Lingfei,
> Thanks for the paper. We do have a composite measure of motion from
> Artifact Reduction Toolbox, and we screened participants scans based
> on composite measures rather than separately by
> translational/rotational measures. But I was thinking to report some
> stats separately for each motion parameter to show the average and
> age effects. Perhaps instead I should report the stats for the
> composite measure?
Without knowing which composite measure you used, I can't say whether it
is more meaningful, but if you used this measure to exclude subjects I
think this is also what you should report.
> It is just I feel that the values of composite
> measures are less interpretable.
I actually think they are more interpretable: if I know how much a point
at a given distance from the origin moved *as a combination of all 6
parameters*, I think this is more tangible, and more meaningful, than
looking at an individual radian value which may or may not add up with
one of the other 5 parameters.
> Are there some guidelines/rule of
> thumb on how to interpret the magnitude of these composite measures?
In resting state fMRI, people have become increasingly wary of even
small motion and use .5 mm translation between scans as an exclusion
criterion. Depending on your cohort, task, question, approach etc. this
may be very strict, though.
Cheers,
Marko
> ----- Original Message ----- From: "Marko Wilke"
> <[log in to unmask]> To: "Lingfei Tang"
> <[log in to unmask]>, [log in to unmask] Sent: Monday, June 8, 2015
> 11:48:28 AM Subject: Re: [SPM] Motion parameter summary
>
> Lingfei,
>
>> When summarizing the motion parameters, I am wondering how I should
>> average. Say I have 3 runs, each run having a rp file, should I
>> average within per run and then average across 3 runs.
>
> I am not completely sure what you mean, and why you want to do it...
> but assuming you want to describe the amount of motion per subject,
> per session, I would like to suggest to not only rely on these
> parameters to assess subject motion. I humbly suggest you take a look
> at a recent concoction of mine at
> http://www.ncbi.nlm.nih.gov/pubmed/25333359 for the reasons why I
> think this is not an excellent idea.
>
> I may have misunderstood your question, though, as averaging motion
> across sessions does not seem to make much sense to me (each session
> is either acceptable, or not - a bad one will not outweigh a good
> one, or vice versa).
>
>> The first line of the rp files are always all 0s and each line is
>> against the first scan. When averaging, should I do a subtraction
>> to get the scan-to-scan movement, or should I just average from 2
>> to last line within each rp file?
>
> The first line is zeros as motion is "corrected for" by realigning
> all subsequent scans to the first one. Therefore, by definition the
> first scan did not move (this is also true if you select "register to
> mean", as explained at
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=SPM;5dfedb11.0801)
>
> Cheers, Marko
>
--
____________________________________________________
PD Dr. med. Marko Wilke
Facharzt für Kinder- und Jugendmedizin
Leiter, Experimentelle Pädiatrische Neurobildgebung
Universitäts-Kinderklinik
Abt. III (Neuropädiatrie)
Marko Wilke, MD, PhD
Pediatrician
Head, Experimental Pediatric Neuroimaging
University Children's Hospital
Dept. III (Pediatric Neurology)
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/
____________________________________________________
|