Charles Darwin University

CDU eSpace
Institutional Repository

 
CDU Staff and Student only
 

Online artifact recognition and removal in EEG signals through wavelet transform and independent component analysis

Richardson, Greg (2014). Online artifact recognition and removal in EEG signals through wavelet transform and independent component analysis. Bachelor of Engineering (4th Year Project) Thesis, Charles Darwin University.

Document type: Thesis
Citation counts: Google Scholar Search Google Scholar

Author Richardson, Greg
Title Online artifact recognition and removal in EEG signals through wavelet transform and independent component analysis
Institution Charles Darwin University
Publication Date 2014
Thesis Type Bachelor of Engineering (4th Year Project)
Subjects 0906 - Electrical and Electronic Engineering
Abstract Electroencephalographic (EEG) signals are signals generated from electrical activity in the brain; these are recorded via a number of electrodes placed across the scalp of a patient. In a Brain Computer Interface (BCI) this technology is utilized and the output data is processed online. EEG signals are highly susceptible to artifact contamination from signals generated by other muscles in the body. These artifacts can often lie in the same frequency range as the brain signals being monitored and, as such, can make readings worthless, as such, these artifacts should be isolated and removed from the signal.

This method uses a combination of the FastICA algorithm and wavelet transforms. First the artifact is observed by identifying a sharp increase in amplitude in the signal (while negating drift). The artifact is then isolated using the FastICA algorithm. Using the relevant independent component (IC), the artifact is recognised by running wavelet transforms using a number of mother wavelets (relating to different potential artifacts). When identified the IC is replaced with a zero vector and the signal is regenerated by performing an inverse FastICA. Artifacts could also be considered features; as such, when a relevant waveform is identified, an output or external function may be triggered prior to removal.
Keyword Artifact
Brain Computer Interface (BCI)
EEG
Wavelet
Independent Component Analysis (ICA)
Feature Recognition
Online
Additional Notes Full text restricted.
 
Versions
Version Filter Type
Access Statistics: 62 Abstract Views, 333 File Downloads  -  Detailed Statistics
Created: Fri, 24 Apr 2015, 10:49:42 CST by Jessie Ng