Why the study?
Does a Laplace probability density function model improve atrial fibrillation detection from surface ECG compared to a Gaussian model?
Does a Laplace probability density function model improve atrial fibrillation detection from surface ECG compared to a Gaussian model?
Modeling RR interval variability with a Laplace probability density function improves automated atrial fibrillation detection from surface ECGs compared to a Gaussian model.
May enhance automated AF detection algorithms; leaves open prospective clinical validation before practice adoption.
We compare two Atrial Fibrillation (AFIB) detection methods from surface ECG, based on RR interval variability in a statistical framework. We obtain the histogram of normalized RR differences for AFIB and non-AFIB episodes using MIT-BIH Arrhythmia database. Two probability density functions (pdf) are employed to model the histograms: Gaussian and Laplace. We then use Neyman-Pearson (NP) detection approach to obtain criteria for AFIB detection. The performance of the two methods is compared using Receiver Operating Curves (ROC) over the different databases. The result shows that the Laplace pdf approximates the histogram of normalized RR differences better than the Gaussian pdf and leads to better AFIB detection performance.
No takes yet. Share an insight, caveat, or question.
Ghodrati et al. (2008) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: