Hi Patrick,
that is the 100.000 $ question ;) I believe that it will depend on too many things as to be easily answered (for example, resting state vs. task-based, clustered vs. single timepoints, other sources of noise, and ultimately [although of course not strictly scientific] "precious patient" vs. "one of many control subjects"). If you want to try the Afyouni & Nichols approach, you can play around with my optimized censoring toolbox which allows you to identify outliers and proceed with an interpolated timeseries. Download is free at
https://www.medizin.uni-tuebingen.de/kinder/en/research/neuroimaging/software/
Cheers
Marko
--
____________________________________________________
Prof. Dr. med. Marko Wilke
Facharzt für Kinder- & Jugendmedizin, Neuropädiater
Leiter, Experimentelle Pädiatrische Neurobildgebung
Geschäftsführender Oberarzt der Abteilung Neuropädiatrie
Universitäts-Kinderklinik
Marko Wilke, MD, PhD
Pediatrician and Pediatric Neurologist
Head, Experimental Pediatric Neuroimaging
Senior Consultant in Pediatric Neurology
University Children's Hospital
Hoppe-Seyler-Str. 1
D - 72076 Tübingen, Germany
Tel. +49 7071 29-83416
Fax +49 7071 29-5473
[log in to unmask]
http://www.medizin.uni-tuebingen.de/kinder/epn/
____________________________________________________
>>> Patrick <[log in to unmask]> 16.08.19 19.56 Uhr >>>
Dear SPM Experts,
Could someone advise me on some guidelines that might be used to decide how many volumes to mark as motion outliers before deciding to remove the subject altogether from an analyses. For example, Afyouni and Nichols (2018) have given an updated definition of DVARS...how many volumes before I should throw the subject out? Similarly, Parkes et al (2018) have recently suggested thresholds of 0.2 and 0.5mm using framewise displacement. Is there some literature and/or rule of thumb to say how many volumes to scrub?
Thanks a lot!
P.S: My data has 140 time points
BestPB
|