Imaging‐Guided Versus Conventional Ventricular Tachycardia Ablation in Ischemic Cardiomyopathy: A Systematic Review and Meta‐Analysis

ABSTRACT

Advanced cardiac imaging can characterize infarct substrate and guide ablation strategy and lesion delivery during ventricular tachycardia (VT) ablation, but its incremental benefit over conventional strategies in ischemic cardiomyopathy (ICM) remains uncertain. We compared imaging‐guided with conventional VT ablation in adults with ICM. This PRISMA‐compliant systematic review and meta‐analysis was registered in PROSPERO (CRD420261399180). MEDLINE, Cochrane CENTRAL, and Scopus were searched from January 1, 2016, through May 21, 2026, with citation tracking. Comparative analyses were restricted to ICM populations or extractable ICM subgroups. The primary outcome was VT recurrence, pooled using random‐effects risk ratios (RRs); single‐arm imaging‐guided cohorts were summarized descriptively. Four comparative studies included 316 patients, of whom 134 underwent imaging‐guided ablation and 182 underwent conventional ablation. Imaging‐guided ablation was associated with lower VT recurrence (pooled RR 0.53, 95% confidence interval 0.36–0.79; p = 0.002; I ² = 0%), with consistent effects across imaging modalities and study designs. Major complications were infrequent and numerically similar between groups, although sparse events and heterogeneous definitions precluded pooling. Three additional single‐arm imaging‐guided studies reported favorable VT‐free or event‐free survival, acceptable safety, and high acute non‐inducibility, but were not used to infer comparative efficacy. In patients with ICM undergoing scar‐related VT ablation, imaging‐guided ablation was associated with lower VT recurrence without an apparent increase in major complications; given the limited comparative evidence, these findings are hypothesis‐generating rather than practice‐changing. Larger multicenter randomized trials are required to confirm the magnitude of benefit and define patient selection, imaging modality, and integration strategy.

Imaging‐Guided Versus Conventional Ventricular Tachycardia Ablation in Ischemic Cardiomyopathy: A Systematic Review and Meta‐Analysis. Selected mapping images were adapted from Uhm et al. and Narayan et al. under CC BY 4.0; cropped and reformatted for this graphical abstract.

Abbreviations

AAD

antiarrhythmic drug

CI

confidence interval

CMR

cardiac magnetic resonance

CT

computed tomography

EAM

electroanatomical mapping

ECGi

electrocardiographic imaging

GRADE

Grading of Recommendations, Assessment, Development and Evaluation

HR

hazard ratio

ICD

implantable cardioverter‐defibrillator

ICM

ischemic cardiomyopathy

IQR

interquartile range

LGE

late gadolinium enhancement

MI

myocardial infarction

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta‐Analyses

PROSPERO

International Prospective Register of Systematic Reviews

RCT

randomized controlled trial

RF

radiofrequency

RoB 2

Cochrane Risk of Bias 2 tool

ROBINS‐I

Risk of Bias In Non‐randomized Studies of Interventions

RR

risk ratio

SD

standard deviation

VT

ventricular tachycardia

1. Introduction

Catheter ablation is an established therapy for scar‐related ventricular tachycardia (VT) in patients with ischemic cardiomyopathy or prior myocardial infarction [1]. However, long‐term freedom from VT remains suboptimal, with recurrence after a single procedure still common in contemporary practice [2, 3]. Recent network meta‐analytic evidence comparing VT ablation with Class III antiarrhythmic drug strategies in ICM further supports the clinical importance of ablation in reducing recurrent ventricular arrhythmia (VA) burden, although recurrence after ablation remains an important limitation [4]. Conventional VT ablation still relies on electroanatomical mapping, pace mapping, activation or entrainment mapping when feasible, and targeting of late or local abnormal ventricular activities with and without functional substrate mapping. These techniques are essential but provide only an indirect representation of the three‐dimensional infarct substrate and may miss intramural, epicardial, or heterogeneous conducting channels [5, 6, 7, 8].

Advanced cardiac imaging may improve substrate characterization and procedural planning. Late gadolinium enhancement cardiac magnetic resonance (LGE‐CMR) can define scar transmurality, border‐zone heterogeneity, and potential conducting channels, while contrast‐enhanced cardiac computed tomography (CT) and CT‐derived three‐dimensional models can identify wall thinning, myocardial thickness heterogeneity, calcification, and potential VT isthmus regions. Integration of these imaging data into electroanatomical mapping systems may refine target selection and reduce reliance on extensive intraprocedural mapping [9, 10, 11]. In this sense, imaging‐guided ablation is better conceptualized as an imaging‐informed substrate modification strategy, in which pre‐procedural imaging reshapes intraprocedural decision‐making regarding access, mapping density, and lesion targeting, rather than as imaging serving merely as a passive adjunct to conventional mapping.

Previous observational studies and more recent CT‐guided workflows suggest that imaging‐guided or imaging‐aided VT ablation may improve procedural efficiency, acute non‐inducibility, and arrhythmia outcomes compared with conventional approaches [10, 11, 12]. A prior meta‐analysis published in 2020 reported improved VT‐free and overall survival with image‐guided ablation, but it included relatively few imaging‐guided cohorts and did not incorporate newer CT‐based, CMR‐based, multimodality, and randomized evidence [13].

Therefore, this systematic review and meta‐analysis aims to evaluate whether imaging‐guided VT ablation using cardiac CT and/or cardiac MRI is associated with reduced VT recurrence compared with conventional non‐image‐guided ablation in adults with ischemic cardiomyopathy or prior myocardial infarction. It also summarizes outcomes from single‐arm imaging‐guided cohorts to describe contemporary clinical performance, acute success, safety, and long‐term arrhythmia outcomes in this population.

2. Methods

This systematic review and meta‐analysis followed recommendations of the Preferred Reporting Items for Systematic Reviews and Meta‐Analysis (PRISMA) statement 2020 guidelines and was prospectively registered in PROSPERO (CRD420261399180) [14]. MEDLINE via PubMed, Cochrane CENTRAL, and Scopus were searched from 1 January 2016 to 21 May 2026. The search combined terms for ventricular tachycardia, catheter ablation, ischemic cardiomyopathy or post‐infarction scar, and imaging guidance using cardiac CT or CMR. Reference lists of relevant articles and previous reviews were screened manually. The full search strategy is provided in the Supporting Information: Methods.

Eligible studies enrolled adults with ICM, prior myocardial infarction, or clearly defined infarct‐related ventricular scar undergoing catheter ablation for scar‐related monomorphic VT. Comparative studies were required to include an imaging‐guided ablation arm and a conventional non‐image‐guided ablation arm. Single‐arm studies were eligible if pre‐procedural CT and/or CMR were used to characterize scar and guide ablation strategy or lesion placement. Mixed‐etiology cohorts were included only when ICM subgroup data were extractable for the primary endpoint. Studies were excluded if they focused on non‐ischemic cardiomyopathy without extractable ICM data, idiopathic VT, premature ventricular complexes, arrhythmogenic cardiomyopathy, congenital heart disease, channelopathies, imaging used only for anatomical description without procedural guidance, sample size < 10 patients, or non‐original reports.

Titles and abstracts were screened, followed by full‐text review of potentially eligible studies. Data extraction was performed using a standardized spreadsheet and reviewed by the senior reviewer. Extracted variables included study design, population, imaging modality, imaging role, ablation context, baseline characteristics, procedural metrics, complications, VT recurrence, appropriate ICD therapies or shocks, mortality, and follow‐up duration. For mixed cohorts, only ICM data were used in comparative analyses when available. Where ICM subgroup recurrence data or time‐to‐event estimates were indirectly derived from published subgroup information or Kaplan–Meier curves, these values were labeled as derived and explored in sensitivity analyses.

The primary outcome was VT recurrence after index ablation in patients with ICM. For arm‐level risk‐ratio analyses, recurrence counts were extracted at the longest available study‐reported follow‐up. Follow‐up ranged from 6 months to approximately 31 months across the comparative studies. Time‐to‐event estimates were evaluated separately in sensitivity analyses.

Risk of bias was assessed using RoB 2 for randomized trials and ROBINS‐I for non‐randomized comparative studies [15, 16]. Certainty of evidence was assessed using GRADE [17]. Comparative studies formed the primary meta‐analysis. Risk ratios (RRs) with 95% confidence intervals (CIs) were calculated from arm‐level recurrence counts and pooled using random‐effects models. Heterogeneity was assessed using I ². Hazard ratios were extracted or reconstructed from Kaplan–Meier curves when possible and used only in sensitivity analyses. Exploratory subgroup and sensitivity analyses assessed study design, imaging modality, follow‐up duration, redo‐only cohorts, indirectly derived data, and risk of bias. Publication bias was assessed only exploratorily because fewer than 10 comparative studies were available.

3. Results

The search identified 766 database records, with one additional eligible study identified through citation tracking. After removal of duplicates, 535 records were screened and 60 reports underwent full‐text assessment. Seven studies met the inclusion criteria: four comparative studies of imaging‐guided versus conventional ablation and three single‐arm imaging‐guided cohorts. The study selection process is shown in Figure 1. The comparative studies included one randomized controlled trial and three non‐randomized cohorts. Imaging guidance was CMR‐based in two comparative studies and CT‐based in two. All comparative analyses were restricted to ICM populations or extractable ICM subgroups. Study characteristics are summarized in Table 1.

FIGURE 1.

PRISMA 2020 flow diagram showing study identification, screening, eligibility assessment, and inclusion for the systematic review and meta‐analysis of imaging‐ guided versus conventional VT ablation in ischemic cardiomyopathy.

TABLE 1.

Characteristics of included studies.

Study	Design	Population	ICM population n (%)	Imaging modality	Imaging protocol	Outcomes available for ICM only focused analysis
Andreu 2017	Prospective observational, single‐center	Mixed	121 (76%)	LGE‐CMR (1.5 T)	PSI scar maps using ADAS‐VT (now marketed as ADAS 3D) integrated into CARTO for scar dechanneling	VT recurrence; reconstructed time‐to‐event estimate
Ghannam 2022	Retrospective observational, single‐center	ICM only	33 (100%)	LGE‐CMR (1.5 T)	LGE‐CMR dark‐core lesions registered to the EAM before repeat ablation	VT recurrence/VT‐free survival; procedure and RF times; acute non‐inducibility; major complications
Englert 2024	Retrospective observational, single‐center	Mixed	49 (51%)	Cardiac CT	Late‐enhancement CT segmented by inHEART into a 3D model and registered with CARTO mapping system	VT recurrence; VT‐free survival (ICM subgroup estimates indirectly derived from Kaplan–Meier curve)
Sacher 2026	Prospective randomized, multicenter	ICM only	113 (100%)	Cardiac CT	AI‐segmented CT with core‐lab isthmus annotations imported into the EAM	Procedure duration; VT recurrence/VT‐free survival; VT burden; major adverse events
Nunes‐Ferreira 2025	Prospective observational, single‐center (single‐arm)	Mixed	75 (73.5%)	Cardiac CT ± LGE‐CMR (3 T)	MDCT and/or LGE‐CMR segmented with ADAS 3D and integrated into high‐density EAM maps	Appropriate ICD shock‐free survival; mortality; complications; procedure data; VT inducibility
Sacristan 2025	Retrospective observational, single‐center (single‐arm)	ICM only	39 (100%)	Cardiac CT	CT wall‐thickness maps generated with inHEART; CT channels merged into the EAM	VT‐free/event‐free survival; acute non‐inducibility; procedure/RF/fluoroscopy times; complications
Penela 2025	Prospective observational registry, multicenter (single‐arm)	Mixed	121 (70.8%)	LGE‐CMR (1.5/3 T) and/or Cardiac CT	CMR PSI and/or MDCT wall‐thinning maps generated with ADAS 3D and integrated into the navigation system to guide a scar dechanneling strategy	VT recurrence‐free survival; overall survival; major adverse events; procedure/RF/fluoroscopy times

Abbreviations: CMR, cardiac magnetic resonance; CT, computed tomography; EAM, electro‐anatomical mapping; ICD, implantable cardioverter‐defibrillator; ICM, ischemic cardiomyopathy; KM, Kaplan–Meier; LGE, late gadolinium enhancement; MDCT, multidetector computed tomography; PSI, pixel signal intensity; RCT, randomized controlled trial; RF, radiofrequency; VT, ventricular tachycardia.

Baseline characteristics were consistent with a contemporary scar‐related VT population, with predominantly male patients, reduced left ventricular ejection fraction, and frequent implantable cardioverter‐defibrillator use. Some imbalance was present in the non‐randomized studies, particularly Ghannam [18], in which the imaging‐guided arm had more electrical storm but higher baseline LVEF and less prior amiodarone use than the conventional arm. These imbalances were incorporated into the risk‐of‐bias assessment. Detailed baseline characteristics are provided in Table S1A,B.

Procedural outcomes were inconsistently reported and were not pooled. ICM‐specific procedure‐time data were available only from Ghannam [18] and Sacher [19]. Procedure duration was similar between arms in the redo‐only cohort by Ghannam [18], whereas CT‐guided ablation was associated with shorter procedure duration in the randomized Sacher [19] trial. Major complications were infrequent and numerically similar across arms, but sparse events and heterogeneous definitions precluded pooled analysis. Procedural characteristics are summarized in Tables S2A and S2B.

Four comparative studies contributed to the primary VT‐recurrence meta‐analysis. Imaging‐guided ablation was associated with a lower risk of VT recurrence than conventional ablation in patients with ICM (pooled RR 0.53, 95% CI 0.36–0.79; p = 0.002; I ² = 0%; Figure 2, Table 2). The direction of effect was consistent across studies despite differences in imaging modality, study design, and ablation context. Excluding Englert 2024, for which the ICM subgroup event counts in both treatment arms were indirectly reconstructed from the published subgroup information and Kaplan–Meier figure produced a similar estimate (RR 0.55, 95% CI 0.36–0.84; Table S4).

FIGURE 2.

Forest plot for the primary outcome (VT recurrence). † Englert 2024 ischemic‐subgroup recurrence event counts in both treatment arms were indirectly reconstructed from published subgroup information and the Kaplan–Meier figure. CI, confidence interval; ICM, ischemic cardiomyopathy; REML, restricted maximum likelihood; RR, risk ratio; VT, ventricular tachycardia.

TABLE 2.

Primary outcome and key secondary outcomes across comparative studies.

Study	Imaging vs. conventional, n	Follow‐up duration, mean/median (months) a	VT recurrence, n/N (%)			Procedure duration (min)			Major complications
			Imaging‐guided arm	Conventional arm	Effect estimate (95% CI)	Imaging‐guided arm	Conventional arm	Difference	Imaging‐guided arm	Conventional arm
Andreu 2017	37 vs. 84	20 months	18.9% (7/37)	42.9% (36/84)	RR 0.44 (0.22–0.90)	NR for ICM subgroup			Overall, 8.8% (14/159); not stratified by arm
Ghannam 2022	19 vs. 14	31 months	21.1% (4/19)	50.0% (7/14)	RR 0.42 (0.15–1.16)	454 ± 93 min	445 ± 146 min	+9 min (NS)	0% (0/19)	0% (0/14)
Englert 2024 b	21 vs. 28	6 months	14.3% (3/21)	35.7% (10/28)	RR 0.40 (0.13–1.28)	NR for ICM subgroup			NR for ICM subgroup
Sacher 2026 c	57 vs. 56	12 months	23.2% (13/56)	32.7% (18/55)	RR 0.71 (0.39–1.30)	120 ± 50 min	149 ± 51 min	−29 min	1.8% (1/57)	3.6% (2/56)
Pooled estimate (random‐effects)	134 vs. 182	6–31 months	Pooled RR 0.53 (0.36–0.79), p = 0.002, I 2 = 0%			Not pooled (procedure times reported inconsistently)			Not pooled (sparse events and heterogeneous definitions)

Abbreviations: CI, confidence interval; ICM, ischemic cardiomyopathy; KM, Kaplan–Meier; NS, not significant; RR, risk ratio; VT, ventricular tachycardia.

^a Follow‐up duration is reported as provided in each comparative study and converted to months. The arm‐level recurrence analysis used the longest available study‐reported follow‐up.

^b Englert 2024 ischemic‐subgroup denominators and recurrence event counts for both treatment arms were indirectly reconstructed from the published subgroup information and Kaplan–Meier figure.

^c Sacher 2026, randomized/procedural arm sizes were 57 CT‐guided and 56 conventional; VT‐recurrence denominators reflect patients with available 12‐month VT‐free survival data, 56 and 55, respectively.

Secondary outcomes generally favored imaging‐guided ablation but were limited by inconsistent reporting. Reconstructed time‐to‐event sensitivity analyses were consistent with the primary risk‐ratio analysis. Acute procedural efficacy was directionally favorable in imaging‐guided cohorts, although endpoint definitions differed across studies and the available comparisons were underpowered. Ghannam [18] reported higher complete noninducibility with imaging guidance, whereas Sacher [19] found similar final monomorphic VT noninducibility and numerically lower inducibility of any sustained ventricular arrhythmia in the CT‐guided arm. All‐cause mortality and cardiovascular death were too sparsely reported for pooling. Major complications were infrequent, but sparse events and heterogeneous definitions precluded a reliable comparative safety estimate. Key secondary outcomes across comparative studies are summarized in Table 2.

The three single‐arm imaging‐guided cohorts were summarized descriptively because endpoint definitions and follow‐up durations differed. These studies reported favorable VT‐free or event‐free survival, acceptable safety, and generally high acute non‐inducibility, but they were not directly comparable with conventional ablation and were not pooled. Individual cohort outcomes are provided in Table S3, and descriptive single‐arm recurrence proportions are shown in Figure S9.

Exploratory subgroup and sensitivity analyses were directionally consistent with the primary finding. The randomized trial and the pooled non‐randomized studies both favored imaging‐guided ablation. CMR‐guided and CT‐guided studies also showed protective point estimates, although the small number of studies prevented formal comparison. Excluding the redo‐only cohort, restricting analysis to studies with at least 12 months of follow‐up, and leave‐one‐out analyses did not materially alter the conclusion. The randomized‐trial‐only analysis remained directionally favorable but was underpowered. Full subgroup, sensitivity, leave‐one‐out, and descriptive recurrence analyses are provided in Table S4, Figures S1–S10.

Certainty of evidence for VT recurrence was graded as moderate, with one‐level downgrading for residual risk of bias arising from non‐randomized treatment allocation and potential unmeasured confounding; no further downgrading was applied for inconsistency or imprecision (Table S5). Domain‐level risk‐of‐bias judgments are provided in Figure S11. Publication‐bias assessment was exploratory and underpowered because only four comparative studies were available. The exploratory funnel plot and Egger's test (intercept −1.91, p = 0.237; Figure S10) are provided for transparency only and were not used to draw conclusions regarding publication bias.

4. Discussion

This systematic review and meta‐analysis evaluated imaging‐guided versus conventional VT ablation in patients with ICM. The principal finding was that CT‐ or CMR‐guided VT ablation was associated with a significantly lower risk of VT recurrence compared with conventional ablation. This association was consistent across four comparative studies, including both CT‐ and CMR‐based workflows, with no statistical heterogeneity for the primary endpoint. It is important to note, however, that the included studies used substantially different imaging‐guided workflows, including LGE‐CMR‐based scar dechanneling, CT‐derived wall‐thickness mapping, AI‐derived CT isthmus annotation, and different segmentation/integration platforms (e.g., ADAS 3D, inHEART) with dark‐core lesion registration. Accordingly, imaging‐guided ablation should not be interpreted as a single, uniform intervention, and the observed benefit may not generalize equally across all imaging modalities, segmentation platforms, and integration workflows.

Although procedural outcomes were reported inconsistently, available data suggested that imaging guidance may improve procedural efficiency in selected settings without an apparent increase in major complications. These findings should be regarded as supportive and hypothesis‐generating rather than definitive or practice‐changing. The true magnitude of benefit remains uncertain, and the certainty of evidence is limited by the small number of comparative studies and the predominance of observational data; adequately powered randomized trials are needed before these findings can inform routine clinical practice.

To our knowledge, this is the first meta‐analysis to specifically compare imaging‐guided versus conventional VT ablation in an ischemic cardiomyopathy‐only population. This distinction is clinically important because restricting the analysis to ICM reduces substrate heterogeneity and focuses the inference on post‐infarction scar‐related VT, where CT‐ and CMR‐derived substrate characterization may be particularly relevant.

The observed benefit is mechanistically plausible. Conventional electroanatomical mapping remains central to VT ablation, but it provides an indirect functional representation of post‐infarction scar and may incompletely capture intramural, epicardial, or heterogeneous conducting channels [5, 6, 20]. Advanced imaging can complement invasive mapping by defining scar architecture before the procedure, guiding access strategy, focusing mapping density, and refining lesion placement. LGE‐CMR can characterize scar transmurality, border‐zone heterogeneity, and potential conducting channels, whereas CT can provide high‐resolution anatomical models, wall‐thickness maps, calcification, thrombus assessment, and spatial relationships relevant to procedural planning [6, 20, 21].

Our findings extend previous evidence in a more contemporary and ICM‐specific manner. Hendriks et al. reported higher VT‐free survival with image‐guided than non‐image‐guided ablation, but their review included heterogeneous populations and predated several recent CT‐based, CMR‐based, and randomized datasets [13]. More recent meta‐analyses evaluating CMR integration in structural heart disease similarly suggested lower ventricular arrhythmia recurrence with CMR‐guided or CMR‐aided workflows than with electroanatomical mapping alone. However, these analyses addressed broader or methodologically distinct populations [12, 22, 23]. In contrast, the present study restricted comparative inference to patients with ICM or extractable ICM subgroups, incorporated both CT‐ and CMR‐guided strategies, and included contemporary randomized evidence.

A key implication is that “imaging‐guided” VT ablation should not be considered a single uniform intervention. Across included studies, imaging was used for pre‐procedural planning, electroanatomical map integration, target annotation, scar‐channel localization, and lesion registration [9, 10, 18, 19]. The benefit of imaging therefore likely depends not only on the modality used but also on how effectively imaging information is translated into procedural targets. The term ‘imaging‐guided’ was used for consistency with prior literature; however, in the included studies, pre‐procedural imaging did not simply supplement conventional maps but actively altered the procedural strategy, informing access route, focusing mapping density, and directing lesion delivery to imaging‐defined channels and crucial isthmuses. The intervention evaluated in this review is therefore better regarded as an imaging‐informed substrate modification strategy rather than imaging per se.

CMR may be most useful when detailed tissue characterization, scar‐channel analysis, or suspected intramural or epicardial substrate is clinically important [5, 6, 12, 24]. CT, on the other hand, offers high spatial resolution and robust anatomical modeling, allowing assessment of wall thinning, myocardial thickness heterogeneity, fatty metaplasia, calcification, thrombus, and anatomical relationships relevant to access strategy and ablation safety. In many ICD patients, CT may also be more practical and reproducible than CMR, particularly for anatomical modeling, wall‐thickness mapping, and procedural streamlining in post‐infarction VT. These differences suggest that CMR and CT should be viewed as complementary rather than interchangeable modalities [5, 6, 19, 20].

The randomized InEurHeart trial provides important contemporary support for CT‐guided workflows, showing shorter procedure duration with CT‐guided ablation while maintaining similar safety and numerically favorable efficacy [19]. CT‐based pilot and mechanistic studies also support the concept that wall‐thickness channels and anatomical corridors can identify regions relevant to VT circuits, although anatomical imaging alone does not replace functional electrophysiological confirmation [25, 26].

Similarly, CMR‐guided workflows can identify scar features and potential arrhythmogenic substrate that may be missed or incompletely sampled by endocardial voltage mapping alone [5, 22, 27]. Soto‐Iglesias et al. demonstrated that image‐based criteria can help identify epicardial arrhythmogenic substrate in patients with transmural myocardial infarction, supporting the clinical utility of imaging in selecting an endocardial versus endo‐epicardial strategy [27]. Likewise, studies examining post‐MI scar phenotype have shown that specific CMR scar characteristics are associated with subsequent ventricular arrhythmic events, reinforcing the concept that imaging‐derived substrate information is clinically meaningful beyond descriptive anatomy [22].

It should be noted, however, that the level of supporting evidence differs between modalities: the only randomized comparative study to date evaluated a CT‐guided workflow (Sacher [19]), whereas CMR‐guided approaches remain supported predominantly by observational data. This asymmetry should be considered when interpreting the clinical implications of each modality and does not, on its own, imply that CT‐guided ablation is inherently superior to CMR guidance [10, 18, 19].

Segmentation and integration platforms such as ADAS 3D and inHEART represent a practical bridge between imaging and ablation delivery. These tools convert CT or CMR datasets into procedural models that can be imported into mapping systems and used to localize scar, border zone, wall‐thickness channels, and potential conducting corridors [9, 19, 28]. Their clinical value depends on image quality, segmentation accuracy, registration reliability, operator experience, and how the resulting model is incorporated into the ablation strategy [21, 28]. The head‐to‐head comparison by Laan et al. suggested good spatial correspondence between imaging‐derived scar and EAM‐defined abnormal substrate, without demonstrating a universally superior platform, implying that different tools may have different strengths depending on imaging modality, substrate type, procedural goal, and institutional workflow [28].

The single‐arm imaging‐guided cohorts were supportive but not confirmatory. They reported favorable VT‐free or event‐free survival, acceptable safety, and generally high acute non‐inducibility but lacked contemporaneous conventional‐control groups and differed in endpoint definitions, imaging protocols, and follow‐up duration [24, 25, 29]. For this reason, they should be interpreted as evidence of feasibility and contemporary performance rather than proof of comparative efficacy.

The recurrence benefit observed in this review may reflect improved substrate targeting rather than uniform reductions in procedural duration or radiofrequency time. Imaging can identify scar channels and regions that may be incompletely recognized during initial invasive mapping, including buried, intramural, epicardial, or partially concealed substrate. In redo ablation, CMR‐defined prior lesion information and dark‐core assessment may help identify residual viable arrhythmogenic substrate after previous ablation failure [18]. Similarly, CT‐guided approaches may allow more deliberate ablation of anatomically defined targets and improve lesion completeness [19, 25].

Emerging non‐invasive mapping technologies may further strengthen anatomical‐functional integration. Non‐invasive electrocardiographic imaging (ECGi) has shown spatial relationships with arrhythmogenic substrate, imaging‐defined scar, VT exit regions, isthmus regions, and successful ablation sites, but current evidence remains observational and validation‐based [30]. This approach is conceptually attractive because it may add functional localization before invasive mapping begins, particularly in patients with poorly tolerated VT, multiple morphologies, or complex scar. Future work should determine whether combining pre‐procedural CT or CMR substrate imaging with ECGi‐based electrical localization can improve procedural efficiency, reduce mapping burden, or improve recurrence‐free survival [21, 31].

Clinically, imaging‐guided VT ablation may be most useful in complex ICM substrate, including redo procedures, extensive scar, suspected intramural or epicardial extension, and cases where conventional mapping may incompletely define the arrhythmogenic substrate [18, 19, 24]. The current evidence does not support viewing imaging as a replacement for expert EAM; instead, imaging should be interpreted as a tool that improves the precision and efficiency of an integrated substrate‐based strategy [6, 26]. By identifying scar channels, wall‐thinning corridors, and possible intramural or epicardial substrate before the procedure, imaging may allow a more targeted ablation strategy and reduce reliance on broad empirical substrate modification. Implementation remains dependent on local expertise, image quality, post‐processing capability, software access, EAM registration, and workflow familiarity, which might make the benefit observed in imaging‐integrated centers not reproducible across all other centers without appropriate infrastructure and training [21, 24].

Future research should define which imaging modality, workflow, segmentation platform, and level of procedural integration provide the greatest benefit for specific ICM substrate phenotypes. Larger multicenter randomized trials powered for clinical outcomes are needed to confirm the recurrence benefit observed here [12]. Future studies should also standardize definitions of imaging‐aided versus imaging‐guided ablation, compare CT‐, CMR‐, multimodality‐, ECGi‐assisted, and conventional workflows, and harmonize reporting of arrhythmia, safety, and procedural endpoints [13, 21]. Platform‐specific validation of ADAS 3D, inHEART, and related segmentation pipelines will also be important because the clinical value of imaging depends not only on acquisition but on reproducible conversion of image data into procedural targets [21, 28]. An additional frontier is post‐ablation LGE‐CMR, which may help assess lesion formation, scar evolution, channel modification, and mechanisms of recurrence, potentially informing both procedural planning and long‐term assessment of ablation durability as supported by the PAM‐VT 2 study [32].

This review has several strengths. First, comparative analyses were restricted to ICM populations or extractable ICM subgroups, reducing etiological and substrate heterogeneity compared with prior reviews that included mixed ischemic and non‐ischemic cohorts. Second, the analysis incorporated both CT‐ and CMR‐guided strategies, allowing assessment of contemporary imaging‐guided workflows rather than a single modality. Third, the inclusion of recent comparative evidence, including randomized data, provides an updated synthesis of a rapidly evolving field. Finally, the consistency of the recurrence benefit across sensitivity and subgroup analyses, despite differences in imaging modality, study design, and ablation context, supports the robustness of the observed association while acknowledging the moderate certainty of evidence and the limited number of comparative studies.

5. Limitations

This review has several limitations. First, only four comparative studies contributed to the primary meta‐analysis, limiting the precision of subgroup analyses and making assessment of publication bias underpowered. Second, most comparative evidence came from non‐randomized studies with residual risk of bias related to non‐random treatment allocation, potential unmeasured confounding, selection processes, baseline imbalance, operator experience, and variation in ablation strategy. Although the pooled estimate was precise and statistically consistent, the randomized trial alone was underpowered to establish a difference in VT recurrence. Third, imaging workflows differed across studies, including CT‐ and CMR‐based approaches, pre‐procedural planning, map integration, lesion registration, and different endocardial or endo‐epicardial strategies. This heterogeneity limits identification of a single optimal imaging modality or workflow. Fourth, some ICM subgroup outcomes were indirectly derived from published subgroup information or Kaplan–Meier curves, which are less certain than directly reported arm‐level data. Fifth, secondary outcomes, including acute non‐inducibility, ICD therapies, shocks, mortality, procedural duration, fluoroscopy time, radiofrequency time, and complications, were inconsistently reported and could not be pooled reliably. Finally, single‐arm imaging‐guided cohorts were summarized descriptively only because they lacked conventional‐control groups and differed in endpoint definitions and follow‐up duration.

6. Conclusion

In patients with ICM undergoing scar‐related VT ablation, CT‐ or CMR‐guided ablation was associated with lower VT recurrence than conventional ablation, without an apparent increase in major complications. However, evidence certainty remains limited by the small number of comparative studies, observational designs, and heterogeneous reporting, and these results should be interpreted as hypothesis‐generating rather than practice‐changing. The true magnitude of benefit remains uncertain until adequately powered randomized trials are available; larger multicenter randomized trials are needed to confirm this benefit and define optimal patient selection, imaging modality, and integration strategy.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Baseline characteristics. (A) comparative studies. (B) Single‐arm studies.

Table S2: Procedural characteristics. (A) Comparative studies. (B) single‐arm studies.

Table S3: Long‐term outcomes of single‐arm imaging‐guided cohorts (ICM patients) — described individually, not pooled.

Table S4: Subgroup and sensitivity analyses for the primary outcome (VT recurrence).

Table S4: GRADE summary of certainty of evidence for key outcomes.

Figure S1: VT recurrence—odds ratio, random‐effects (ICM).

Figure S2: Leave‐one‐out sensitivity analysis—VT recurrence (RR).

Figure S3: Subgroup: follow‐up ≥ 12 months (k = 3).

Figure S4: Subgroup: CMR‐guided ablation (LGE‐CMR; k = 2).

Figure S5: Subgroup: CT‐guided ablation (k = 2).

Figure S6: Subgroup: observational studies only (k = 3).

Figure S7: Sensitivity: excluding Ghannam 2022 (redo‐only; k = 3).

Figure S8: VT recurrence rate—imaging‐guided arm proportion (comparative studies; ICM).

Figure S9: VT recurrence rate—single‐arm studies (descriptive).

Figure S10: Funnel plot and Egger's regression (exploratory).

Figure S11: Risk of bias: ROBINS‐I and RoB 2 assessments.

Data Availability Statement

The data that supports the findings of this study are available in the Supporting Information of this article.