Please use this identifier to cite or link to this item:
|Title:||Applying ICA to single trial auditory P300 and CNV evoked potentials to provide biomarkers|
Camilleri, Kenneth P.
Fabri, Simon G.
Linden, David E. J.
Independent component analysis
Evoked potentials (Electrophysiology)
|Citation:||Jervis, B. W., Belal, S. Y., Camilleri, K., Cassar, T., Fabri, S., Linden, D. E. J., ... Besleaga, M. (2007). Applying ICA to single trial auditory P300 and CNV evoked potentials to provide biomarkers. 3rd International Conference on Computational Intelligence in Medicine and Healthcare (CIMED 2007).|
|Abstract:||The Independent Components (ICs) of P300 and CNV Evoked Potentials (EPs) have been found on a truly single trial basis using Independent Component Analysis (ICA) and cluster analysis. P300 data was obtained from healthy participants and patients with Alzheimer’s disease (AD), while Contingent Negative Variation (CNV) data came from a healthy subject. The ICs found corresponded to the spatiotemporal profiles reported in the literature, which is based mainly upon averaged waveforms. Some positively and negatively peaking components of the P300 ICs were found to coincide with known peaks in the averaged P300. However, there were trial-to-trial differences where a particular IC might occur or not, and where a positive or a negative component could occur with the same latency. These variations would be obscured by the standard averaging practice in EP analysis. The mean latencies of the ICs differed between healthy participants and AD patients, particularly in the IC associated with the P3b peak (p<0.01). In the case of the CNV, our approach correctly identified both the orienting and the expectancy components, even though a short 1s interstimulus interval was used.|
|Appears in Collections:||Scholarly Works - FacEngSCE|
Files in This Item:
|Conference paper||280.56 kB||Adobe PDF||View/Open Request a copy|
Items in OAR@UM are protected by copyright, with all rights reserved, unless otherwise indicated.