[Eeglablist] Baseline correction in ERSP

Scott Makeig smakeig at gmail.com
Fri May 28 06:40:55 PDT 2021


Hasan -

You don't say what operations you intend on the single trials, nor whether
you intend to use amplitude, power, or log power...?
I originally defined the ERSP using log(trial-mean-power) baseline
correction since because of phase instability the STFT can give ~0 power
values that convert to log(0) = -inf 'holes' in the trial-mean log-power
ERSP...
In your trial-by-trial analysis, could you take into account /make use of
trial baseline differences?  In this case, the above baseline correction
should be useful.

Scott Makeig

Makeig, S, Auditory event-related dynamics of the EEG spectrum and effects
of exposure to tones," <http://www.sccn.ucsd.edu/~scott/ersp93.html>
*Electroencephalogr
clin Neurophysiol*, 86:283-293, 1993.

On Thu, May 27, 2021 at 2:30 PM Mohammad Hasan Ghafari Saravi via
eeglablist <eeglablist at sccn.ucsd.edu> wrote:

> Dear EEGLAB experts,
> I have a question regarding baseline correction methods and their impact
> on the interpretation of results.
> Some brief background; we are measuring event-related spectral
> perturbation (ERSP) to measure the magnitude of responses to external
> periodic stimuli (in 80 trials) for each participant. In the 'newtimef'
> function in EEGLAB, there are several methods for baseline correction
> (e.g., classical baseline correction and single-trial baseline correction).
> However, after applying STFT to each trial, we need to perform
> single-trial (pre-stimulus) baseline correction due to additional analyses
> across single trials (instead of averaging the trials). As Grandchamp and
> Delorme (2011) mentioned in their study, single-trial (pre-stimulus)
> baseline correction is associated with a bias in the post-stimulus period.
> Is this baseline correction method the optimal way to measure the
> magnitude of responses? Does this bias affect further statistical analyses
> significantly in terms of Interpretability? Are there other alternatives I
> could try in EEGLAB?
> Your help and advice on this matter would be highly appreciated.
> Thanks and best regards,
> Hasan Ghaffari?
>
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to
> eeglablist-request at sccn.ucsd.edu
>


-- 
Scott Makeig, Research Scientist and Director, Swartz Center for
Computational Neuroscience, Institute for Neural Computation, University of
California San Diego, La Jolla CA 92093-0559, http://sccn.ucsd.edu/~scott


More information about the eeglablist mailing list