On Sun, 10 Dec 2006 14:55:41 +1100, Matthew Kirkcaldie
<m.kirkcaldie At removethis.unsw.edu.au> wrote:
>In article <1165701310.527446.233760 At n67g2000cwd.googlegroups.com>,
> "Marcin" <marcin.gierlicki At gmail.com> wrote:
>>> Hi,
>>>> I'm looking for informations about signals frequencies in brain
>> depending on area of it. For example range of frequencies for area
>> responsible for hearing or motor function. I need this information to
>> my master thesis about recordin extracellular signals from neuronal
>> cells.
>> Can someone help ?
>>>> Kind of animal dosen't matter ;)
>>>> Thanks,
>> Marcin
>>Maximum spike rate in neurons can be up to 2kHz - I believe there are
>phase-locked responses in cochlear afferents about this fast.
>>In cortex, 500-1000 spikes per second is the upper limit. Almost never
>see 2 action potentials close than 1ms apart, aside maybe from burst
>firing in a basket cell.
>>In cortical extracellular recording, we typically use bandpass filtering
>to include, say, 300Hz - 10kHz, and would generally sample at
>10kHz-20kHz depending on specific requirements (and yes, we know about
>Nyquist limits!). Including lower frequencies causes a lot of baseline
>shifts in potential to interfere with spike discrimination. Sadly the
>baseline shifts are probably very important to cortical function, we
>just don't know how!
>
An important point to add is that the recording bandwidth has to do
with the time course of a single impulse, not the frequency of
impulses. Although typical impulse frequencies are seldom as high as
these limits, usually below 100/sec, you still need the 10kHz
bandwidth to capture the pulse shape accurately..
