[FieldTrip] Optimal settings for ft_artifact_zvalue for detecting electrode pops

[Alain de Cheveigne]

Does anyone know how those "pops" are produced?  Their sharp onset and slow decay suggest a first-order high-pass filter, but that is not consistent with the DC characteristics of the device. 

My guess is that the basic phenomenon is a step-like change in electrode offset, similar to a SQUID jump, and that the device compensates for it to keep in range, according to some non-linear and/or switching mechanism. The slow decay would reflect that mechanism. If so, it would be nice to know more about it.  

Any insights?

Alain



> On 5 Feb 2024, at 08:19, Schoffelen, J.M. (Jan Mathijs) via fieldtrip <fieldtrip at science.ru.nl> wrote:
> 
> Hi Jan,
> 
> I think that the morphology of an electrode pop looks a bit like a (low amplitude) eye blink (but then occurring at an isolated electrode, not necessarily a frontal one). In that context, I would imagine that a sequence of processing steps that highlight eyeblinks might also detect the larger pops.
> 
> Good luck,
> Jan-Mathijs
> 
> 
>> On 30 Jan 2024, at 12:19, Jan Ostrowski via fieldtrip <fieldtrip at science.ru.nl> wrote:
>> 
>> Dear Fieldtrippers,
>> 
>> I was wondering whether you could provide some advice on the optimal configuration for ft_artifact_zvalue function so that it might more or less reliably detect electrode pop artifacts. When I talk about electrode pops, I mean those artifacts with a very sharp, high amplitude onset and a subsequent gradual decay. (We use EEG, not MEG)
>> 
>> I have been trying the configuration for detecting SQUID jumps as suggested on the corresponding tutorial page, but I realized SQUID jumps look quite different than electrode pops, especially after the initial sharp onset. 
>> 
>> My approach is to use ft_artifact_zvalue to generate sample markers for the artifacts, and then use ft_rejectartifact with the preprocessed data loaded into the workspace to remove those trials.
>> 
>> This has been my configuration so far:
>> 
>> % +++ Jump artifact config +++
>> cfgJump = [];
>> cfgJump.continuous   = 'no';
>> % channel selection, cutoff and padding
>> cfgJump.artfctdef.zvalue.channel = target_chn;
>> cfgJump.artfctdef.zvalue.cutoff  = 60;
>> cfgJump.artfctdef.zvalue.trlpadding = 0;
>> cfgJump.artfctdef.zvalue.artpadding = 0;
>> cfgJump.artfctdef.zvalue.fltpadding = 0;
>> % algorithmic parameters
>> cfgJump.artfctdef.zvalue.cumulative = 'yes';
>> cfgJump.artfctdef.zvalue.medianfilter = 'yes';
>> cfgJump.artfctdef.zvalue.medianfiltord = 9;
>> cfgJump.artfctdef.zvalue.absdiff = 'yes';
>> 
>> I appreciate any help on this!
>> 
>> Cheers,
>> Jan Ostrowski
>> 
>> 
>> -- 
>> Jan Ostrowski, M.Sc.
>> Department of Systems Neuroscience
>> University Medical Center Hamburg-Eppendorf
>> Martinistr. 52
>> Building W34, Room 320b
>> D-20246 Hamburg, Germany
>> 
>> 
>> Universitätsklinikum Hamburg-Eppendorf; Körperschaft des öffentlichen Rechts; Gerichtsstand: Hamburg | www.uke.de
>> Vorstandsmitglieder: Prof. Dr. Christian Gerloff (Vorsitzender), Joachim Prölß, Prof. Dr. Blanche Schwappach-Pignataro, Matthias Waldmann (komm.)
>> 
>> SAVE PAPER - THINK BEFORE PRINTING
>> 
>> _______________________________________________
>> fieldtrip mailing list
>> https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
>> https://doi.org/10.1371/journal.pcbi.1002202
> 
> _______________________________________________
> fieldtrip mailing list
> https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
> https://doi.org/10.1371/journal.pcbi.1002202