Independent Components Analysis (ICA) in the study of electroencephalographic signals
Urrestarazu E., Iriarte J.
Unidad de Epilepsia, Departamento de Neurología, Servicio de Neurofisiología, Clínica Universitaria de Navarra, Universidad de Navarra, Pamplona
Date: Jul 1, 2005Neurophysiology [SP]
Independent Component Analysis (ICA) is a mathematical tool able to separate complex signals in statistically independent components. It solves the blind source separation problem (BSS). The EEG satisfies most of the assumptions of ICA, so it may be an adequate signal for ICA for its use. In this paper we review the method and the applications of ICA in EEG. The studied applications are: a) artifacts removal; b) source estimation of spikes; c) analysis of seizures.
Several studies have demonstrated that ICA is useful to remove artifacts from contaminated EEG records without distorting cerebral activity. It is able to decompose epileptic discharges and seizures in independent spatio-temporal components. In combination with techniques of source localisation, as dipole modelling, ICA improves the localization of the epileptic focus. Finally, we discuss the future role of ICA in the study of epileptic patients.
CITATION Neurologia. 2005 Jul-Aug;20(6):299-310
you maybe interested
The Clínica is the greater private hospital with technological equipment of Spain, all in a single center.
The professionals of the Clínica perform continuous research and training, always to the benefit of the patient.
Learn why we are different from other healthcare centers. Quality, speed, comfort and results.