Łukasz Lenart
Łukasz Lenart
Krakow University of Economics
Date(s) : 09/02/2026 iCal
14h00 - 17h00
Bayesian non-Gaussian autoregressive model with outliers for cardiac signals
Anna E. Dudek (AGH University of Krakow, Aix Marseille University)
Łukasz Lenart (Krakow University of Economics)
Wu & H. Ombao (King Abdullah University of Science and Technology)
This presentation presents a novel model for analyzing electrocardiogram (ECG) time series data. The particular focus here is on the RR-interval, which is the time elapsed between two successive R waves of the QRS signal on the ECG. A non-Gaussian autoregressive (AR) model is developed to analyze about 25 minutes of the RR time series recorded from 3 clinical groups: controls, congestive heart failure patients, and atrial fibrillation patients. Outliers in the signals were identified through the latent variables of the model. A suitable MCMC sampler was applied and validated in a simulation study.
The simulation studies and analysis of the ECG signal demonstrate that the proposed approach (in particular, the AR structure with latent variables) is flexible in modeling the complex structure of RR segments, while providing the additional benefit of being able to filter out outliers.
Emplacement
Saint-Charles - FRUMAM (2ème étage)
Catégories
