[FieldTrip] Fwd: Question about unit of the wavelet analysis

[Marcello Venzi]

Hello,

I'm a first year PhD student in Neuroscience, so this is really my two cents...

1) If you sum two sine waves of the same amplitude but different frequency then you would expect the wavelet spectrum to show the higher-frequency peak smaller than the lower-frequency peak (while this would not happen with the Fourier power spectrum) . Have a look at 'Bias of the Global Wavelet Spectrum'

on http://paos.colorado.edu/research/wavelets/faq.html

for an explanation.

2) My limited understanding of the 1/f scaling in EEG/MEG is that there is no complete agreement on what its origin is. There are at two main hypothesis that I'm aware of :

-  The 1/f scaling is due to low-pass filtering of the extracellular medium. See this paper and references therein 

Bédard, C., & Destexhe, A. (2009). Macroscopic models of local field potentials and the apparent 1/f noise in brain activity. Biophysical journal, 96(7), 2589-603. doi:10.1016/j.bpj.2008.12.3951

- If instead the extracellular medium is purely resistive (as suggested in this Logothethis paper: Logothetis, N. K., Kayser, C., & Oeltermann, A. (2007). In vivo measurement of cortical impedance spectrum in monkeys: implications for signal propagation. Neuron, 55(5), 809-23. doi:10.1016/j.neuron.2007.07.027 ) the scaling could represent an organized phenomenon produced by neurons, ie. an aspect of neuronal computation.

I hope people in the mailing list will be so kind to correct me if this is just gibberish talk: I'm learning about this topics as I'm writing!

All the best,

Marcello Venzi

Marcello Venzi
PhD Student in Integrative Neuroscience
School of Biosciences
Cardiff University
+44(0)29 208 7515

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20120511/00f8dcf3/attachment-0002.html>