|Institution:||Curtin University of Technology|
|Department:||Electrical and Computer Engineering|
|Full text PDF:||http://espace.library.curtin.edu.au:80/R/?func=dbin-jump-full&object_id=204960&local_base=gen01-era02|
Event-related potentials (ERPs) are neural activities that are generated in response to an event. In recent decades, multi-channel EEG recordings have been increasingly adopted in ERP studies. In this thesis, to extract an ERP-of-interest, single-trial multi-channel extraction methods have been developed. The developed methods are data-driven and thus avoid human intervention during ERP analyses. For this reason, they are also practical for real-time applications such as brain-computer interface (BCI).