What does this research mean for the field?

A machine learning model improves lesion metric predictions in radiofrequency catheter ablation (RFCA), achieving r² values of 0.87 for depth and 0.86 for volume, outpacing conventional methods. Novelty: ClaimNovelty.NOVEL_FINDING. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

The aim is to develop a machine learning model that enhances accuracy in estimating lesion metrics during radiofrequency catheter ablation.

February 8, 2026

Accurate estimation of lesion metrics in radiofrequency ablation using machine-learning model

Key Result

The machine learning model predicted lesion depth (r²=0.87), length (r²=0.82), volume (r²=0.86), and surface area (r²=0.69) more accurately than conventional metrics in RFCA.

Key Points

The aim is to develop a machine learning model that enhances accuracy in estimating lesion metrics during radiofrequency catheter ablation.
Applied RF energies (30-50W) to excised ventricular myocardium using RFCA.
Used contact forces of 10g or 20g for durations between 10 and 180 seconds.
Evaluated correlations between total ablation energy, force-time integral, impedance-drop, and lesion metrics against machine learning predictions.
Employed eXtreme Gradient Boosting (XGBoost) for predictive modeling with a dataset split of 75% for training and 25% for validation.
Assessed feature importance for each lesion metric.
Analyzed a total of 1,142 ablations.
Total ablation energy showed strong correlations with max depth (r² = 0.63), max length (r² = 0.50), and volume (r² = 0.69).
Machine learning model predictions yielded high accuracy: r² = 0.87 for depth, 0.82 for length, 0.86 for volume, and 0.69 for surface area.
Impedance drop was notably linked to surface area with r² = 0.48.
Total ablation energy and duration were key predictors, enhancing accuracy over traditional metrics.

Structured PICO

Does a machine learning model using multiple RFCA parameters improve the accuracy of lesion metric estimation compared to conventional parameters in excised ventricular myocardium?

Population

Excised ventricular myocardium (n=1,142 ablations)

Intervention

Machine learning model (eXtreme Gradient Boosting - XGBoost) using multiple RFCA parameters (RF energies 30-50W, contact forces 10g or 20g, durations 10-180 seconds, perpendicular and parallel catheter orientations)

Comparator

Conventional single parameters (total ablation energy, force-time integral, impedance-drop)

Outcome

Accuracy of lesion metric estimation (max depth, max length, volume, surface area) measured by r²surrogate

A machine learning model using multiple RFCA parameters significantly improves the prediction of ablation lesion metrics compared to conventional single parameters in an ex vivo model.

Main Result

Absolute Event Rate: 0% vs 0%

Abstract

Abstract Aims Conventional parameters for estimating lesions in radiofrequency catheter ablation (RFCA), such as ablation energy (AE), contact force, and impedance variation, often yield suboptimal results. This study aimed to develop a machine learning model to improve the accuracy of lesion metric estimation in RFCA. Methods RF energies (30-50W) were applied to excised ventricular myocardium using RFCA with contact forces of 10g or 20g for durations between 10 and 180 seconds, with both perpendicular and parallel catheter orientations. Correlations between total AE, force-time integral, impedance-drop, and lesion metrics were evaluated and compared to machine learning model predictions, using eXtreme Gradient Boosting (XGBoost). The dataset was split for training (75%) and validation (25%). Feature importance for each lesion metric was also assessed. Results A total of 1,142 ablations were analyzed. Total AE had the strongest correlation with max depth, max length, and volume (r² = 0.63, 0.50, 0.69), followed by force-time integral (r² = 0.54, 0.45, 0.62) and impedance-drop (r² = 0.33, 0.45, 0.31). Impedance drop was most strongly associated with surface area (r² = 0.48). The machine learning model accurately predicted lesion metrics: r² = 0.87 for depth, 0.82 for length, 0.86 for volume, and 0.69 for surface area, with low root mean square error values. Total AE and ablation duration were key predictors, with impedance drop contributing more to surface area and length predictions. Conclusions Machine learning using multiple RFCA parameters improves lesion metric predictions, enhancing lesion estimation beyond conventional metrics, potentially improving procedural guidance and safety.

Bookmark