[FieldTrip] Calculating ERPs with tf_timelockbaseline

Julian Keil julian.keil at gmail.com
Tue Apr 9 17:58:27 CEST 2019 

Dear Luca,

yes, this is what baselinecorrection in ERPs does: You take the mean value of a certain interval and subtract this value from all other values.
This will result in a linear shift. In other words, the waveforms will look identical, but shifted vertically.

Best,

Julian

> Am 09.04.2019 um 16:08 schrieb Luca Moretti <moretti.l1991 at gmail.com>:
> 
> Dear Julian,
> 
> I have tried to follow your useful suggestion: I have plot ERP and Color_C together. What I have found is that they generally don't overlap: however, the waveforms are still identical. The only difference is that the baseline corrected waveform starts (and of course ends) at more positive values. 
> 
> Now I am confused: Is this what baseline correction should do?
> 
> Best,
> Luca
> 
> Il giorno mar 9 apr 2019 alle ore 12:57 Julian Keil <julian.keil at gmail.com <mailto:julian.keil at gmail.com>> ha scritto:
> Dear Luca,
> 
> have you plotted both datasets in one figure?
> 
> e.g.
> 
> figure
> cfg = [];
> cfg.layout = layout;
> cfg.interactive = 'yes';
> cfg.showoutline = 'yes';
> ft_multiplotER(cfg, ERP, Color_C);
> 
> This way, you’ll see whether the two ERP traces actually overlap.
> However, if the average of your baseline is indeed „close to zero“ as you state, then the baselinecorrection will not have a huge impact. In other words, if there is nothing to subtract, the ERP will not change.
> 
> Best,
> 
> Julian
> 
> 
>> Am 09.04.2019 um 11:55 schrieb Luca Moretti <moretti.l1991 at gmail.com <mailto:moretti.l1991 at gmail.com>>:
>> 
>> Dear Julian,
>> 
>> I mean the second one: the signal does not change from Color_C to ERP. Inspecting Color_C as you suggested, it looks like there is a difference with ERP. However, ft_multiplotER seems to fail in plotting the ERP data. 
>> At first I also thought that maybe I simply couldn't see the differences because the avarage was close to 0 in my baseline, but I tend to exclude this case as the figures are Always identical even when specifying different baseline periods, when plotting different electrodes, and when analyzing different subjects. 
>> 
>> Thank you again,
>> Luca
>> 
>> Il giorno mar 9 apr 2019 alle ore 10:28 Julian Keil <julian.keil at gmail.com <mailto:julian.keil at gmail.com>> ha scritto:
>> Dear Luca,
>> 
>> could you be a bit more precise in your description?
>> By saying that "figure Always turn out the same for each electrode“, do you mean that the ERP is identical at each electrode?
>> In that case, I suggest going back a step and checking whether the signal at each electrode is identical. That would indicate a problem earlier in your pipeline, e.g. that you overwrite the single electrodes with a common signal.
>> 
>> Or do you mean, that the signal does not change from Color_C to ERP? 
>> In that case, I suggest inspecting Color_C in close detail, and computing the mean of the baseline interval by hand. If that value is close to zero, you can’t expect drastic changes from the baseline correction.
>> 
>> Hope that helps,
>> 
>> Julian
>>  
>> ________________
>> Prof. Dr. Julian Keil
>> 
>> Biological Psychology
>> Olshausenstrasse 62 - R. 306
>> 24118 Kiel, Germany
>> 
>> +49 - 0431 - 880 - 4872
>> http://www.biopsych.uni-kiel.de/en <http://www.biopsych.uni-kiel.de/en>
>> 
>> 
>>> Am 09.04.2019 um 10:11 schrieb Luca Moretti <moretti.l1991 at gmail.com <mailto:moretti.l1991 at gmail.com>>:
>>> 
>>> Dear Fieldtrippers,
>>> 
>>> I have a problem concerning the calculation of ERPs with ft_timelockbaseline. After I have preprocessed the data (re-referenced, filtered, cleaned from artifact), this is what I have done:
>>> 
>>> inputfile = (strcat('Subject_2_clean.mat'));
>>> load (inputfile);
>>> layout = 'easycapM1';
>>> 
>>> %% Calculating ERP
>>> 
>>> cfg = [];
>>> cfg.trials = find(data_clean_2.trialinfo==60);
>>> Color_C = ft_timelockanalysis(cfg, data_clean_2);
>>> 
>>> %% Plotting
>>> 
>>> cfg = [];
>>> cfg.baseline = [-.3 0];
>>> cfg.channel = {'all'};
>>> cfg.parameter = 'avg';
>>> ERP = ft_timelockbaseline(cfg, Color_C);
>>> 
>>> figure
>>> cfg = [];
>>> cfg.layout = layout;
>>> cfg.interactive = 'yes';
>>> cfg.showoutline = 'yes';
>>> ft_multiplotER(cfg, ERP)
>>> 
>>> %%
>>> 
>>> My problem is that it doesn't matter what kind of baseline period I am chosing, the figure Always turn out the same for each electrode. I have even tried to set the baseline to the Whole trial period, but the figure I get from plotting individual channels from the multiplot is Always the same.
>>> 
>>> Can someone find something wrong with my code?
>>> 
>>> Many thanks!
>>> Luca
>>> 
>>> _______________________________________________
>>> fieldtrip mailing list
>>> https://mailman.science.ru.nl/mailman/listinfo/fieldtrip <https://mailman.science.ru.nl/mailman/listinfo/fieldtrip>
>>> https://doi.org/10.1371/journal.pcbi.1002202 <https://doi.org/10.1371/journal.pcbi.1002202>
>> 
>> _______________________________________________
>> fieldtrip mailing list
>> https://mailman.science.ru.nl/mailman/listinfo/fieldtrip <https://mailman.science.ru.nl/mailman/listinfo/fieldtrip>
>> https://doi.org/10.1371/journal.pcbi.1002202 <https://doi.org/10.1371/journal.pcbi.1002202>
>> _______________________________________________
>> fieldtrip mailing list
>> https://mailman.science.ru.nl/mailman/listinfo/fieldtrip <https://mailman.science.ru.nl/mailman/listinfo/fieldtrip>
>> https://doi.org/10.1371/journal.pcbi.1002202 <https://doi.org/10.1371/journal.pcbi.1002202>
> 
> _______________________________________________
> fieldtrip mailing list
> https://mailman.science.ru.nl/mailman/listinfo/fieldtrip <https://mailman.science.ru.nl/mailman/listinfo/fieldtrip>
> https://doi.org/10.1371/journal.pcbi.1002202 <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

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20190409/622127a5/attachment-0002.html>
-------------- next part --------------
_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202

More information about the fieldtrip mailing list