ICA is a blind signal separation technique that is widely used to decompose the contributions of neurogenic and artifactual sources of EEG activity (Delorme et al., 2007, 2012).
Neurogenic independent components (ICs) have been shown to arise from synchronous activity in distinct cortical and subcortical patches, despite ICA being blind to electrode locations and source conduction patterns (Głąbska et al., 2014).
The source anatomy of neurogenic EEG ICs can be accurately modeled using one or two (bilateral) equivalent dipoles. Moreover, equivalent dipole fitting of IC scalp projections often results in lower residual variance compared to fitting non-decomposed data (e.g. scalp topography of ERP components or condition contrasts), suggesting markedly improved localization accuracy (Delorme et al., 2012).
Accurate source location may require EEG systems with at least 128 sensors, as the spatial Nyquist rate of the EEG signal is thought to be 128 points (Srinivasan et al., 1998). Increasing sensor count generally increases localization accuracy.
In my experience with this pipeline (using recordings from CMU's 128ch BioSemi), equivalent dipole fitting of neurogenic ICs generally results in model residual variance under 10%, sometimes under 1%.
Figure from Onton & Makeig (2006):
Fifteen seconds of EEG data at 9 (of 100) scalp channels (top panel) plus contributing activities of 9 (of 100) independent components (ICs, bottom panel) extracted from the whole session data. While nearby electrodes (upper panel) record highly similar mixtures of brain and non-brain activities, ICA component activities (lower panel) are temporally distinct (i.e., maximally independent over time), even when their scalp maps are overlapping. Compare, for example, IC1 and IC3, accounting for different phases of eye blink artifacts produced by this subject after each visual letter presentation (grey background) and ensuing auditory performance feedback signal (colored lines). Compare, also, IC4 and IC7, which account for overlapping frontal (4-8 Hz) theta band activities appearing during a stretch of correct performance (seconds 7 through 15). Typical ECG and EMG artifact ICs (IC12, IC55) are also shown, as well as overlapping posterior (8-12 Hz) alpha band bursts (IC5, IC9�8) that appear when the subject waits for the next letter presentation (white background) For comparison, the repeated average visual event-related potential of a bilateral occipital IC process (IC5) time locked to the visual stimulus is repeated (in red) on the same (relative) scale as the component activity. Clearly the unaveraged activity dynamics of this IC process are not well summarized by its averaged response a dramatic illustration of the independence of phase-locked and phase-incoherent activity.