Last Updated on February 24, 2023 by pg@petergamma.org
Why not to combine 4 OpenBCI Cyton boards with WIFI sield to get a 32 channel EEG device and stream it to InfluxDB and use MNE Python software for data analysis, or two OpenBCI Cyton boards with two OpenBCI daisy boards to do the same?
Not all the parameters are available if we do it like this, phase and coherence are missing, if we do it like this, according to William Croft from OpenBCI. But do we need phase and coherence for our application, or can we do without it? This needs to be investigated.
We think this is worth to evaluate. The advantage of this method is, that we can build up multi sensor EEG devices from 8 or 16 channel components. Affordable clones for these components are available. If we build up multi-sensor EEG devices from these modules, costs for multi-sensor EEG devices trop dramatically. Often these clone components offer only a very limited warranty time. But if we build up these devices from individual 8 or 16 modules, the risk is lower, if a single module breaks, the costs are not that high to replace it, as for instance if we buy a 32 channel EEG device, which have also a very limited warranty time.
This could be a revolution for EEG devices as we have discussed here with up to 32 or 64 channels. The costs would trop, they would become easier to build, and risk is lower if single components fail.
According to William Croft from OpenBCI, a hardware change would be necessary to get all the parameters, such as phase and coherence. Is a simple cache storage enough to achieve this goal? A simple modification for instance of the OpenBCI Cyton board, to make it suitable to be used for InfluxDB applications, to support all the parameters?
This is a challenge for EEG soft- and hardware developers. But we think it is a challenge which might be interesting for an EEG engineer to develop something nobody ever has developed before.