[FieldTrip] Computing average power within a band following ft_freqanalysis
acskwara at ucdavis.edu
Thu Feb 13 02:03:25 CET 2020
I am doing an IAF-based analysis that requires me to extract average power
(or PSD) within frequency bands (e.g., alpha, beta) for subsequent
comparison. This means that I need an estimate in the form of a single
number for the whole range of frequencies subsumed within each frequency
My code when processing my data using ft_freqanalysis is as follows:
cfg.method = 'mtmfft';
cfg.output = 'pow';
cfg.taper = 'hanning';
cfg.channel = 'all';
cfg.foilim = [lowerlim upperlim]; %of the band of interest
cfg.trials = 'all';
cfg.keeptrials = 'no';
freq = ft_freqanalysis(cfg_band, data_segmented);
The resulting output is a struct containing a 2D matrix (powspctrm) with
the dimensions of 36 frequencies (in this case ranging from 12.5 to 30,
counting by .5) and 73 channels, with each cell containing a power estimate.
My questions are:
1) What are the units of what is contained in each cell, using the "pow"
output option? Are my units mV2 (power) or mV2/Hz (PSD)?
2) What is the most appropriate way to go from these frequency-wise
estimates to an estimate of band power (e.g., alpha power or beta power)?
If the units are in PSD, my understanding is that it would be appropriate
to take the simple mathematical average to calculate power within the band.
If they are in mV2, I am unclear what the appropriate approach would be.
Using the bandpower function in matlab? Or does FieldTrip have a built in
function to calculate power within a band that I am somehow not finding?
Thanks in advance for your help,
Alea C. Skwara, M.A.
PhD Candidate | Saron Lab
Center for Mind and Brain | University of California, Davis
267 Cousteau Place | Davis CA 95616
Cell: (828) 273-8595
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the fieldtrip