Skip to main content
Springer logoLink to Springer
. 2026 Aug 15;69(9):2999–3011. doi: 10.1007/s10840-026-02414-7

Revisiting low-voltage interpretation: Chamber-specific relationships between myocardial wall thickness and omnipolar voltage

Masateru Takigawa 1,2,✉, Ryosuke Kato 3, Junji Yamaguchi 1, Hidehiro Iwakawa 3, Iwanari Kawamura 1, Masaki Honda 1, Miho Negishi 1, Kentaro Goto 1,2, Takuro Nishimura 1, Kensuke Ihara 1, Susumu Tao 1, Takehiro Iwanaga 4, Sayaka Suzuki 5, Iichiro Onishi 6, Shinsuke Miyazaki 1,2, Tetsuo Sasano 1
PMCID: PMC13769752  PMID: 42603237

Abstract

Background

Electrogram voltage reflects multiple factors, including viable myocardial mass, structural integrity, electrode configuration, catheter orientation, and activation sequence. How local myocardial wall thickness (WT) contributes to voltage amplitude in structurally intact myocardium remains incompletely defined.

Purpose

This study examined the chamber-specific relationships between directly measured WT and unipolar voltage (Uni-V), omnipolar voltage (Omni-V), and peak frequency (PF) in structurally intact swine myocardium.

Methods

High-resolution electroanatomic maps were created in 25 swine. Radiofrequency (RF) lesions were used as anatomical tags to ensure precise spatial correlation between electrophysiological data and post-mortem WT measurements.

Results

A total of 340 atrial and 304 ventricular sites were analyzed. In the atria, WT correlated significantly with Omni-V (R²=0.54, P < 0.0001). This correlation improved after excluding trabeculated regions (R²=0.60, P < 0.0001) and was strongest in non-trabeculated regions with WT < 4 mm (R²=0.65, P < 0.0001). Uni-V also correlated with atrial WT, but less strongly (R²=0.29, P < 0.0001). In contrast, ventricular WT showed only weak associations with Omni-V (R²=0.02, P = 0.07) and Uni-V (R²=0.016, P = 0.06). PF did not correlate with WT in either chamber.

Conclusions

In structurally intact swine myocardium, WT was associated with voltage amplitude primarily in the atrium, with a stronger relationship for Omni-V than Uni-V, particularly in thin, non-trabeculated regions. Ventricular WT showed only weak associations with voltage, and PF was independent of WT. These findings indicate that WT is one anatomical contributor to voltage amplitude under healthy conditions, particularly in the atrium, but should not support using voltage as a direct surrogate for physical tissue thickness or for interpreting voltage behavior in diseased substrate.

Capsule Summary

In structurally intact swine myocardium, atrial wall thickness correlated with omnipolar voltage, whereas ventricular wall thickness showed only weak voltage associations, highlighting chamber-specific anatomical contributors to electrogram amplitude.

Graphical Abstract

graphic file with name 10840_2026_2414_Figa_HTML.webp

In a 25-swine structurally intact heart model, high-resolution electroanatomic maps were generated and compared with direct anatomical measurements of myocardial WT at sites marked by RF fiducial tags. In the atria, Omni-V showed a strong positive association with WT, particularly in thin, non-trabeculated myocardium with WT <4 mm. This relationship was weaker for Uni-V and was reduced in trabeculated atrial regions. In contrast, WT–voltage coupling was weak in the ventricles, and PF did not correlate with WT in either chamber. These findings suggest that local WT may be one anatomical contributor to voltage amplitude in structurally intact atrial myocardium, particularly when assessed using Omni-V, whereas voltage amplitude should not be interpreted as a direct surrogate for tissue thickness or extrapolated to diseased ventricular substrate.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s10840-026-02414-7.

Keywords: Electrophysiology, Wall thickness, Musculature volume, Unipolar, Omnipolar, Peak frequency

Clinical Perspective

What’s new

  • This study provides point-by-point, spatially matched assessment between directly measured myocardial WT and electrogram indices across atrial and ventricular myocardium using RF fiducial tags.

  • In structurally intact atrial myocardium, WT showed the strongest association with Omni-V, particularly in thin, non-trabeculated regions, whereas the association with Uni-V was weaker.

  • In structurally intact ventricular myocardium, WT showed only weak associations with both Omni-V and Uni-V, and PF did not correlate with WT in either chamber.

What Are the Clinical Implications?

  • These findings suggest that, in structurally intact atrial myocardium, local WT may be one anatomical contributor to voltage amplitude by indexing viable myocardial mass within the effective sensing volume, particularly when assessed using Omni-V in thin, non-trabeculated regions.

  • Trabeculated atrial regions, including appendage and pectinate muscle areas, reduce the fidelity of the WT–voltage relationship and should be interpreted cautiously when considering anatomical contributors to voltage amplitude.

  • Conversely, higher Omni-V in structurally intact atrial myocardium may identify regions with greater viable myocardial mass, where careful lesion placement or adjustment of ablation strategy may be required to achieve durable lesion formation; however, this hypothesis requires dedicated validation.

  • Because most voltage amplitudes in this study were within a range generally considered compatible with healthy tissue, these findings should not be interpreted as redefining clinical low-voltage thresholds, as supporting voltage amplitude as a direct surrogate for physical tissue thickness, or as informing voltage interpretation in diseased ventricular substrate.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s10840-026-02414-7.

Introduction

Voltage mapping is widely used to infer myocardial health and guide substrate-based arrhythmia interventions. However, electrogram amplitude reflects a composite of viable myocardial mass, structural integrity, electrode configuration, catheter orientation, contact, and activation sequence, rather than tissue pathology alone [1–10]. Therefore, anatomically thin myocardium may exhibit relatively low voltage even in the absence of diseased substrate, potentially complicating interpretation of low-voltage areas.

Prior studies have suggested that myocardial thickness and recording field size influence electrogram amplitude. Glashan et al. showed that microelectrode-derived bipolar voltage correlated more closely with myocardial thickness than conventional bipolar recordings, whereas human atrial biopsy data demonstrated that lower bipolar voltage was associated with fibrosis, intercellular space expansion, myofibrillar disarray, and reduced nuclear density [4, 11]. These findings support voltage as an integrative marker of both viable myocardial mass and structural integrity.

Nevertheless, most prior studies relating voltage to anatomy have relied on CT- or MRI-derived wall thickness (WT), which may limit precise point-by-point correspondence between electrogram recordings and local myocardial thickness [4, 5, 8–10, 12–16]. In addition, imaging-derived WT in clinical populations may include fibrosis, fat, edema, or other nonmyocyte components, making it difficult to isolate the contribution of viable myocardial mass. Omnipolar mapping may reduce wavefront-direction dependence and provide more stable voltage estimates, but validation against spatially matched ground-truth WT remains limited [17].

Therefore, we sought to define chamber-specific relationships between directly measured WT and electrogram indices in structurally intact swine myocardium. Using RF fiducial tags for precise post-mortem localization, we examined the associations of WT with unipolar voltage, omnipolar voltage, and peak frequency at matched atrial and ventricular sites.

Methods

Study design and ethical approval

This was an acute in-vivo swine study designed to quantify chamber-specific relationships between myocardial WT and electrogram indices at spatially matched sites. The protocol was approved by the Ethical Committee of Institute of Science Tokyo (A2024-125 C). All procedures conformed to institutional guidelines for animal research and complied with ARRIVE reporting standards.

Animals and anesthesia

Twenty-five castrated male swine (age 2.8–4.2 months; weight 48–59 kg; Table 1) were sedated with intramuscular ketamine hydrochloride (10 mg/kg) and xylazine (2 mg/kg). Castrated male swine were selected to reduce biological variability related to sex, hormonal cycling, and body-size differences in this exploratory mechanistic study. After induction with inhaled isoflurane, animals were endotracheally intubated and mechanically ventilated (Carestation/Carescape, GE Healthcare, Chicago, IL) using room air supplemented with oxygen. Anesthesia was maintained with 2%–3% isoflurane. An internal jugular venous sheath was placed for medication and fluid administration. Intravenous amiodarone was infused to reduce ventricular arrhythmia risk. Arterial blood gases were monitored periodically and ventilator settings adjusted to maintain physiological ranges.

Table 1.

Characteristics of the experimental animals

Sex Age, month old Weight, kg
Male (castrated) 3.8 49
Male (castrated) 3.4 50
Male (castrated) 3.6 54
Male (castrated) 3.4 59
Male (castrated) 3.3 50
Male (castrated) 3.5 54
Male (castrated) 3.6 56
Male (castrated) 3.8 52
Male (castrated) 4.2 55
Male (castrated) 3.4 54
Male (castrated) 3.3 51
Male (castrated) 3.6 52
Male (castrated) 3.6 52
Male (castrated) 3.3 55
Male (castrated) 3.3 54
Male (castrated) 3.2 54
Male (castrated) 3.3 52
Male (castrated) 3.5 54
Male (castrated) 2.9 55
Male (castrated) 3.5 53
Male (castrated) 3.0 51
Male (castrated) 3.3 52
Male (castrated) 3.0 54
Male (castrated) 2.8 48
Male (castrated) 3.2 50

Catheter placement and mapping protocol

A decapolar catheter was positioned in the coronary sinus (CS) via the internal jugular vein. High-density electroanatomic mapping was performed in both atria and ventricles using the Advisor™ HD Grid Mapping Catheter, Sensor Enabled™ (Abbott) during CS pacing (cycle length 500 ms). Mapping was performed under fluoroscopic guidance and intracardiac echocardiography (ICE).

Electrogram measurements (Uni-V, Omni-V, PF)

Electrograms were acquired using the Advisor™ HD Grid mapping catheter (Abbott, Abbott Park, IL, USA) and the EnSite X mapping system (Abbott). The Advisor™ HD Grid catheter consists of 16 electrodes arranged in a 4 × 4 grid configuration with 3-mm interelectrode spacing.

Uni-V was recorded between each HD Grid mapping electrode and the Wilson central terminal (WCT), which served as the indifferent electrode. Unipolar electrograms were filtered with a high-pass filter of 2 Hz and a low-pass filter of 100 Hz. Omni-V was automatically calculated by the EnSite X omnipolar algorithm using orthogonally oriented bipoles derived from adjacent HD Grid electrodes. Omnipolar electrograms were filtered with a high-pass filter of 30 Hz and a low-pass filter of 300 Hz. A 50-Hz noise filter was applied during recordings. Peak frequency (PF) was calculated from omnipolar electrograms using fast Fourier transform analysis, as previously described.(18,19).

Adequate catheter–tissue contact was assessed using a prespecified multi-criterion approach: (i) continuous endocardial apposition on fluoroscopy, (ii) a stable catheter–tissue interface on ICE without separation or excessive tissue tenting, (iii) electrogram stability with < 10% amplitude variation over a 3–5 s recording window, and (iv) a proximity indicator within the highest confidence zone. Electrogram acquisition and site inclusion were based on predefined objective recording-quality criteria rather than post hoc subjective signal selection. Points not meeting any of these criteria were excluded a priori. Because the HD Grid catheter does not provide contact-force sensing, recordings were obtained using standardized light-to-moderate pressure guided by fluoroscopy and ICE while avoiding excessive tenting or instability.

All recordings were reviewed offline according to the same predefined quality criteria, and points with mechanical artifact, amplitude fluctuation, or baseline drift were excluded. Because each RF fiducial tag corresponded to several adjacent HD Grid mapping points within a limited local region, the highest stable near-field voltage among points fulfilling all predefined quality criteria was selected as the representative electrogram for that tagged site to reduce underestimation caused by subtle differences in contact or electrode orientation.

Therefore, the present findings should be interpreted in the context of this specific electrode geometry, interelectrode spacing, and omnipolar recording configuration [3, 4].

Radiofrequency applications for fiducial tagging (no electrogram acquisition after RF)

Radiofrequency (RF) energy was delivered with the TactiFlex™ SE catheter using saline irrigation at 13 mL/min (CoolPoint, Abbott Medical) and an Ampere RF generator (St. Jude Medical) delivering 550 kHz unmodulated sine-wave energy in temperature-controlled mode (target 43 °C). RF applications were performed only to create fiducial tags for spatial referencing on the 3D mapping system: 50 W for 5–8 s in the right and left atria (defined as the true atrial walls, excluding pulmonary veins in the left atrium [LA] and including the proximal superior vena cava [SVC] for right atrial [RA] mapping) and 50 W for 10–30 s in the right and left ventricles. All RF applications were tagged on the electroanatomic map. No electrograms were acquired after RF tagging. RF marking lesions were created with sufficient spacing and anatomical orientation to minimize localization ambiguity. Right atrial tags were systematically distributed across predefined anatomical regions, whereas ventricular tags were placed at regular intervals along anatomically recognizable linear trajectories. Because of the small size of the porcine left atrium, left atrial tagging was intentionally limited to a maximum of five lesions per animal.

Post-mortem wall thickness measurement and trabeculation assessment

At the end of the mapping protocol, animals were euthanized under deep general anesthesia by intravenous potassium chloride administration, followed by confirmation of cardiac arrest. The hearts were immediately excised and opened before fixation for direct WT measurement.

To minimize potential deformation from RF energy, WT was measured at the lesion edge and confirmed in adjacent myocardium with preserved architecture using electronic calipers. Measurements were repeated three times at each site, and the median value was used for analysis. During measurement, the tissue was gently supported in a neutral, non-stretched configuration to minimize compression or deformation. Post-mortem WT measurements were performed by an operator blinded to the electrophysiological measurements, including Uni-V, Omni-V, and PF, and to the RF ablation parameters at each site. Trabeculation—defined as pectinate muscle projecting from the endocardial surface—was documented because of its potential influence on electrogram amplitude and near-/far-field composition. Anatomical WT was then correlated with Uni-V, Omni-V, and PF.

Statistical analysis

Associations between WT and electrogram indices (Omni-V, Uni-V, PF) were evaluated using linear regression models. Because multiple measurements were obtained from each animal, cluster-robust standard errors were used with animal as the clustering variable to account for within-animal correlation. Regression coefficients are reported per 1-mm increase in WT with 95% confidence intervals. Thin atrial myocardium was additionally examined using WT ≤ 4 mm, corresponding to the 90th percentile of atrial WT in this cohort. A two-sided P value < 0.05 was considered statistically significant. Analyses were performed using JMP and R. Model assumptions were assessed by residual inspection.

Results

Measured sites

A total of 644 sites were analyzed, including 340 atrial sites (RA, n = 275; LA, n = 65) and 304 ventricular sites (right ventricle [RV], n = 178; left ventricle [LV], n = 126). Among these sites, 337 of 340 atrial sites (99.1%) showed Omni-V ≥ 0.5 mV, and 301 of 304 ventricular sites (99.0%) showed Omni-V ≥ 1.5 mV. All RF tags were clearly identifiable at post-mortem examination, and no tagged sites were excluded because of failed identification, overlap, or uncertain anatomical correspondence. Among atrial sites, 71 (20.9%) were trabeculated, most commonly observed in the left atrial appendage, right atrial appendage, and the anterior portion of the crista terminalis along the RA lateral wall; among ventricular sites, 104 were septal. These sites formed the basis for chamber-specific analyses linking directly measured WT to electrogram indices (Fig. 1).

Fig. 1.

Fig. 1

Spatially matched measurement of electrogram indices and wall thickness using RF fiducial tags. High-density electroanatomic mapping was performed, and radiofrequency applications were delivered solely as fiducial tags to enable precise post-mortem localization of recorded sites. Wall thickness was measured directly at tagged locations, and electrogram indices (Omni-V, Uni-V, and PF) were analyzed at spatially matched sites. Representative examples illustrate differences in atrial WT and corresponding Omni-V values across thin, thicker, and trabeculated atrial regions

Atrial wall thickness and voltage indices

The relationships between atrial WT and voltage indices are summarized in Fig. 2.

Fig. 2.

Fig. 2

Relationship between atrial wall thickness and voltage amplitude. Correlations between atrial WT and Omni-V or Uni-V are shown for all atrial sites and for analyses excluding trabeculated sites and/or restricting the analysis to thin atrial myocardium with WT < 4 mm. Omni-V showed a stronger association with atrial WT than Uni-V, particularly in thin, non-trabeculated atrial myocardium. These findings indicate that local WT may be one anatomical contributor to voltage amplitude under structurally intact atrial conditions, but voltage should not be interpreted as a direct measure of tissue thickness

- Omni-V -

Omni-V correlated strongly with atrial WT (R²=0.54, P < 0.0001). The association strengthened when the analysis was restricted to anatomically more homogeneous regions: excluding trabeculated sites (R²=0.60, P < 0.0001) and further limiting to non-trabeculated thin atrial myocardium (WT ≤ 4 mm) (R²=0.65, P < 0.0001). The WT ≤ 4 mm analysis was performed to focus on thin atrial myocardium, because WT ≤ 4 mm encompassed 90% of atrial sites in this cohort (90th percentile). This restriction also reduced architectural heterogeneity introduced by thicker and/or trabeculated regions.

-Uni-V-

Uni-V also correlated with atrial WT but more modestly (R²=0.29, P < 0.0001). Similar to Omni-V, the relationship improved after excluding trabeculated sites (R²=0.35, P < 0.0001) and was strongest in non-trabeculated thin atrium (WT ≤ 4 mm) (R²=0.42, P < 0.0001). Across all atrial analyses, Omni-V demonstrated a consistently stronger association with wall thickness than Uni-V.

Ventricular wall thickness and voltage indices

Ventricular relationships are shown in Fig. 3.

Fig. 3.

Fig. 3

Relationship between ventricular wall thickness and voltage amplitude. Correlations between ventricular WT and Omni-V or Uni-V are shown for all ventricular sites and by ventricular chamber. In contrast to the atrium, ventricular WT showed only weak associations with both Omni-V and Uni-V in structurally intact myocardium. These findings indicate that local WT alone is not a dominant determinant of voltage amplitude in the healthy ventricle and should not be extrapolated to diseased ventricular substrate

- Omni-V

In contrast to the atrium, Omni-V showed only a weak, non-significant association with ventricular WT (R²=0.02, P = 0.07; Fig. 3A). Neither RV nor LV demonstrated a significant WT–Omni-V relationship. A limited trend was observed in non-septal ventricular regions, whereas septal sites did not show an apparent relationship.

- Uni-V -

Ventricular Uni-V similarly showed a weak, non-significant association with WT (R²=0.016, P = 0.06; Fig. 3B). Overall, Uni-V exhibited slightly lower WT association than Omni-V in the ventricle, yet neither RV nor LV showed a significant WT–Uni-V relationship. As with Omni-V, a weak association was observed in non-septal regions.

Additional exploratory analysis restricted to thinner ventricular myocardium (WT ≤ 6 mm; approximately 25% of ventricular sites) demonstrated modestly improved WT–voltage relationships compared with the overall ventricular cohort (Uni-V: R²=0.14; Omni-V: R²=0.07). However, these associations remained substantially weaker than those observed in the atrium.

PF and wall thickness

PF showed no significant correlation with WT in either the atrium or ventricle (Supplementary Fig. 1).

Histological assessment

To confirm the absence of significant structural remodeling in these young animals, histological evaluation was performed in 7 swine. Representative sections (including myocardium adjacent to tagged regions) are shown in Fig. 4. No evidence of cardiomyocyte nuclear loss, myofibrillar reduction, intercellular space expansion, or fibrosis was observed.

Fig. 4.

Fig. 4

Representative electrograms and corresponding myocardial tissue examples. Representative Omni-V, Uni-V, and PF electrograms are shown with corresponding anatomically matched tissue examples from thin non-trabeculated SVC myocardium (A), thicker LA myocardium (B), trabeculated LA appendage (C), lateral RV myocardium (D), lateral LV myocardium (E), and septal RV myocardium (F). These examples illustrate how voltage amplitude may vary according to local atrial WT and tissue architecture under structurally intact conditions, while also demonstrating that PF does not show a consistent relationship with WT. Representative histological assessment supported interpretation of these electrogram–tissue examples as structurally intact myocardium rather than diseased substrate EGV, Elastica van Gieson staining; HE, Hematoxylin and Eosin staining; LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle; SVC, superior vena cava

Discussion

This study provides a spatially matched assessment of the relationship between myocardial wall thickness and electrogram characteristics across cardiac chambers. Three principal findings emerged. First, atrial wall thickness showed a strong positive association with omnipolar voltage, particularly in thin, non-trabeculated atrial myocardium. Second, ventricular wall thickness demonstrated only weak associations with electrogram voltage. Third, peak frequency was not related to wall thickness in either chamber.

In this acute in-vivo swine study, RF fiducial tags enabled spatially matched comparison between electrogram characteristics and directly measured myocardial wall thickness. This approach reduced point-to-point spatial uncertainty between electrogram sampling and anatomical measurement; however, unlike in-vivo imaging modalities such as CT or MRI, direct post-mortem measurements do not preserve the full three-dimensional myocardial geometry and may be influenced by tissue deformation after extraction [4, 5, 8–10, 12–16]. This approach allowed integration of anatomical and electrophysiological data under controlled conditions across both atrial and ventricular chambers. The present cohort consisted of healthy juvenile swine, and histological evaluation showed no evidence of fibrosis or significant structural remodeling; therefore, the study should be interpreted as a structurally intact myocardial model. In this structurally intact cohort, directly measured WT likely served as a practical anatomical proxy for local viable myocardial mass; however, this relationship may be weaker in clinical populations, where fibrosis, edema, fat, or other nonmyocyte components can contribute to apparent wall thickness. Importantly, because the present study was performed in structurally intact swine myocardium and most voltage amplitudes were within a range generally considered compatible with healthy tissue, the findings should be interpreted as describing physiological contributors to recorded voltage amplitude rather than mechanisms of voltage reduction in diseased substrate.

Atrial wall thickness–voltage coupling in structurally intact myocardium

Relationship between atrial wall thickness and Omni-V

We observed a robust positive association between atrial WT and electrogram amplitude, with the strongest correlation for Omni-V. The relationship strengthened when analyses were restricted to non-trabeculated, thin atrial myocardium (WT ≤ 4 mm), suggesting that anatomical homogeneity and stable tissue contact improve the fidelity of thickness–voltage coupling. In trabeculated regions, correlations were weaker—likely reflecting both complex and non-uniform myocardial architecture and practical recording challenges, including difficulty achieving stable, perpendicular catheter–tissue contact that can underestimate voltage [1, 3, 4, 7]. Accordingly, trabeculated atrial sites, including appendage and pectinate muscle regions, should be interpreted cautiously when considering anatomical contributors to voltage amplitude.

Biological interpretation and technical rationale

From a biological standpoint, voltage can reflect both viable myocardial mass and structural integrity [4, 8–11]. In human atrial septal biopsy data, Takahashi/Yamaguchi et al. showed that lower bipolar voltage was associated with greater fibrosis, intercellular space expansion, myofibrillar disarray, and reduced nuclear density, supporting voltage as an integrative marker of structural remodeling rather than a marker of tissue thickness alone [11]. Consistent with this structurally intact model, local WT may represent one anatomical determinant of viable electrically active myocardial mass within the effective sensing volume and may therefore contribute to electrogram amplitude, particularly in atrial myocardium. Importantly, omnipolar mapping was selected to reduce direction dependence of voltage, which is a known limitation of conventional bipolar recordings driven by wavefront orientation, electrode spacing, and catheter angle [1, 3–5, 7–10, 17]. Although the direction-independent design of omnipolar mapping may be relevant to the observed stronger WT–voltage association for Omni-V, the present study was not designed to directly compare the point-by-point variability of Omni-V with conventional bipolar voltage. Therefore, we avoided interpreting Omni-V as intrinsically less variable than conventional bipolar voltage. Thus, in structurally intact atrial myocardium, local WT may contribute to voltage amplitude by indexing viable electrically active myocardial mass within the effective sensing volume, and Omni-V showed a stronger WT–voltage association than Uni-V under standardized mapping conditions. This should not be interpreted as indicating that voltage directly estimates physical tissue thickness. Rather, WT may contribute to voltage amplitude by indexing viable electrically active myocardial mass within the effective sensing volume in structurally intact atrial myocardium.

Relationship to prior imaging-based studies

These results also help reconcile inconsistencies in prior reports that relied primarily on imaging-derived atrial WT [13–15]. For example, Nakatani et al. reported that thinner LA WT predicted low-voltage areas in paroxysmal AF [14], and Makimoto et al. showed that LA WT correlated with Uni-V more strongly than with bipolar voltage [15]. Across studies, varying degrees of correlation may reflect (i) limited spatial resolution of CT for submillimeter thickness variation, (ii) spatial mismatch between imaging measurements and voltage sampling sites, and (iii) variable degrees of fibrosis and remodeling across patient populations [11, 13–15]. By minimizing spatial mismatch between anatomy and electrogram sampling, this workflow provides a spatially matched anatomical reference for understanding contributors to voltage amplitude in atrial myocardium. In addition, because imaging-derived WT may incorporate nonmyocyte components in remodeled atria, “WT–voltage” relationships in patients may reflect a mixture of viable myocardial mass and structural integrity rather than myocardial mass alone.

Interpretation within the context of structurally intact myocardium

Taken together, these chamber-specific findings should be interpreted within the specific context of structurally intact swine myocardium. Most recorded voltage amplitudes were within a range generally considered compatible with healthy myocardium. Therefore, the present findings should not be interpreted as redefining clinical low-voltage thresholds, explaining voltage behavior in diseased substrate, or supporting voltage amplitude as a direct surrogate for physical tissue thickness. In diseased myocardium, physical WT and viable electrically active myocardial mass may be dissociated by scar, fibrosis, fat, edema, or other nonmyocyte components. For example, a physically thick wall may generate low voltage if scarred or fibrotic, whereas a thin wall may maintain preserved voltage if the viable electrically active myocardial mass within the sensing volume remains intact. Rather, these data provide empirical evidence that local WT may be one anatomical contributor to voltage amplitude in healthy atrial myocardium by indexing viable electrically active myocardial mass within the effective sensing volume, particularly when assessed using Omni-V in thin, non-trabeculated regions. This atrial finding represents the main physiological implication of the present study. By contrast, the weak WT–voltage relationship observed in healthy ventricular myocardium indicates that local WT alone is not a dominant determinant of voltage amplitude in the intact ventricle. Accordingly, the present ventricular data should not be extrapolated to diseased ventricular substrate, where voltage amplitude and tissue thickness may both be influenced by scar, fibrosis, and tissue heterogeneity. Thus, the present findings should be viewed as physiological observations in structurally intact myocardium and should not be extrapolated to clinical low-voltage areas without dedicated validation.

Factors contributing to limited ventricular WT–voltage coupling

Unlike the atrium, ventricular WT showed only weak, non-significant associations with both Uni-V and Omni-V in this healthy swine model. Several factors may explain the limited ventricular WT–voltage coupling observed in the intact ventricle.

  • (A) Absence of tissue pathology.

In this structurally intact model, the absence of scar-related signal attenuation, fibrotic replacement, or major remodeling may explain why local WT alone was not a dominant determinant of surface-recorded voltage in ventricular myocardium. Because voltage attenuation in diseased ventricular substrate is often influenced by scar, fibrosis, and tissue heterogeneity, the present ventricular findings should be interpreted as physiological observations under structurally intact conditions rather than as evidence applicable to diseased ventricular substrate.

  • (B) Effective sensing volume and electrode configuration.

Differences in electrode configuration may also contribute to chamber-specific differences in WT–voltage coupling. Larger electrodes and wider interelectrode spacing may increase the contribution of broader myocardial activation and far-field signals to recorded voltage. In contrast, the high-density HD Grid catheter used in the present study is designed to improve near-field resolution. Thus, the effective “reach” of the recording electrodes should be interpreted relative to local myocardial thickness. In relatively thin atrial myocardium, the effective sensing volume may capture a substantial proportion of the local wall, whereas in thick ventricular myocardium with complex fiber orientation, the same sensing volume may capture only part of the transmural myocardial architecture. This mismatch between effective sensing volume and total ventricular WT may weaken a simple relationship between local endocardial WT and recorded voltage [4, 8–10]. This interpretation is also consistent with the field-of-view concept proposed by Omara et al. [8] They demonstrated that voltage recordings reflect viable myocardial mass within an effective three-dimensional sensing volume, rather than purely local subendocardial tissue characteristics. In their study, the effective field-of-view of unipolar recordings was estimated to be approximately 8–10 mm. However, this should be regarded as a conceptual and context-dependent sensing volume rather than a fixed geometric radius, because recorded voltage may be influenced by electrode configuration, catheter orientation, tissue geometry, activation sequence, and the spatial distribution of viable myocardium.

Consistent with this concept, in the relatively thin atrial myocardium, local WT may correspond more closely to the viable myocardial mass captured within the effective sensing volume, particularly in non-trabeculated regions. In contrast, in the substantially thicker and geometrically complex ventricular myocardium, the same effective sensing volume may include deeper myocardial layers, trabeculations, papillary muscles, and adjacent viable myocardium, rather than local endocardial WT alone. This broader and more heterogeneous sensing volume may partly explain why ventricular WT showed only limited coupling with Uni-V and Omni-V in the present study. This concept is illustrated schematically in Fig. 5.

Fig. 5.

Fig. 5

Conceptual illustration of chamber-specific differences in electrogram field-of-view and effective sensing volume

  • (C) Ventricular geometry and activation complexity.

Ventricular myocardium has greater transmural thickness, anisotropy, and complex anatomical structures such as trabeculations and papillary muscles. These features may weaken the relationship between local WT and surface-recorded voltage. In addition, endocardial electrograms in the ventricle may be influenced by electrical activity arising from deeper myocardial layers and adjacent structures within the sensing volume. Therefore, voltage amplitude in the ventricle may reflect not only local endocardial wall thickness but also the three-dimensional distribution of viable myocardium and the complexity of ventricular geometry.

Activation pattern may further contribute to this chamber-specific difference. In the present study, ventricular activation during CS pacing may have involved substantial His–Purkinje system recruitment and relatively synchronous ventricular activation. This may have increased the contribution of broad ventricular activation to recorded electrograms and reduced the dependence of endocardial voltage amplitude on local WT. Different relationships may be observed during direct myocardial pacing, where activation becomes more localized and anisotropic.

Additional exploratory analysis restricted to thinner ventricular regions (WT ≤ 6 mm) demonstrated modestly improved WT–voltage relationships compared with the overall ventricular cohort; however, the correlations remained substantially weaker than those observed in the atrium. These findings further support the concept that ventricular electrograms are influenced by multiple structural and electrophysiological factors beyond local WT alone.

Collectively, these findings suggest that endocardial voltage amplitude in healthy ventricular myocardium is influenced by multiple structural and electrophysiological factors beyond local WT alone. The absence of scar-related signal attenuation, the effective sensing volume of high-density electrodes, the complex three-dimensional geometry of the ventricle, and activation through the His–Purkinje system may all contribute to the limited ventricular WT–voltage coupling observed in the present intact-heart model [4, 8–10].

Peak frequency: A marker of signal sharpness rather than myocardial mass

PF did not correlate with WT in either chamber. We initially hypothesized that PF might increase in thinner myocardium where far-field influence is reduced and electrograms appear temporally sharper. PF represents the highest-frequency component of the sharpest electrogram deflection and is less dependent on amplitude. In intact myocardium surrounded by abundant viable tissue, far-field contributions can smooth electrograms and reduce PF, which may explain the lack of PF–thickness coupling observed here [17]. Conversely, high PF can be observed in electrically isolated regions, such as discrete muscular sleeves or localized near-field potentials, where far-field contamination is minimized. Consistent with this concept, we previously reported that “low voltage, high PF” regions may identify arrhythmogenic substrate, suggesting PF may complement voltage as an independent descriptor of signal sharpness and near-/far-field balance [17]. The narrower dynamic range of PF compared with voltage may also limit its sensitivity to gradual anatomical differences. Taken together, PF may be better interpreted as a descriptor of electrogram sharpness and near-/far-field balance rather than a surrogate of WT-indexed viable myocardial mass.

Limitations

This was an acute study in healthy juvenile swine with minimal remodeling; therefore, relationships may differ in human myocardium with fibrosis, inflammation, fat infiltration, or chronic AF remodeling. Accordingly, the extent to which directly measured WT reflects local viable myocardial mass may be reduced in remodeled clinical myocardium. WT was measured directly post-mortem using RF fiducial tags, but patient-specific clinical imaging–mapping co-registration was not performed. Although direct post-mortem WT measurement enabled precise spatial matching with electrogram recordings, this approach may be affected by tissue deformation after heart extraction and does not capture the full three-dimensional myocardial architecture available from in-vivo imaging modalities such as CT or MRI, particularly in the ventricle [4, 5, 8–10, 12–16]. All maps were acquired during CS pacing, and paired maps during intrinsic rhythm or alternative pacing sites were not obtained. Therefore, the influence of activation sequence on WT–voltage relationships, particularly in the ventricle, could not be directly assessed. PF analysis was performed using clinically available mapping-system filtering and sampling settings rather than unrestricted high-frequency spectral analysis; therefore, a potential ceiling effect related to acquisition bandwidth, sampling rate, or filtering cannot be excluded and may have reduced sensitivity for detecting subtle WT-related differences in PF. Because the HD Grid catheter does not provide contact-force monitoring, residual variation in catheter–tissue contact cannot be fully excluded despite standardized contact assessment using fluoroscopy, ICE, electrogram stability, and the proximity indicator. Although Omni-V showed a stronger association with atrial WT than Uni-V, this study did not include a prespecified analysis comparing the point-by-point variability of Omni-V with conventional bipolar voltage at matched WT sites. A rigorous comparison of voltage variability would require repeated recordings at identical sites under controlled catheter orientation, tissue contact, and activation conditions. Therefore, whether Omni-V provides lower point-by-point variability than conventional bipolar voltage requires further dedicated study. Finally, findings reflect the recording geometry of the HD Grid catheter and omnipolar computation; extrapolation to other catheter designs or conventional bipolar mapping should be performed cautiously. Although post-mortem WT measurements were performed by an operator blinded to electrophysiological measurements and RF ablation parameters, formal independent blinded adjudication for electrogram/site exclusion was not performed.

Conclusions

In this structurally intact swine model, atrial WT was positively associated with endocardial voltage amplitude, with a stronger relationship for Omni-V than Uni-V, particularly in thin, non-trabeculated atrial myocardium. In contrast, ventricular WT showed only weak associations with voltage amplitude, and PF was not related to WT in either chamber. Because the study was performed in healthy myocardium and most voltage amplitudes were within a range generally considered compatible with healthy tissue, these findings should be interpreted as evidence that local WT may be one anatomical contributor to voltage amplitude under physiological conditions, particularly in the atrium, rather than as a basis for interpreting voltage behavior in diseased substrate or using voltage amplitude as a direct surrogate for tissue thickness.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1 (430.6KB, jpeg)
Supplementary Material 2 (389.3KB, jpeg)

Acknowledgements

We appreciate a technical support for this experiment from Mr. Fumiya Tahara, Mr. Yoshikazu Takano, Mr. Tomoya Watanabe, Mr. Ken Yasui, Mr. Taisei Nigo, Miss. Karen Kasuga, Mr. Kosuke Takahashi, Mr. Heitaro Kawanaka, and Miss. Misa Yoshida from Abbott Medical Japan LLC.

Abbreviations

AF

Atrial Fibrillation

CS

Coronary Sinus

CT

Computed Tomography

ICE

intracardiac echocardiography

LA

Left Atrium

LGE

Late Gadolinium Enhancement

LV

Left Ventricle

LVA

Low-Voltage Area

MRI

Magnetic Resonance Imaging

Omni-V

Omnipolar Voltage

PF

Peak Frequency

RA

Right Atrium

RF

Radiofrequency

RV

Right Ventricle

SVC

Superior Vena Cava

Uni-V

Unipolar Voltage

WT

Wall Thickness

Author contributions

Conception and design of the studyMTFundingMTData CollectionMT, RK, JY, HI, IK, MH, MN, KG, TN, KI, ST, TI, SS, IOAnalysis and interpretation of dataMT, RK, JY, HI, IK, MH, MN, KG, TN, KI, ST, TI, SS, IO, SM, TS. Drafting of the articleMTCritical revision of the article for important intellectual contentMT, SM, TSFinal approval of the articleMT, RK, JY, HI, IK, MH, MN, KG, TN, KI, ST, TI, SS, IO, SM, TS.

Funding

This work was partially supported by Japan Society for the Promotion of Science, Grant/Award Number: 22K16068 and 24K11212.

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest

Drs. Goto, Takigawa and Miyazaki received endowments from Medtronic Japan, Boston Scientific, Japan Lifeline, and WIN international. Dr. Takigawa has received honoraria, speaker fees from Abbott. Takigawa received a research grant from Abbott, Boston Scientific, Medtronic, Biosense Webster, and Japan Life Line, but not associated with this study. No other authors have conflict of interest to declare.

IRB information

This study was approved by Ethical Committee of Institute of Science Tokyo Hospital (A2024-125 C).

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Takigawa M, Relan J, Martin R, et al. Detailed Analysis of the Relation Between Bipolar Electrode Spacing and Far- and Near-Field Electrograms. JACC Clin Electrophysiol. 2018;5:66–77. 10.1016/j.jacep.2018.08.022. [DOI] [PubMed] [Google Scholar]
  • 2.Takigawa M, Martin R, Cheniti G, et al. Detailed comparison between the wall thickness and voltages in chronic myocardial infarction. J Cardiovasc Electrophysiol. 2019;30:195–204. 10.1111/jce.13767. [DOI] [PubMed] [Google Scholar]
  • 3.Takigawa M, Relan J, Martin R, et al. Effect of bipolar electrode orientation on local electrogram properties. Heart Rhythm. 2018;15:1853–61. 10.1016/j.hrthm.2018.07.020. [DOI] [PubMed] [Google Scholar]
  • 4.Glashan CA, Beukers HKC, Tofig BJ, et al. Mini-, Micro-, and Conventional Electrodes: An in Vivo Electrophysiology and Ex Vivo Histology Head-to-Head Comparison. JACC Clin Electrophysiol. 2021;7:197–205. 10.1016/j.jacep.2020.08.014. [DOI] [PubMed] [Google Scholar]
  • 5.Glashan CA, Androulakis AFA, Tao Q, et al. Whole human heart histology to validate electroanatomical voltage mapping in patients with non-ischaemic cardiomyopathy and ventricular tachycardia. Eur Heart J. 2018;39:2867–75. 10.1093/eurheartj/ehy168. [DOI] [PubMed] [Google Scholar]
  • 6.Tedrow UB, Stevenson WG. Recording and interpreting unipolar electrograms to guide catheter ablation. Heart Rhythm. 2011;8:791–6. 10.1016/j.hrthm.2010.12.038. [DOI] [PubMed] [Google Scholar]
  • 7.Takigawa M, Kitamura T, Basu S, et al. Effect of electrode size and spacing on electrograms: Optimized electrode configuration for near-field electrogram characterization. Heart Rhythm. 2022;19:102–12. 10.1016/j.hrthm.2021.09.011. [DOI] [PubMed] [Google Scholar]
  • 8.Omara S, Glashan CA, Tofig BJ, et al. Multisize Electrode Field-of-View: Validation by High Resolution Gadolinium-Enhanced Cardiac Magnetic Resonance. JACC Clin Electrophysiol. 2024;10:637–50. 10.1016/j.jacep.2023.12.003. [DOI] [PubMed] [Google Scholar]
  • 9.Glashan CA, Tofig BJ, Tao Q, et al. Multisize Electrodes for Substrate Identification in Ischemic Cardiomyopathy: Validation by Integration of Whole Heart Histology. JACC Clin Electrophysiol. 2019;5:1130–40. 10.1016/j.jacep.2019.06.004. [DOI] [PubMed] [Google Scholar]
  • 10.Glashan CA, Tofig BJ, Beukers H, et al. Multielectrode Uni-V Mapping and Electrogram Morphology to Identify Post-Infarct Scar Geometry: Validation by Histology. JACC Clin Electrophysiol. 2022;8:437–49. 10.1016/j.jacep.2021.11.012. [DOI] [PubMed] [Google Scholar]
  • 11.Takahashi Y, Yamaguchi T, Otsubo T, et al. Histological validation of atrial structural remodelling in patients with atrial fibrillation. Eur Heart J. 2023;44:3339–53. 10.1093/eurheartj/ehad396. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Glashan CA, Stevenson W, Zeppenfeld K. Lesion Size and Lesion Maturation after Radiofrequency Catheter Ablation for Ventricular Tachycardia in Humans with Nonischemic Cardiomyopathy. Circ Arrhythm Electrophysiol. 2021;14(8):E009808. 10.1161/CIRCEP.121.009808. [DOI] [PubMed] [Google Scholar]
  • 13.Hirata M, Nagashima K, Watanabe R, et al. Where is the gap after a 90 W/4 s very-high-power short-duration ablation of atrial fibrillation? Association with the left atrial-pulmonary vein voltage and wall thickness. J Arrhythm. 2024;40:256–66. 10.1002/joa3.13009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Nakatani Y, Sakamoto T, Yamaguchi Y, Tsujino Y, Kataoka N, Kinugawa K. Left atrial wall thickness is associated with the low-voltage area in patients with paroxysmal atrial fibrillation. J Interv Card Electrophysiol. 2020;58:315–21. 10.1007/s10840-019-00611-1. [DOI] [PubMed] [Google Scholar]
  • 15.Kamioka M, Makimoto H, Watanabe T, et al. Uni-V-based evaluation of left atrial tissue properties and ablation outcome in patients with atrial fibrillation. Europace. 2023;25(9). 10.1093/europace/euad240. [DOI] [PMC free article] [PubMed]
  • 16.Wijnmaalen AP, Van Der Geest RJ, Van Huls CFB, et al. Head-to-head comparison of contrast-enhanced magnetic resonance imaging and electroanatomical voltage mapping to assess post-infarct scar characteristics in patients with ventricular tachycardias: Real-time image integration and reversed registration. Eur Heart J. 2011;32:104–14. 10.1093/eurheartj/ehq345. [DOI] [PubMed] [Google Scholar]
  • 17.Takamiya T, Takigawa M, Noda M, et al. Distribution of peak frequency and omnipolar voltage in electrograms across the atrial body and thoracic veins in a normal heart. J Interventional Cardiac Electrophysiol. 2024;67:2003–15. 10.1007/s10840-024-01845-4. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplementary Material 1 (430.6KB, jpeg)
Supplementary Material 2 (389.3KB, jpeg)

Data Availability Statement

The data supporting the findings of this study are available from the corresponding author upon reasonable request.


Articles from Journal of Interventional Cardiac Electrophysiology are provided here courtesy of Springer

RESOURCES