Ablation Lesion Patterns in Atrial Fibrillation: A Cardiac MRI Comparison of Different Ablation Strategies

Thursday, February 5, 2026

3:45 PM - 3:52 PM

Location: Galapagos / Aruba / Ilhabela (2nd Lower Level)

Presenting Author(s)

Luuk H.G.A Hopman, PhD

Postdoctoral Researcher Cardiology
Amsterdam UMC
Amsterdam, Noord-Holland, Netherlands

Co-Author(s)

Luuk H.G.A Hopman, PhD

Postdoctoral Researcher Cardiology
Amsterdam UMC
Amsterdam, Noord-Holland, Netherlands

Primary Author(s)

Nv

Nikki van Pouderoijen, MSc

PhD Student
Amsterdam UMC, Netherlands

Co-Author(s)

Mv

Marisa van der Graaf

MD
St. Antonius Hospital Nieuwegein, Netherlands
BA

Bob Abeln

MD
St. Antonius Hospital Nieuwegein, Netherlands
MK

Michiel J.B. Kemme, MD, PhD

Cardiologist-Electrophysiologist
Amsterdam Medical Center, Netherlands
ML

Max Liebregts

MD, PhD
St. Antonius Hospital Nieuwegein, Netherlands
LB

Lucas Boersma

MD, PhD
St. Antonius Hospital Nieuwegein, Netherlands
Cor P. Allaart, MD, PhD

Cardiologist
Amsterdam Medical Center
Amsterdam, Noord-Holland, Netherlands
MG

Marco J.W Götte, MD, PhD

MD, PhD
Stephenson Cardiac Imaging Centre
Calgary, Alberta, Canada

Background: Pulmonary vein isolation (PVI) is a cornerstone of atrial fibrillation (AF) ablation, with techniques evolving from point-by-point radiofrequency (RF) ablation to single-shot strategies such as cryoballoon and, most recently, pulsed field ablation (PFA). Late gadolinium enhancement cardiac magnetic resonance (LGE-CMR) has emerged as a noninvasive tool for visualizing and quantifying ablation lesions, offering the potential to characterize “lesion fingerprints” specific to each ablation modality. Therefore, we aim to identify distinctive LGE-CMR–derived lesion patterns for various PVI techniques by integrating original prospective data with a targeted review of the existing literature.

Methods: We studied 42 patients undergoing first-time PVI using one of five ablation modalities: moderate-power RF (MPMD, n=10), very high-power short-duration RF (vHPSD, n=10), RF balloon (n=10), ultra-low temperature cryoablation (ULTC, n=4), or PFA (n=8). High-resolution, ECG-triggered, respiratory-navigated 3D LGE-CMR was performed 3 months post-procedure. Left atrial (LA) enhancement was quantified using ADAS 3D and the full-width at half-maximum (FWHM) method, defining LGE as >40% of maximal signal intensity and subclassifying into borderzone (40–60%) and dense fibrosis ( >60%). Lesion location, extent, distribution, and homogeneity were analyzed. A literature review was conducted to compare observed findings with previously reported LGE-CMR lesion characteristics for each modality.

Results: The cohort (mean age 62±7 years; 78.6% male; 59.5% paroxysmal AF) had a mean LA body LGE burden of 23.5±10.3%. Point-by-point RF (MPMD and vHPSD) generated the most continuous circumferential PV encirclement. Single-shot thermal techniques (RF balloon, ULTC) produced broader circumferential enhancement but with more frequent gaps in PV encirclement. PFA created the widest and most heterogeneous patterns, often with incomplete circumferential coverage. Literature review findings aligned with these observations: cryoballoon ablation typically yields the largest scar burden but less complete PV encirclement than point-by-point RF, and PFA lesions display variable visibility, potentially reflecting differences in injury mechanisms.

Conclusion: LGE-CMR reveals distinct ablation lesion “fingerprints” across PVI modalities. Point-by-point RF produces the most continuous PV encirclement, single-shot thermal techniques create broader but less contiguous lesions, and PFA results in heterogeneous, wide but often incomplete patterns. LGE patterns after PFA may reflect distinct tissue responses to this novel ablation strategy or differences in LGE kinetics compared to thermal energy sources. Validated quantitative thresholds are needed to ensure consistent assessment of lesion durability and its correlation with long-term clinical outcomes.

First author	Year	Ablation modality	Scanner	LGE quantification		Most important CMR findings
McGann et al.	2008	RF point-by-point¹ (n= 46)	1.5T^a	Software: OsiriX & Matlab Analysis: LGE defined as a voxel intensity of 3 SD above normal/healthy tissue mean pixel intensity.		LA scar burden: PVI non-responders: 12.4 ± 5.7% PVI responders: 19.3 ± 6.7%
Halbfass et al.	2014	Cryoballoon² (n= 30)	3.0T^b	Software: Corview Analysis: LGE is identified with a normalized voxel intensity.		Ratio of circumferentially encircled PVs: 33% Complete circumferential ablation lesions: 62% of left PVs vs. 7% of right PVs (p < 0.001)
Akoum et al.	2015	RF point-by-point¹ (n= 157) Cryoballoon² (n= 12) RF single-shot³ (n= 8)	Both 1.5T and 3.0T^f	Software: Corview Analysis: LGE defined as a voxel intensity of 2 SD above normal/healthy tissue mean pixel intensity.		Procedure: PVI-only n= 111 (62.7%); PVI + additional ablation n= 52 (29.4%) Other n= 14 (7.9%) LA scar burden: Overall: 10.7 ± 4.9% PVI-only group: RF: 10.5 ± 4.3% Cryoballoon: 13.4 ± 6.2% RF PVAC: 7.1 ± 2.3% Ratio of circumferentially encircled PVs: 7.3%
Khurram et al.	2015	RF point-by-point¹ (n= 7) Cryoballoon⁴ (n= 5)	1.5T^a	Software: OsiriX and QMass MR Analysis: LGE quantified based on the IIR relative to the blood pool, threshold of ≥0.97.		LA scar burden: RF: 46.3 ± 3.6% (pre-ablation: 29.0 ± 6.8%) Cryoballoon: 50.0 ± 8.6% (pre-ablation: 38.0 ± 10.7%)
Jefairi et al.	2018	RF single-shot⁵ (n= 23) RF point-by-point¹ (n= 28)	1.5T^a	Software: MUSIC Analysis: LGE defined as a threshold of 50-70% of maximum signal intensity.		PV scar burden: 7.8 ± 2.1 mL, higher in patients treated with the RF single-shot catheter (8.6 ± 1.7 mL vs. 7.1 ± 2.2 mL) Gaps in scar formation around PVs were found in: 76% of patients (single-shot: 65% vs. point-by-point: 86%) 39% of PVs (single-shot: 41% vs. point-by-point: 56%)
Kurose et al.	2020	RF point-by-point⁶ (n= 12) Cryoballoon² (n= 18)	1.5T^c	Software: Ziostation Analysis: LGE defined as a voxel intensity of 2 SD above normal/healthy tissue mean pixel intensity. NB: CMR performed 1-3 months		Lesion width and total number of gaps in scar formation: RF 6.3 ± 2.2mm; gaps 13% Cryoballoon: 8.1 ± 2.2mm; gaps 22%
Nakatani et al.	2021	PFA⁷ (n= 18) RF point-by-point¹ (n= 7) Cryoballoon² (n= 16)	1.5T^d	Software: MUSIC Analysis: FWHM, the maximum signal intensity as an internal reference and a threshold set at 50% maximum intensity.		No discrete values of scar at 3 months but reversibility of 60% for PFA and 18% for thermal as compared to MRI within 3 hours post-PVI.
Nelson et al.	2022	RF point-by-point¹ (n= 313) Cryoballoon² (n= 51)	1.5T^f	Software: Merisight Analysis: LGE defined as a voxel intensity of 2-4 SD above normal/healthy tissue mean pixel intensity.		LA scar burden: RF 8.8 ± 4.2% Cryoballoon 6.4 ± 3.6%
Regany et al.	2023	RF point-by-point¹ (n= 47) Cryoballoon² (n= 40)	3.0T^e	Software: ADAS 3D Analysis: LGE quantified based on the IIR relative to the blood pool, threshold of >1.2.		Ratio of circumferentially encircled PVs and number of gaps in isolation per patient: RF: 39%, gaps: 2.7 Cryoballoon: 24%, gaps: 3.2
Sohns et al.	2023	PFA⁷ (n= 10)	Not reported	Software: Merisight Analysis: LGE defined as a voxel intensity of 2-4 SD above normal/healthy tissue mean pixel intensity.		Procedure: PVI and posterior wall isolation: LA scar burden: 8.1 ± 2.1% Mean scar width: 12.8 ± 2.1mm Bilateral PV encirclement: 37 out of 40 PVs (92.7%)
Sciacca et al.	2023	vHPSD RF point-by-point⁸ (n= 30)	3.0T^c	Software: Merisight Analysis: LGE defined as a voxel intensity of 2-4 SD above normal/healthy tissue mean pixel intensity.		LA scar burden: 9.5 ± 1.9% Mean scar width: 13.6 ± 2.9mm Gaps in scar formation around PVs: 10 patients (33.3%), mean 1.8 ± 1 gap/patient
Fink et al.	2025	PFA⁹ (n= 20)	Not reported	Software: Merisight Analysis: LGE defined as a voxel intensity of 2-4 SD above normal/healthy tissue mean pixel intensity.		Bilateral PV encirclement: 16 patients (80%) LA scar burden: 6.1 ± 1.9% Mean scar width: left PVs 12.1 ± 3.1mm; right PVs: 10.7 ± 2.3mm.
Regany‑Closa et al.	2025	RF point-by-point¹ (n= 43) vHPSD RF point-by-point⁸ (n= 25) Cryoballoon² (n= 40) PFA⁷ (n= 30)	3.0T^e	Software: ADAS 3D Analysis: LGE quantified based on the IIR relative to the blood pool, threshold of ≥1.2.		Ratio of circumferentially encircled PVs and mean lesion width: RF: 26%, lesion width: 8.7 mm vHPSD: 40%, lesion width: 10.9 mm Cryoballoon: 24%, lesion width: 13.3 mm PFA: 12%, lesion width: 12.7 mm
Abbreviations: CMR, cardiac magnetic resonance; FWHM, full-width half-max; IIR, image intensity ratio; PV, pulmonary vein; PVI, pulmonary vein isolation; others as reported in previous tables.					Ablation modality ¹THERMOCOOL SMARTTOUCH/NAVISTAR, Biosense Webster ² ArcticFront (Advance), Medtronic ³PVAC, Medtronic ⁴Freezor MAX, Medtronic ⁵nMARQ, Biosense Webster ⁶TactiCath, Abbott ⁷FARAPULSE, Boston Scientific ⁸QDOT Micro, Biosense Webster ⁹VLC, Varipulse, Biosense Webster		Scanner ^aAVANTO, Siemens ^bVerio, Siemens ^cAchieva, Philips Medical ^dAERA, Siemens ^ePRISMA, Siemens

Overview of ablation lesion patterns in atrial fibrillation

LGE lesion fingerprints 3 months post-PVI across different ablation modalities