HRX 2024 AbstracX Library-Machine Learning Algorithm for QTc Measurement and AAD Dosing Recommendation sin Long QT mECG Dataset

Machine Learning Algorithm for QTc Measurement and AAD Dosing Recommendation sin Long QT mECG Dataset

Pdf Summary

The document discusses a machine learning (ML) algorithm developed to improve the measurement of QTc intervals and recommend antiarrhythmic drug (AAD) dosages using a Long QT mECG dataset. This research is supported by the National Heart, Lung, and Blood Institute (NHLBI) with a disclosure noting that several authors are employed by SafeBeat Rx Inc. and have received equity compensation.<br /><br />Globally, atrial fibrillation (AF) impacts 43 million individuals, with a predicted 12.1 million cases in the U.S. by 2030. Despite these figures, only 4% of AF patients undergo ablation. The 2023 Heart Rhythm Society (HRS) guidelines suggest early rhythm control within one year of diagnosis, yet there is no existing software for streamlined AAD titration or monitoring, highlighting a gap in AF care.<br /><br />The study presents a comparison between the ML algorithm and manual measurements by electrophysiologists (EP). Although there were minor differences in the R-R interval and QTc measurements, the ML algorithm was highly predictive of EP dosing decisions, achieving similar or more conservative dosing recommendations in 95.2% of sotalol and 94.2% of dofetilide cases compared to EP cardiologists.<br /><br />This ML-based software not only matches EP decisions for QTc measurement and AAD dosing but also has potential applications beyond diagnostics to influence treatment decisions. The work represents a significant step towards more streamlined AAD workflows in managing AF patients.

Keywords

machine learning

QTc intervals

antiarrhythmic drug

Long QT mECG dataset

atrial fibrillation

Heart Rhythm Society

AAD titration

electrophysiologists

sotalol

dofetilide