[FieldTrip] Choice of ft_freqanalysis baseline type

Jeremy Harper harpe300 at umn.edu
Tue Sep 28 19:06:28 CEST 2021


Dear FieldTrip list members,

There are several options for time-frequency baseline adjustment in ft_freqbaseline (e.g., absolute difference, relative change, dB). 

My question is: when would one choose to use the ‘normchange’ option ([activity - baseline] / [activity + baseline]) over other options? I’m having some trouble understanding in what situations (or with what sort of TF data) this approach would be preferred over the more common (and perhaps more intuitive?) decibel or relative percentage change baseline adjustments.

I’ve only seen the normchange approach used in one recent article (https://doi.org/10.1016/j.bandl.2020.104894 <https://doi.org/10.1016/j.bandl.2020.104894>) and a recent FieldTrip tutorial (https://www.fieldtriptoolbox.org/tutorial/sleep/ <https://www.fieldtriptoolbox.org/tutorial/sleep/>). The closest explanation of the normchange metric I can find is not in EEG work but rather spectral analysis in vegetation science (https://en.wikipedia.org/wiki/Normalized_difference_vegetation_index#Rationale <https://en.wikipedia.org/wiki/Normalized_difference_vegetation_index#Rationale>), and I can’t recall seeing this approach coded in other common EEG processing toolboxes.

Thank you,
Jeremy


--
Jeremy Harper, PhD

NIDA T32 Postdoctoral Fellow
Department of Psychiatry | University of Minnesota
Minnesota Center for Twin and Family Research

ORCID: https://orcid.org/0000-0001-8994-6879 <https://orcid.org/0000-0001-8994-6879>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20210928/793bd8fb/attachment.htm>


More information about the fieldtrip mailing list