Detecting the Intention to Move Upper Limbs from Electroencephalographic Brain Signals

Table 1

Description of the state of the art of works reporting decoding of motor information preceding actual movements from electroencephalographic brain signals. ME: motor execution; MI: motor imagery; MRCP: movement related cortical potentials; LDA: linear discriminant analysis.

	State-of-the-art movement intention detection from EEG
	Participants	Motor task	EEG features	Classifier	Decoding	Accuracy	Detection time
[27]	8: healthy	ME/MI	Spectral power	Bayesian	Static	Varied between subjects	—
[28]	4: stroke	ME	Spectral power	LDA	Static	—	From to  ms
[29]	6: healthy	ME	Spectral power	Bayesian	Static	%	—
[30]	19: healthy; 5: stroke	ME/MI	MRCPs	Template matching	Continuous	Healthy: ME: %, MI: %; stroke: ME: %	Healthy: ME:  ms
[31]	20: healthy; 5: stroke	ME/MI	MRCPs	Template matching	Continuous	Healthy: ME: %, MI: %; stroke: ME: %	—
[32]	10: healthy; 2: stroke	ME/MI	MRCPs	LDA	Continuous	—	ms
[33]	10: healthy; 2: stroke	ME	MRCPs	LDA	Continuous	Healthy: %; stroke: %	ms
[34]	20: healthy	ME	Common spatial patterns	LDA	Static	% in subjects	—
[35]	6: healthy; 3: SCI	ME	Spectral power and MRCPs	SDA	Continuous	From % to %	From  ms to  ms