Skip to main content
Journal of Arrhythmia logoLink to Journal of Arrhythmia
. 2026 Nov 8;42(6):e70507. doi: 10.1002/joa3.70507

Impedance‐Based Contact Visualization for PulseSelect Pulmonary Vein Isolation: A Single‐Center Feasibility Study

Masahiro Mizobuchi 1,✉, Junya Matsumoto 2, Kaito Oka 2, Hotaka Takemura 2, Tomoki Yamashita 2, Shigeru Nakamura 1
PMCID: PMC13745798  PMID: 43005343

ABSTRACT

Background

PulseSelect pulmonary vein isolation (PVI) lacks a dedicated contact indicator.

Methods

We retrospectively compared 63 patients undergoing PVI before and after EnSite Contact Index (CI) implementation. Because electrical isolation does not confirm lesion transmurality, post‐procedural unipolar electrograms (u‐EGMs) were assessed as acute surrogate markers.

Results

CI guidance was associated with more applications, shorter fluoroscopy, and in an exploratory analysis of analyzable patients, fewer residual high‐voltage u‐EGMs (1/20 vs. 8/24; p = 0.027). No complications occurred.

Conclusions

CI‐guided PVI was feasible; these procedural and acute electrophysiological associations require prospective validation.

Keywords: atrial fibrillation, catheter‐tissue contact, impedance, pulmonary vein isolation, pulsed field ablation, unipolar electrogram


Adaptation of the EnSite Contact Index module to the PulseSelect catheter was technically feasible and was associated with shorter fluoroscopy time. Separately, an exploratory analysis of patients with interpretable maps found fewer residual high‐voltage u‐EGMs after CI implementation. Because of unequal map analyzability, more PFA applications after implementation, sparse events, and a non‐clinically validated CI threshold, the u‐EGM findings require prospective validation.

graphic file with name JOA3-42-e70507-g003.webp


Abbreviations

AF

atrial fibrillation

CI

contact index

LVA

low‐voltage area

PFA

pulsed field ablation

PS

PulseSelect

PVI

pulmonary vein isolation

u‐EGM

unipolar electrogram

1. Background

The PulseSelect (PS) catheter (Medtronic, Minneapolis, MN, USA) is a circular nine‐electrode catheter used for pulsed field ablation of atrial fibrillation. Because PS lacks a dedicated contact indicator, catheter‐tissue contact is usually judged by visual movement of the fifth electrode, fluoroscopy, and intracardiac echocardiography. The Contact Index (CI) module in EnSite X version 3.1.1 (Abbott, Abbott Park, IL, USA) visualizes real‐time electrode impedance changes for a third‐party pentaspline PFA catheter. Although prior reports suggest CI applicability to PS [1, 2], clinical feasibility data remain limited. We evaluated the feasibility of CI‐guided PS pulmonary vein isolation (PS‐PVI) and its association with procedural outcomes and acute electrophysiologic findings.

2. Methods

2.1. Study Design and Ablation Procedure

Between March 2025 and April 2026, 78 consecutive patients underwent AF ablation with PS at Kyoto‐Katsura Hospital and provided written informed consent. This retrospective analysis included 63 patients who underwent PS‐PVI without additional left atrial procedures. Procedures were performed under general anesthesia using fluoroscopy, intracardiac echocardiography, and EnSite X geometry. Bipolar electrograms were recorded with 30‐Hz high‐pass and 300‐Hz low‐pass filters. Low‐voltage areas were defined as < 0.5 mV. At least four applications were delivered at or within each pulmonary vein and four along the antral lesion set. Post‐PFA voltage mapping confirmed PVI, and additional PFA was performed at the operators' discretion. Patients treated before and after CI implementation were classified as Group 1 and Group 2, respectively. Groups were compared for procedure time, radiation exposure, complications, and acute electrophysiologic findings on post‐procedural mapping.

2.2. Electrode Configuration and CI Settings

To apply CI to PS, electrode channels were reassigned to the pentaspline configuration using virtual electrodes formed from adjacent electrode pairs, yielding four independent local‐impedance measurement sites. Details are provided in the Material S1 and Figure S1. Impedance was measured between each electrode and a virtual bipolar electrode formed by its two neighboring electrodes (“tripole sensing”) to reduce interference. CI represents the percentage impedance change from a baseline measured without endocardial contact (Figure 1). Based on the ex vivo experiment, a CI threshold of ≥ 25% was empirically selected as a pragmatic study‐specific target for robust tissue engagement rather than minimal apposition (Material S2 and Figure S2). Sites with CI ≥ 25% were manually tagged on the EnSite voltage map and used as real‐time visual feedback during lesion delivery.

FIGURE 1.

FIGURE 1

Implementation of CI guidance for PS‐PVI. PulseSelect electrodes 1, 3, 5, 7, and 9 were mapped to the five‐electrode pentaspline catheter configuration via jumper cables (Material S1 and Figure S1). The EnSite X system visualized percentage impedance change in real time. Electrodes showing an impedance increase ≥ 25% were highlighted in light blue. ATP–ETP indicates the virtual bipolar electrode formed by two neighboring electrodes. Sites with CI ≥ 25% were manually tagged on the EnSite voltage map and used as real‐time visual feedback during lesion delivery. In the left panel, blue tags indicate the PV orifice and red tags indicate the antral lesion set.

2.3. Unipolar Map Analysis

Post‐procedural unipolar maps were retrospectively analyzed to assess residual high‐voltage unipolar electrograms (u‐EGMs). Signals were recorded with 2‐Hz high‐pass and 100‐Hz low‐pass filters using the EnSite X surface patch as the indifferent electrode. The unipolar low‐voltage cutoff was < 1.08 mV [3]. Because loss of the terminal S wave on atrial u‐EGMs has been associated with lesion transmurality [4], residual u‐EGM morphology was assessed by two independent observers. Offline u‐EGM analysis was considered uninterpretable when reliable visual assessment was precluded by clipped or saturated waveforms, persistent post‐pacing artifacts with baseline instability, or baseline drift of undetermined origin.

3. Results

3.1. Procedural Characteristics and Safety

Patient characteristics and procedural outcomes are summarized in Tables 1 and 2. Group 2 had more applications than Group 1 (51 ± 8.1 vs. 44 ± 5.6, p < 0.001) and shorter fluoroscopy time (27 ± 7.6 vs. 37 ± 13 min, p = 0.001). Procedure time and fluoroscopic dose were numerically lower in Group 2, with significance only for lateral air kerma (p = 0.030). No procedural complications occurred in either group.

TABLE 1.

Patient characteristics.

Background Group 1 (N = 27) Group 2 (N = 36) p
Age (yrs) 72 ± 11 73 ± 11 0.651
Female 7 (26%) 11 (31%) 0.782
Non‐PAF 11 (41%) 16 (44%) 0.802
Prior session 0 (0%) 3 (8.4%) 0.502
BMI (kg/m2) 24 ± 4.3 24 ± 3.2 0.681
Cr (mg/dl) 1.0 ± 0.3 1.1 ± 1.3 0.815
CHA2DS2‐VASc 2.3 ± 2.7 1.8 ± 1.7 0.375
LVEF (%) 60 ± 15 66 ± 7.8 0.055
Left atrial diameter (mm) 40 ± 8.0 40 ± 5.7 0.769
Left atrial volume index (mL/m2) 36 ± 14 35 ± 12 0.876

TABLE 2.

Procedural outcome and u‐EGM analysis.

Procedural outcome (N = 63) Group 1 (N = 27) Group 2 (N = 36) p
PFA applications 44 ± 5.6 51 ± 8.1 < 0.001
Procedure time (min) 94 ± 35 80 ± 20 0.057
Fluoroscopy time (min) 37 ± 13 27 ± 7.6 0.001
Air kerma (Front) (mGy) 262 ± 181 194 ± 120 0.082
Air kerma (Lateral) (mGy) 429 ± 319 293 ± 139 0.030
Dose area product (Gy·cm2) 69 ± 54 50 ± 23 0.073
Offline analysis of u‐EGM in PVI area (N = 44) Group 1 (N = 24) Group 2 (N = 20) p
Patients with residual high‐voltage u‐EGMs (> 1.08 mV) 8 1 0.027
Residual high‐voltage u‐EGM sites 12 2 n/a a
Terminal S wave in residual u‐EGM 12/12 0/2 n/a a
a

Lesion‐level p values were not calculated because multiple sites were clustered within patients and the number of sites was insufficient for reliable cluster‐adjusted analysis.

3.2. Unipolar Map Findings

Offline u‐EGM analysis was possible in 44 of 63 patients (Group 1, n = 24; Group 2, n = 20), after excluding 19 patients because of missing data (n = 1) or uninterpretable u‐EGM recordings during coronary sinus pacing (n = 18). Among the included (n = 44) and excluded (n = 19) patients, Group 2 was more common in the excluded group (84.2% vs. 45.5%; p = 0.005), whereas fluoroscopy time was longer in the included group (34 ± 12 vs. 26 ± 4.7 min; p = 0.012), likely because of group imbalance. No other characteristics differed significantly (Table S1). Representative cases are shown in Figure 2. Residual high‐voltage u‐EGMs (> 1.08 mV) were observed in eight of 24 Group 1 patients and one of 20 Group 2 patients (p = 0.027), corresponding to 12 and two residual sites, respectively. Terminal S waves were present in all 12 residual sites in Group 1 but absent in the two Group 2 sites. Specifically, in Case 45 of Group 2, the posterior RSPV residual high‐voltage region directly overlapped with CI ≥ 25% tags. At the right PV carina, direct correspondence between the residual region and an individual CI ≥ 25% tag could not be determined reliably owing to catheter movement and manual‐tagging uncertainty. Nevertheless, both residual sites showed loss of the terminal S wave. These findings support false‐positive classification by the 1.08‐mV u‐EGM criterion when used alone. Combining voltage amplitude with terminal S‐wave morphology allowed more appropriate discrimination and highlighted the limited specificity of the 1.08‐mV cutoff when used in isolation. Because multiple sites originated from the same patients, these lesion‐level findings are presented descriptively (Table 2 and Figure 3).

FIGURE 2.

FIGURE 2

Representative cases from Group 1 and Group 2. (Left) In Case 5, PVI was performed without CI guidance. A residual lesion at the bottom of the right inferior pulmonary vein was detected only on the unipolar map (> 1.08 mV). The local atrial u‐EGM demonstrated a residual S wave, suggesting a residual intramural lesion. (Right) In Case 45, PVI was performed primarily under CI guidance. Although residual lesions were detected at the carina of the right pulmonary veins and the posterior side of the RSPV, the local atrial u‐EGM demonstrated loss of the S wave, consistent with an acute unipolar morphology associated with transmural lesion formation. Yellow tags indicate CI ≥ 50%, and red tags indicate CI ≥ 25%–< 50%; both were manually confirmed and tagged at each application site. Blue tags indicate applications inside the PVs without numerical recording.

FIGURE 3.

FIGURE 3

The distribution of residual u‐EGMs. The distribution of residual u‐EGMs is shown schematically for both pulmonary veins. Residual sites in Groups 1 and 2 are indicated by pink and light blue circles, respectively. In Group 1, all residual electrograms demonstrated terminal S waves on atrial u‐EGMs (shown in black rectangles). In contrast, residual sites in Group 2, shown as crossed light blue circles, showed no terminal S waves on atrial u‐EGMs (shown in the red rectangle).

Additionally, given the imbalance in PFA application count and the small number of patients with residual high‐voltage u‐EGMs, a post hoc exploratory Firth penalized logistic regression analysis was performed. After adjustment for PFA application count, CI‐guided treatment was associated with lower odds of residual high‐voltage u‐EGMs (adjusted odds ratio, 0.193; 95% CI, 0.021–0.973; p = 0.046), whereas PFA application count was not significantly associated with the outcome (adjusted odds ratio per five applications, 0.758; 95% CI, 0.460–1.352; p = 0.303).

4. Discussion

In this single‐center retrospective study, adaptation of the EnSite CI module to PS was feasible and was associated with shorter fluoroscopy time and fewer residual high‐voltage u‐EGMs after PS‐PVI. Although impedance‐based contact assessment is established, its application to PS, which lacks a dedicated contact indicator, remains insufficiently characterized. This study provides a reproducible workflow for CI‐guided PS‐PVI and preliminary acute electrophysiologic observations after implementation, whereas a clinically validated CI threshold has not been established. Recent clinical studies have proposed lower CI thresholds. Fujiwara et al. identified a 10% cutoff for transesophageal echocardiography‐confirmed apposition with PS [5], whereas Watanabe et al. reported that a ≥ 10% impedance change predicted the absence of acute pulmonary vein gaps with a pentaspline PFA catheter [6]. However, visual apposition may include minimal or feather‐touch contact, which may not ensure robust tissue engagement. Moreover, differences in catheter geometry and electrode configuration may affect CI measurements and limit direct comparison between catheters. Therefore, we empirically selected a more conservative threshold of ≥ 25% based on our preclinical ex vivo experiment, although its clinical validity requires prospective evaluation. Intraprocedural confirmation of lesion completeness after PFA remains challenging. We therefore combined a unipolar voltage criterion reported by Yavin et al. [3] with u‐EGM morphology described by Otomo et al. [4], particularly terminal S‐wave loss. CI guidance was associated with fewer residual high‐voltage u‐EGMs, and the residual sites in Group 2 showed terminal S‐wave loss, whereas all residual sites in Group 1 retained terminal S waves. These findings suggest that CI‐guided PS‐PVI may be associated with more favorable acute surrogate markers after PFA. In addition, shorter fluoroscopy time may reflect reduced reliance on subjective catheter‐motion assessment under fluoroscopy, although this hypothesis requires prospective confirmation.

Several limitations should be acknowledged. First, the retrospective, single‐center, sequential design precludes causal inference, and differences may reflect procedural learning or temporal workflow changes. Second, Group 2 received more applications, which may have contributed to fewer residual u‐EGMs; thus, the observed findings may reflect contact visualization, a more intensive lesion strategy, or both. Third, differential exclusion, mainly due to uninterpretable u‐EGM recordings, may have introduced selection bias; therefore, the u‐EGM analysis was exploratory. Fourth, residual high‐voltage u‐EGMs and terminal S waves are surrogate markers, and offline analysis did not affect intraprocedural decision‐making. Finally, follow‐up remains immature, precluding assessment of durable PVI or long‐term clinical outcomes.

5. Conclusion

Adaptation of the EnSite CI module to the PS catheter was technically feasible and was associated with shorter fluoroscopy time and fewer residual high‐voltage u‐EGMs. Because of unequal availability of interpretable u‐EGM maps between the groups, a greater number of applications after implementation, sparse outcome events, and the use of a non‐clinically validated CI threshold, these findings should be considered exploratory and do not establish improved lesion durability or clinical efficacy.

Funding

The authors have nothing to report.

Ethics Statement

The study was approved by the Institutional Review Board of our institution (Approval No. 2026‐40), and written informed consent was obtained from all individual participants.

Conflicts of Interest

Masahiro Mizobuchi received lecture fees from Abbott Medical Japan, Boston Scientific K.K., Biotronik Japan Co. Ltd., and Medtronic Japan Co. Ltd.

Supporting information

Figure S1: The PulseSelect (PS) catheter and the superimposed pentaspline catheter are shown. Because the Contact Index (CI) was designed for a pentaspline catheter, applying it to the circular PS catheter required reassignment of the electrode channels. For the PS catheter, the odd‐numbered electrodes (E1, E3, E5, E7, and E9) were assigned to the electrode‐3 channels of the five pentaspline splines (red circle). Virtual electrodes (VEs) are formed by two neighboring electrode pairs (VE1: E1–E2, VE3: E3–E4, VE5: E5–E6, and VE7: E7–E8). Impedance measurements for CI were performed between each odd‐numbered electrode and the corresponding VE (E1–VE1, E3–VE3, E5–VE5, and E7–VE7).

Figure S2: The relationship between the electrical impedance rise, measured as the Contact Index (CI) (%), and the distance of the PS catheter from the tissue surface (0–10 mm) is shown. Because CI values may fluctuate, the minimum and maximum values at each distance are plotted. In this ex vivo setup, direct‐contact measurements were ≥ 31%, whereas measurements obtained 2 mm from the tissue were ≤ 16%. A threshold of CI ≥ 25% was therefore selected empirically as a pragmatic study‐specific target for robust tissue engagement. This threshold was not optimized statistically and has not been clinically validated for PS.

Table S1: Baseline and procedural characteristics between included and excluded patients.

Acknowledgments

J.M., K.O., H.T., and T.Y. contributed to patient data collection and analysis. S.N. critically reviewed and proofread the manuscript. The authors gratefully acknowledge Mr. Mitsuhiro Fujita, Ms. Hono Okamoto, and Mr. Ryo Koami of Abbott Japan LLC (Tokyo, Japan), and Mr. Kazuma Yatani of Medtronic Japan Co. Ltd. (Tokyo, Japan), for their technical assistance during the procedures and in the preparation of the preclinical experiment.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

References

  • 1. Mizutani Y., Fujiwara G., Yanagisawa S., and Murohara T., “Enhanced Contact of the PulseSelect Catheter Using the Contact Index During Pulmonary Vein Isolation,” Heart Rhythm 23, no. 8 (2026): 1862–1863, 10.1016/j.hrthm.2025.08.039. [DOI] [PubMed] [Google Scholar]
  • 2. Yamashita K., Kikuchi Y., Yoshiyama K., Kumazawa D., Mizuno Y., and Onodera K., “Novel Impedance‐Guided Contact Mapping Technique for the Circular Multielectrode Pulsed‐Field Ablation Catheter,” Clinical Case Reports 14, no. 3 (2026): e72142, 10.1002/ccr3.72142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Yavin H., Younis A., Zilberman I., et al., “Atrial Endocardial Unipolar Voltage Mapping for Detection of Viable Intramural Myocardium: A Proof‐Of‐Concept Study,” Circulation. Arrhythmia and Electrophysiology 16, no. 1 (2023): e011321, 10.1161/CIRCEP.122.011321. [DOI] [PubMed] [Google Scholar]
  • 4. Otomo K., Uno K., Fujiwara H., Isobe M., and Iesaka Y., “Local Unipolar and Bipolar Electrogram Criteria for Evaluating the Transmurality of Atrial Ablation Lesions at Different Catheter Orientations Relative to the Endocardial Surface,” Heart Rhythm 7, no. 9 (2010): 1291–1300, 10.1016/j.hrthm.2010.06.014. [DOI] [PubMed] [Google Scholar]
  • 5. Fujiwara G., Mizutani Y., Yanagisawa S., and Murohara T., “Contact Index for PulseSelect Catheter Apposition: Validation by Transesophageal Echocardiography,” Pacing and Clinical Electrophysiology (2026), 10.1111/pace.70376. [DOI] [PubMed] [Google Scholar]
  • 6. Watanabe R., Nagashima K., Masuda H., et al., “Contact Index‐Guided Pulsed Field Ablation: A Two‐Phase Validation of the Impedance Change Threshold to Prevent Acute Pulmonary Vein Gaps,” Journal of Arrhythmia 42, no. 2 (2026): e70333. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1: The PulseSelect (PS) catheter and the superimposed pentaspline catheter are shown. Because the Contact Index (CI) was designed for a pentaspline catheter, applying it to the circular PS catheter required reassignment of the electrode channels. For the PS catheter, the odd‐numbered electrodes (E1, E3, E5, E7, and E9) were assigned to the electrode‐3 channels of the five pentaspline splines (red circle). Virtual electrodes (VEs) are formed by two neighboring electrode pairs (VE1: E1–E2, VE3: E3–E4, VE5: E5–E6, and VE7: E7–E8). Impedance measurements for CI were performed between each odd‐numbered electrode and the corresponding VE (E1–VE1, E3–VE3, E5–VE5, and E7–VE7).

Figure S2: The relationship between the electrical impedance rise, measured as the Contact Index (CI) (%), and the distance of the PS catheter from the tissue surface (0–10 mm) is shown. Because CI values may fluctuate, the minimum and maximum values at each distance are plotted. In this ex vivo setup, direct‐contact measurements were ≥ 31%, whereas measurements obtained 2 mm from the tissue were ≤ 16%. A threshold of CI ≥ 25% was therefore selected empirically as a pragmatic study‐specific target for robust tissue engagement. This threshold was not optimized statistically and has not been clinically validated for PS.

Table S1: Baseline and procedural characteristics between included and excluded patients.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.


Articles from Journal of Arrhythmia are provided here courtesy of Japanese Heart Rhythm Society

RESOURCES