Key result
Beats alignment in time and amplitude achieved a sensitivity of 0.96 and specificity of 0.52 for detecting ventricular late potentials, compared to 0.72 and 0.40 without alignment.
Why the study?
Does time-amplitude alignment using derivative dynamic time-warping improve the detection of ventricular late potentials in high-resolution ECGs compared to standard averaging?
Does time-amplitude alignment using derivative dynamic time-warping improve the detection of ventricular late potentials in high-resolution ECGs compared to standard averaging?
Time-amplitude alignment using derivative dynamic time-warping improves the sensitivity and specificity of detecting ventricular late potentials on high-resolution ECG.
May enhance ventricular late potential detection on high-resolution ECG; hypothesis-generating and requires validation before adoption.
Ventricular late potentials (VLPs) are small-amplitude waves with a short duration that appear at the end part of the QRS complex, making a QRS complex duration larger. The signal-averaged electrocardiography (ECG) technique enhances VLPs and beats, assuming noise as the only random variable. However, ECG signals are not completely stationary and different elongations appear in both time and amplitude in each beat. This research proposes to use piecewise linear approximation to segment each beat and performs the alignment of the beats using the technique known as derivative dynamic time-warping to have beats better aligned and consequently enhance the presence of VLPs. We recorded high-resolution ECGs (HRECGs) from 50 subjects in supine position with no heart-stroke antecedents. VLPs were created synthetically and added to the HRECGs. Two cases were evaluated: (i) duration of the QRS complexes with VLPs without beats alignment, and (ii) duration of QRS complexes with VLPs using beats alignment in time and amplitude. Considering QRS duration as an indicative of VLP presence, results show that when using beats alignment in time and amplitude it is possible to reach a sensitivity of 0.96 and a specificity of 0.52, as opposed to 0.72 and 0.40, respectively, when using only averaging without beats alignment in time and amplitude.
No takes yet. Share an insight, caveat, or question.
Avitia et al. (2013) studied No heart-stroke antecedents (n=50). Beats alignment in time and amplitude using derivative dynamic time-warping vs. Averaging without beats alignment in time and amplitude was evaluated on Sensitivity of detecting ventricular late potentials based on QRS duration. Beats alignment in time and amplitude achieved a sensitivity of 0.96 and specificity of 0.52 for detecting ventricular late potentials, compared to 0.72 and 0.40 without alignment.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: