Key result
ECG-based K-means clustering algorithm distinguishes overdrive excited from suppressed ventricular beats with ~93% sensitivity.
Answers are grounded in this paper and cite their sources.
A novel ECG-based K-means clustering algorithm using normalized recovery cycle length and QRS duration can accurately distinguish between overdrive suppressed (automatic) and overdrive excited (triggered) ventricular beats in a preclinical model.
Rapid ventricular pacing rates induces two types of beats following pacing cessation: recovery cycle length (RCL) prolongation (overdrive suppression) and RCL shortening (overdrive excitation). The goals of this study were to compare common experimental protocols for studying triggered activity in whole-heart preparations and differentiate between recovery beats using a new methodology. Post-pacing recovery beat cycle length (RCL) and QRS were normalized to pre-paced R-R and QRS intervals and analyzed using a K-means clustering algorithm. Control hearts only produced suppressed beats: RCL ratio increased with rapid pacing (25 ± 4.0%, n = 10) without changing QRS duration. Rapid pacing during hypercalcemia + hypothermia (5.5 mM and 34°C) produced significantly earlier excited beats (53 ± 14%, n = 5) with wider QRS durations (58 ± 6.3%, n = 5) than suppressed beats. Digoxin + hypothermia (0.75 μM) produced the most excited beats with significantly earlier RCL (44 ± 3.2%, n = 6) and wider QRS (60 ± 3.1%, n = 6) ratios relative to suppressed beats. Increasing pacing further shortened RCL (30 ± 7.8%, n = 6). In a prospective study, TTX (100 nM) increased RCL ratio (15 ± 6.0%, n = 10) without changing the QRS duration of excited beats. The algorithm was compared to a cross-correlation analysis with 93% sensitivity and 94% specificity. This ECG based algorithm distinguishes between triggered and automatic activity.
No takes yet. Share an insight, caveat, or question.
Greer-Short et al. (2015) studied Ventricular arrhythmias (experimental model) (n=37). Rapid ventricular pacing and pharmacological interventions (digoxin, hypercalcemia, TTX) vs. Control (normal Tyrode solution, 37°C, slow pacing) was evaluated on Sensitivity and specificity of K-means clustering algorithm to distinguish overdrive excited vs suppressed beats. An ECG-based K-means clustering algorithm using normalized recovery cycle length and QRS duration distinguished between overdrive excited and suppressed ventricular beats with 93% sensitivity and 94% specificity.
Synapse has enriched 10 related papers. Consider them for comparative context: