Dear Martin,
Peak-to-peak should be quite similar to just thresholding and the
threshold should be somewhere between 80 and 160. Adjacent samples
check is there to detect discontinuous jumps so the question is
whether there is such such a thing in your data and what is their
typical amplitude. In any case it will be less than 80 because
otherwise you'd detect them by thresholding.
Best,
Vladimir
On 5 Jan 2011, at 13:45, Martin Dietz wrote:
> Dear Vladimir & list,
>
> For EEG artefact detection I have been using simple thresholding at
> 80 microvolts which I think works very well in conjunction with
> robust averaging. However, I would like to compare overall
> thresholding with peak-to-peak and difference between adjacent
> samples thresholding.
>
> What would sensible thresholds be for the two latter ?
>
> Very best wishes,
> Martin
|