Generally this requires a large number of channels, evenly spread across the scalp. QEEG / Loreta neurofeedback uses the 19 channels of the 10-20 map. Research groups typically use more channels: 32, 64, even more. All of these setups require an EEG amplifier that supports the required number of channels, and a headset / cap system that matches.
I don't have the exact formula off the top of my head, but as the channel count goes down, the spatial resolution drops quickly. Which is the number and size of 'voxels' volume-pixels that are produced by the localization. The 4 channels of the Ganglion will not be enough. Even 8 or 16 channels are way below the numbers normally used. Although 16 might be on the edge.
Many EEG analysis systems can compute this. The most famous being *Loreta variations. MNE and NeuroPype support localization. Try a search,
Section 2.4, Table 1 in this paper has a good summary of localization error distances for various numbers of channels, from 256 all the way down to 16.
Clinical Neuroscience
EEG source localization: Sensor density and head surface coverage
Comments