Automated differentiation of wide QRS complex tachycardia using QRS complex polarity

Automated differentiation of wide QRS complex tachycardia using QRS complex polarity

Abstract Background Wide QRS complex tachycardia (WCT) differentiation into ventricular tachycardia (VT) and supraventricular wide complex tachycardia (SWCT) remains challenging despite numerous 12-lead electrocardiogram (ECG) criteria and algorithms. Automated solutions leveraging computerized ECG...

Full description

Saved in:
Bibliographic Details
Main Authors: Adam M. May, Bhavesh B. Katbamna, Preet A. Shaikh, Sarah LoCoco, Elena Deych, Ruiwen Zhou, Lei Liu, Krasimira M. Mikhova, Rugheed Ghadban, Phillip S. Cuculich, Daniel H. Cooper, Thomas M. Maddox, Peter A. Noseworthy, Anthony Kashou
Format: Article
Language:English
Published: Nature Portfolio 2024-12-01
Series:Communications Medicine
Online Access:https://doi.org/10.1038/s43856-024-00725-2
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Abstract Background Wide QRS complex tachycardia (WCT) differentiation into ventricular tachycardia (VT) and supraventricular wide complex tachycardia (SWCT) remains challenging despite numerous 12-lead electrocardiogram (ECG) criteria and algorithms. Automated solutions leveraging computerized ECG interpretation (CEI) measurements and engineered features offer practical ways to improve diagnostic accuracy. We propose automated algorithms based on (i) WCT QRS polarity direction (WCT Polarity Code [WCT-PC]) and (ii) QRS polarity shifts between WCT and baseline ECGs (QRS Polarity Shift [QRS-PS]). Methods In a three-part study, we derive and validate machine learning (ML) models—logistic regression (LR), artificial neural network (ANN), Random Forests (RF), support vector machine (SVM), and ensemble learning (EL)—using engineered (WCT-PC and QRS-PS) and previously established WCT differentiation features. Part 1 uses WCT ECG measurements alone, Part 2 pairs WCT and baseline ECG features, and Part 3 combines all features used in Parts 1 and 2 Results Among 235 WCT patients (158 SWCT, 77 VT), 103 had gold standard diagnoses. Part 1 models achieved AUCs of 0.86–0.88 using WCT ECG features alone. Part 2 improved accuracy with paired ECGs (AUCs 0.90–0.93). Part 3 showed variable results (AUC 0.72–0.93), with RF and SVM performing best. Conclusions Incorporating engineered parameters related to QRS polarity direction and shifts can yield effective WCT differentiation, presenting a promising approach for automated CEI algorithms.
ISSN:2730-664X

Similar Items