Whereas some functionally meaningful components help explain the contribution of specific topographic activity patterns in the ERP time course (Makeig et al. 1999a,b), one or more of the ICA components that are not related to brain processes (i.e., ocular, cardiac, and muscular artifacts) can be removed (Iriarte et al. 2003). ICA algorithms have been used to identify topographic patterns of ERPs associated with specific diagnostic categories, such as mild cognitive impairment (Li et al. 2013; Missonnier et al. 2013b) and voluntary hypoxic state (Menicucci et al. 2013). In alcoholism research, Olbrich and colleagues (2002) studied ICA-derived spatial components of ERPs in a visual CNV paradigm and found increases in the ICA components of N2 and negative slow waves as well as decreases in P3 in alcoholics compared with control subjects. Evidence suggests that ICA is becoming a useful signal processing method for analyzing electrophysiological data, and may become an important tool in alcoholism research as well.
