Spectral Estimation of Non-Stationary EEG using Particle Filtering with Application to Event-Related Desynchronization (ERD)
This paper proposes non-Gaussian models for parametric spectral estimation with application to event-related desynchronization (ERD) estimation of nonstationary EEG. Existing approaches for time-varying spectral estimation use time-varying autoregressive (TVAR) state-space models with Gaussian state...
Saved in:
Main Authors: | , , , |
---|---|
Format: | Article |
Published: |
IEEE Explore
2010
|
Subjects: | |
Online Access: | http://eprints.utm.my/id/eprint/25989/ http://dx.doi.org/10.1109/TBME.2010.2088396 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
id |
my.utm.25989 |
---|---|
record_format |
eprints |
spelling |
my.utm.259892018-10-21T04:29:58Z http://eprints.utm.my/id/eprint/25989/ Spectral Estimation of Non-Stationary EEG using Particle Filtering with Application to Event-Related Desynchronization (ERD) Ting, Chee Ming Salleh, Sh. Hussain M. Zainuddin, Zaitul Bahar, Arifah Q Science (General) This paper proposes non-Gaussian models for parametric spectral estimation with application to event-related desynchronization (ERD) estimation of nonstationary EEG. Existing approaches for time-varying spectral estimation use time-varying autoregressive (TVAR) state-space models with Gaussian state noise. The parameter estimation is solved by a conventional Kalman filtering. This study uses non-Gaussian state noise to model autoregressive (AR) parameter variation with estimation by a Monte Carlo particle filter (PF). Use of non-Gaussian noise such as heavy-tailed distribution is motivated by its ability to track abrupt and smooth AR parameter changes, which are inadequately modeled by Gaussian models. Thus, more accurate spectral estimates and better ERD tracking can be obtained. This study further proposes a non-Gaussian state space formulation of time-varying autoregressive moving average (TVARMA) models to improve the spectral estimation. Simulation on TVAR process with abrupt parameter variation shows superior tracking performance of non-Gaussian models. Evaluation on motor-imagery EEG data shows that the non-Gaussian models provide more accurate detection of abrupt changes in alpha rhythm ERD. Among the proposed non-Gaussian models, TVARMA shows better spectral representations while maintaining reasonable good ERD tracking performance. IEEE Explore 2010 Article PeerReviewed Ting, Chee Ming and Salleh, Sh. Hussain and M. Zainuddin, Zaitul and Bahar, Arifah (2010) Spectral Estimation of Non-Stationary EEG using Particle Filtering with Application to Event-Related Desynchronization (ERD). IEEE Trans. On Biomedical Engineering, 58 (2). pp. 321-331. ISSN 0018-9294 http://dx.doi.org/10.1109/TBME.2010.2088396 10.1109/TBME.2010.2088396 |
institution |
Universiti Teknologi Malaysia |
building |
UTM Library |
collection |
Institutional Repository |
continent |
Asia |
country |
Malaysia |
content_provider |
Universiti Teknologi Malaysia |
content_source |
UTM Institutional Repository |
url_provider |
http://eprints.utm.my/ |
topic |
Q Science (General) |
spellingShingle |
Q Science (General) Ting, Chee Ming Salleh, Sh. Hussain M. Zainuddin, Zaitul Bahar, Arifah Spectral Estimation of Non-Stationary EEG using Particle Filtering with Application to Event-Related Desynchronization (ERD) |
description |
This paper proposes non-Gaussian models for parametric spectral estimation with application to event-related desynchronization (ERD) estimation of nonstationary EEG. Existing approaches for time-varying spectral estimation use time-varying autoregressive (TVAR) state-space models with Gaussian state noise. The parameter estimation is solved by a conventional Kalman filtering. This study uses non-Gaussian state noise to model autoregressive (AR) parameter variation with estimation by a Monte Carlo particle filter (PF). Use of non-Gaussian noise such as heavy-tailed distribution is motivated by its ability to track abrupt and smooth AR parameter changes, which are inadequately modeled by Gaussian models. Thus, more accurate spectral estimates and better ERD tracking can be obtained. This study further proposes a non-Gaussian state space formulation of time-varying autoregressive moving average (TVARMA) models to improve the spectral estimation. Simulation on TVAR process with abrupt parameter variation shows superior tracking performance of non-Gaussian models. Evaluation on motor-imagery EEG data shows that the non-Gaussian models provide more accurate detection of abrupt changes in alpha rhythm ERD. Among the proposed non-Gaussian models, TVARMA shows better spectral representations while maintaining reasonable good ERD tracking performance. |
format |
Article |
author |
Ting, Chee Ming Salleh, Sh. Hussain M. Zainuddin, Zaitul Bahar, Arifah |
author_facet |
Ting, Chee Ming Salleh, Sh. Hussain M. Zainuddin, Zaitul Bahar, Arifah |
author_sort |
Ting, Chee Ming |
title |
Spectral Estimation of Non-Stationary EEG using Particle Filtering with Application to Event-Related Desynchronization (ERD) |
title_short |
Spectral Estimation of Non-Stationary EEG using Particle Filtering with Application to Event-Related Desynchronization (ERD) |
title_full |
Spectral Estimation of Non-Stationary EEG using Particle Filtering with Application to Event-Related Desynchronization (ERD) |
title_fullStr |
Spectral Estimation of Non-Stationary EEG using Particle Filtering with Application to Event-Related Desynchronization (ERD) |
title_full_unstemmed |
Spectral Estimation of Non-Stationary EEG using Particle Filtering with Application to Event-Related Desynchronization (ERD) |
title_sort |
spectral estimation of non-stationary eeg using particle filtering with application to event-related desynchronization (erd) |
publisher |
IEEE Explore |
publishDate |
2010 |
url |
http://eprints.utm.my/id/eprint/25989/ http://dx.doi.org/10.1109/TBME.2010.2088396 |
_version_ |
1643647646052646912 |
score |
13.18916 |