Why the study?
Does a threshold detector operating on the 0-4 Hz band effectively detect single-trial P300 signals?
Does a threshold detector operating on the 0-4 Hz band effectively detect single-trial P300 signals?
Answers are grounded in this paper and cite their sources.
A threshold detector for single-trial P300 detection achieved a ~70% detection rate, suggesting brain-machine interfaces can use short interstimulus intervals and higher target probabilities to increase throughput.
A threshold detector for single-trial P300 detection has been evaluated. The detector operates on the 0-4 Hz band, isolated from the raw electroencephalogram using low-pass filtering, wavelet transforms, or the piecewise prony method (PPM). A detection rate around 70% was found, irregardless of stimulus type, interstimulus interval (ISI), probability of occurrence (Pr) of the target stimuli, intrasession and intersession effects, or filtering method. This suggests that P300-based brain-machine interfaces can use an ISI as short as 1 s and a Pr of 45%, to increase throughput.
No takes yet. Share an insight, caveat, or question.
Jansen et al. (2004) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: