2
\$\begingroup\$

Currently, all 'wireless' EEG systems acquire signals from electrodes which are wired to an amplifier, a filter, a digitizer and then finally a wireless module.

I want to design a system that's completely wireless, maybe using wireless sensor networks. Is it possible to transmit data wirelessly (RF or bluetooth), from the electrode itself, to a station where the amplification/filtering would be done?

I do realize the raw EEG signal from a single electrode would have a LOT of noise.

\$\endgroup\$
  • 1
    \$\begingroup\$ There is no way you're going to digitize and transmit a raw EEG signal without amplification at least. It's such a small signal that it it would take a ton of bits, of which most would be noise. \$\endgroup\$ – Matt Young Jun 20 '14 at 19:21
  • \$\begingroup\$ Depending on the nature of the raw signal then this is entirely possible. Although apparently it would be best to amplify this signal prior to transmitting it. You are essentially only moving part of the logic diagram to an earlier stage in the global process. You only have to be aware of the affects that wireless transmission may impose on your signal (i.e. noise and the like) that would have otherwise not be there at that point in the process. \$\endgroup\$ – sherrellbc Jun 20 '14 at 19:25
6
\$\begingroup\$

You can't transmit the voltage of an electrode, because that does not exist: a voltage is always between two points, so you need at least two electrodes connected to your transmitter. In practice you need three to cancel out the 'common mode' noise.

To transmit the signal, you must first filter it (to remove the ever-present 50 or 60 Hz), amplify it (to bring it in the range of the A/D converter), the convert it to digital and finally transmit it. This is doable (for someone who can design such circuits, which is far from trivial), but there is not that much advantage over a wired version.

| improve this answer | |
\$\endgroup\$
3
\$\begingroup\$

Assuming you would use analogue techniques (as the signal will not be digital until you sample it), the short answer is no, you would need to amplify first to get a big enough signal with a reasonable signal-to-noise ratio. That said, a gain stage and 24 bit A/D converter for 8 channels won't be much bigger than the electrode itself. For example, TI makes a number of analog front ends for ECG which might be suitable. Add a small wireless module capable of interfacing with the A/D and you're good to go.

| improve this answer | |
\$\endgroup\$
0
\$\begingroup\$

This falls into the category of "blue-sky" thinking, but you might be able to wireless transmit a EEG signal via a RFID style backscatter. The trick would be modulating an interrogator transmission with the EEG signal.

One method of doing so would be to translate the EEG signal into a variable impedance load on an antenna, doing so changes the radar cross-section (RCS) of the antenna in relation to the load; variations in the RCS could be seen at the integrator. Another idea that comes to mind would be use the interrogator and EEG signal in a sub-harmonic style mixer where the EEG signal would be modulated onto a frequency two times or one-half the interrogator frequency.

With all these methods I suspect that system noise would be a problem. The low-level EEG signal may end up being covered by the noise.

| improve this answer | |
\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.