Wavelet-Based Semblance Methods to Enhance the Single-Trial Detection of Event-Related Potentials for a BCI Spelling System

Algorithm 1

Multichannel EEG thresholding by similarity (METS).

(1)	Input: given the EEG signal matrix X, with C channels and T temporal samples
(2)	Output: the denoised signals
(3)	Set the correlation threshold ,
(4)	for to C do
(5)	Compute the Hilbert transform of
(6)	Compute the DWT of signal and of
(7)	end for
(8)	for to T do
(9)	Compute the MRL(t) using equation (6)
(10)	if then
(11)	set to zero at time t,
(12)	end if
(13)	end for
(14)	for to C do
(15)	Reconstruct the signal for channel c using the new
(16)	end for