Independent Component Analysis

Definition: Independent component analysis (ICA) is a blind source separation method which separates a multivariate data (e.g., fNIRS signal) into statistically independent components which are assumed to be statistically independent from each other and have a non-gaussian distribution. Different types of ICA are available (e.g. linear and non-linear ICAs). The removal of physiological noise (such as heart rate and respiration) and motion artifacts, as well as the identification of neuronal activity sources, have been demonstrated to be effectively achievable in fNIRS data using ICA.

Alternative definition:

Synonym:

References:

https://doi.org/10.3389/fninf.2021.683735

https://doi.org/10.1016/0165-1684%2894%2990029-9

https://doi.org/10.1109/embc.2013.6611118

https://doi.org/10.1098/rsta.2011.0534

https://doi.org/10.1063/1.4812785

https://doi.org/10.1016/j.neuroimage.2010.02.080

Related terms: Principal Component Analysis, Cocktail Party Problem, Central limit theorem, noise removal, functional connectivity