Related Experiment Video
Updated: Jul 7, 2026
08:10
Estimating Bilateral Atrial Function by Cardiovascular Magnetic Resonance Feature Tracking in Patients with Paroxysmal Atrial Fibrillation
Published on: July 20, 2022
Frequency tracking of atrial fibrillation using hidden Markov models
Frida Sandberg1, Martin Stridh, Leif Sörnmo
1Department of Electrical and Information Technology, Box 118, Lund University, Lund SE-22100, Sweden. frida.sandberg@eit.lth.se
IEEE Transactions on Bio-Medical Engineering
|February 14, 2008
Summary
A hidden Markov model (HMM) significantly enhances noise robustness for tracking atrial fibrillation (AF) dominant frequencies in electrocardiograms (ECGs). This method substantially reduces frequency tracking errors, improving diagnostic accuracy even in noisy signals.
Area of Science:
- Biomedical Engineering
- Signal Processing
- Cardiology
Background:
- Atrial fibrillation (AF) detection and monitoring rely on accurate analysis of electrocardiogram (ECG) signals.
- Tracking the dominant frequency in AF is crucial for understanding disease dynamics and guiding treatment.
- Existing methods often struggle with noise robustness, limiting their clinical applicability.
Purpose of the Study:
- To develop and evaluate a noise-robust method for tracking the dominant frequency of AF in ECG signals.
- To leverage hidden Markov models (HMMs) to improve the accuracy of frequency estimation in the presence of noise.
Main Methods:
- A hidden Markov model (HMM) was applied to ECG signals after QRST cancellation.
- Frequency states were observed using the short-time Fourier transform on the residual ECG.
- The Viterbi algorithm was used to find the optimal state sequence, integrating AF characteristics and signal-to-noise ratio (SNR) information.
Main Results:
- The HMM-based tracking method demonstrated significant improvement in noise robustness.
- Root-mean-square (rms) error in frequency tracking was substantially reduced.
- At a 4-dB SNR, the rms error decreased from 0.2 Hz to 0.04 Hz, a fivefold improvement.
Conclusions:
- Hidden Markov models offer a powerful approach to enhance noise robustness in AF dominant frequency tracking.
- The proposed method significantly improves the accuracy of frequency estimation in noisy ECG signals.
- This technique holds promise for more reliable AF monitoring and analysis in clinical settings.
