S Petersen2,
  15. R Schilling2,
  16. P Kanagaratnam1,
  17. N S Peters1
  1. 1Imperial College London, London, UK
  2. 2Barts and the London Biomedical Research Unit, London, UK


Introduction Visualisation of atrial scar using delayed-enhanced MRI (DE-MRI) may reveal causes for atrial fibrillation recurrence following ablation. To develop an objective method for delineating ablation-scar we compared DE-MRI before and after Cryoballoon ablation on the basis that a predictable antral lesion set would be created. The method was subsequently used in patients undergoing radiofrequency or cryoballoon ablation. Endocardial voltage was measured and compared with corresponding enhancement levels on DE-MRI. Pre and post ablation enhancement was then compared with clinical outcome and risk factors for stroke.

Methods Pre-ablation DE-MRI was performed in 50 patients from two centres, undergoing pulmonary vein isolation; 25 with Arctic Front cryoballoon and 25 with conventional circumferential pulmonary vein ablation. Post-ablation DE-MRI was performed at 3 months. 10 patients at the first procedure and 12 patients returning for a redo procedure underwent endocardial voltage mapping with registration to MRA segmentation for comparison with the DE-MRI. The free-breathing late gadolinium enhanced sequence was also registered to the MRA segmentation and surface intensities were normalised using mean intensity of the blood pool (BlP). Normalised intensity levels were projected onto LA surface as multiples of SD above BlP mean.

Results Ostial and LA intensity were similar in pre ablation scans and greater in the ostia in post ablation scans (p=0.48 and p<0.001 respectively). Bipolar voltage measurements were compared for each intensity level. No significant differences were noted between SD 0 and 1, however significant difference were noted between SD 1 and 2 (1.41±0.03 vs 1.26±0.04 mV p<0.001), SD 2 and 3, (1.26±0.04 vs 0.90 ±0.05 mV p<0.001), SD 3 and 4 (0.90±0.05 vs 0.63±0.04 mV p<0.001) and SD 4 and 5, (0.63±0.04 vs 0.45±0.05 p=0.004). No significant differences were found between intensities > SD 5. Intensity > SD 3 was identified as low voltage atrial tissue <1 mV (scar). % scar was higher in patients with high risk of stroke (CHADS2 >1) compared to low (CHADS2=0), (low vs high 3.19±3.17% vs 7.13±7.38% p=0.035). % scar was also greater in pre ablation scans from patients with AF recurrence vs none following ablation (6.6±6.7% vs 3.5±3.0% p=0.032).

Conclusion We have demonstrated the feasibility of an objective, automated method of DE-MRI analysis of left atrial ablation-scar. Atrial enhancement is reflective of endocardial low-voltage myocardium over a range of intensities and voltages. Pre-ablation atrial scar identified patients at higher risk of stroke and reduced success from AF ablation, suggesting potential future roles for DE-MRI in the management of patients with AF.

Abstract 050 Figure 1

Voltage type. **p<0.001, *p<0.005 compared to preceding intensity level.

  • Atrial fibrillation
  • catheter ablation
  • MRI

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.