Research Method Question

Question

I am a grad student( neuro) and taking an engineering course, my background is more cognitive/behavioral so I am learning about signals and systems, but some concepts remain elusive. For single neuron(unit) recording, why would someone chose to apply bandpass filter of 100Hz to 8kHz? Some context: I am reading a paper about a task which involves rhesus monkeys. Im aware that 200-300APs/second is the frequency to assess, but im not sure why the range is so large. I know band pass for electrical data from stim does high pass filter then amplification, then low pass filter, so im wondering if "100hz" is their high pass filter and "8kHz" is their low pass filter after amplification?

Answer 1

Hi Jade T.,

Your interpretation is correct. What you are attempting to do with the high pass is reject all spikes from quiescent neuron times (quiescent is a relative term!), amplify so that your counters will eventually work, and then also cut out any noise spike activity (which is at high-frequency) by the low-pass filter. I'm not sure why that is the preferred order of processing, but I assume that best satisfies the operating requirements (voltages, etc.) of the respective filters and signal amplifier.

Now, why might one choose the larger range, compared with what is thought optimal? Remember that neurons can have quite different excitabilities (latencies to the stimulus, frequency, duration, number of bursts) in a logical circuit -- but still be part of the processing. So they are looking to catch the largest possible set of participating neurons, perhaps.