Key result
Deep learning model predicts OT-VA origins with ~89% accuracy, outperforming conventional algorithms.
Why the study?
Existing algorithms for differentiating RVOT and LVOT arrhythmia origins from 12-lead ECGs are complicated, prompting the development of a visualized deep learning model to classify arrhythmia origins more accurately.
Does a visualization deep learning model improve the accuracy of classifying ventricular arrhythmia origins from 12-lead ECGs compared to conventional algorithms in patients undergoing catheter ablation?
Observational (n=80)
Does a visualization deep learning model improve the accuracy of classifying ventricular arrhythmia origins from 12-lead ECGs compared to conventional algorithms in patients undergoing catheter ablation?
Absolute Event Rate: 89.4% vs 71.3%
p-value: p=<0.01
A novel visualization deep learning model using 12-lead ECGs accurately classifies the origin of ventricular arrhythmias (RVOT vs LVOT) and outperforms conventional algorithms while providing interpretability.
Supports DL for ECG-based OTVA localization; hypothesis-generating, requiring prospective validation before clinical adoption.
No takes yet. Share an insight, caveat, or question.
Nakasone et al. (2022) conducted an observational in Idiopathic outflow tract ventricular arrhythmias (OT-VAs) (n=80). Deep learning (DL) model vs. Conventional algorithm was evaluated on Accuracy of predicting the origin of OT-VAs (95% CI 88.2-90.6, p=<0.01). A visualized deep learning model classified the origins of outflow tract ventricular arrhythmias with an accuracy of 89.4%, significantly outperforming a conventional algorithm (71.3%).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: