Optimal near-field unipolar voltage thresholds using a lattice-tip mapping catheter corresponding to bipolar low-voltage areas in atrial fibrillation ablation
T Ozaki, K Inoue, S Kato, M Masuda, S Osaki, T Ohashi, T Mishima, S Kudo, K Hamamoto, J Fujii, H Nakasaki, T Sumigawa, K Nanri, H Miyaguchi, Y UedaAbstract
Background
Conventional electroanatomical mapping for atrial fibrillation (AF) ablation relies on bipolar voltage to identify low-voltage areas (LVAs), with ≥0.5 mV commonly defining healthy myocardium. Recently, integrated lattice-tip catheters have been introduced, allowing for high-resolution mapping using near-field unipolar electrograms alongside radiofrequency and pulsed-field ablation. However, near-field unipolar voltage thresholds corresponding to established bipolar criteria remain undefined, limiting clinical interpretation.
Purpose
To determine the near-field unipolar voltage threshold corresponding to conventional bipolar LVA definitions.
Methods
Five patients undergoing initial AF ablation using a lattice-tip catheter at two centres were retrospectively analysed. After ablation, left atrial voltage maps were sequentially acquired during right atrial pacing using both a bipolar reference mapping system and a near-field unipolar system. Bipolar maps defined healthy myocardium as ≥0.5 mV and scar as <0.05 mV. Near-field unipolar maps were generated with the scar threshold fixed at <0.05 mV and prespecified healthy cut-offs of 0.50, 0.40, 0.30, 0.25, 0.20, and 0.15 mV. Identical anterior–posterior and posterior–anterior views were exported, yielding 10 datasets (5 cases × 2 views). For each dataset, one bipolar reference image and six candidate unipolar images were presented side-by-side in a blinded fashion (Figure 1). Thirteen independent evaluators experienced in electroanatomical mapping selected the best and second-best matches primarily based on the spatial distribution and continuity of preserved voltage regions corresponding to healthy myocardium on the bipolar reference maps. The mean proportion of datasets in which each threshold was recognised as either the best or second-best match was calculated across evaluators, based on 10 datasets per evaluator. These proportions were compared using the Friedman test, followed by post-hoc analysis with Holm correction to account for multiple comparisons.
Results
A total of 130 possible best-match selections were obtained across 10 datasets, of which 124 were included in the analysis after exclusion of non-assessable responses. The 0.20 mV threshold was most frequently selected as the best match (42.7%), followed by 0.15 mV (20.2%) and 0.25 mV (19.4%), whereas higher thresholds ≥0.40 mV were rarely chosen (Figure 2A). Recognition proportions differed significantly across thresholds (Figure 2B; Friedman test, p < 0.001). The 0.20 mV threshold demonstrated the highest recognition proportion (68.4 ± 14.6%), which was significantly higher than other thresholds (p < 0.05 for all vs. 0.20 mV), except for 0.25 mV (p = 0.118).
Conclusion
A near-field unipolar voltage threshold of 0.20 mV most closely corresponded to conventional bipolar LVA definitions. This lower cut-off may prevent the overestimation of LVAs in this near-field unipolar mapping approach.