In the following I hope to demystify some of the aspects of neurofeedback training with bipolar placements, or at least provide substantiation for some of the statements I have been making about it. We are concerned with reward-based training with narrow-band filters that select a particular training frequency. The bipolar montage feeds a differential amplifier that senses the difference in voltage waveforms appearing at the two sites. The signal that is common to the two sites cancels out in the amplifier circuit. This helps us with respect to cancellation of electronic interference that is common to both electrodes, but the same considerations apply to the signal itself. We only get to see the differential signal between the two sites.
For reasons that will ultimately become clear, I am particularly interested in the phase dependence of the signal, mostly because this has been missing from the discussion to date. How can we best illustrate this? One would like to have the same waveform, for example a sinusoidal waveform at the center frequency of the filter, appearing at both electrodes, and to see what happens as we shift the phase between the two signals. The resulting signal would be time-dependent, which cannot be easily shown in a static graphic in a newsletter. So we will do the next-best thing: We will simulate two slightly different frequencies at the two sites. They will therefore progressively move out of phase and then back into phase in yet another sinusoidal pattern, namely at the beat frequency or the difference frequency. This can be shown as a frozen image, as we have done in Figure 1. Continue reading “Understanding Bipolar Placement”