[FieldTrip] Problem with ft_megrealign

Kirsten Herfurth kirsten at carrot-village.de
Thu Nov 9 12:58:02 CET 2017


Dear Jan-Mathijs,

thank you very much for your answer! I will fix the bug and commit it to Github. 
Do you have any suggestions for an alternative approach for realigning data obtained with a non-CTF-system (we have a 4d-MEG-system)? Unfortunately we don’t have a head position time-series obtained online, just the position in the beginning and in the end of a run. I assume I cannot use the function ft_regressconfound or ft_headmovement.
I’m sorry for this basic question, but I haven’t found a satisfying answer to my problem yet. The reason for accounting for head movement is that I would like to use the Machine Learning toolbox and do statistics using k-fold cross-validation, but have too few trials in one run per subject. So I thought to concatenate the trials of different runs with ft_appenddata using the realignment function first. Would it be ok to just go without the realignment in subjects that didn’t move that much, especially if it gives me the expected results? 

Thanks again and best wishes

Kirsten




More information about the fieldtrip mailing list