Abstract
AbstractThe “virtual electrode” is an appealing metaphor as it suggests that we can – with the right tools and skills – use macroscopic recording techniques such as EEG or MEG to place an electrode anywhere in the brain, and record from it as if it were a real, microscopic electrode. Like every good metaphor, it can sometimes lead us astray. This short paper reviews what we can, and cannot, expect of a “virtual electrode”. It is suggested that a more accurate metaphor – if less appealing – is that of a “virtual scalpel” by which activity from a source within the brain can be suppressed within the macroscopic recording. WithJchannels, up toJ− 1 sources can be suppressed while still retaining some information about other sources. In addition, with appropriate source and forward models, it may be possible to accurately localize any source that can be accurately suppressed. This paper sets out the possibilities and limits of what linear analysis applied to multichannel data can, or cannot, achieve.
Publisher
Cold Spring Harbor Laboratory