Hi Daniela,
Answers inline with questions below:
> On 29 Jul 2022, at 14:00, Daniela de Albuquerque <[log in to unmask]> wrote:
>
> Hi there,
>
> Sorry for posting this again (as a separate Q), but since I got no reply since Wed, I thought I should just try posting it again. Based on Matthew's answer to my original question, I had a couple of f/u questions/clarifications:
>
> 1) The de-meaning piece Matthew mentioned - is mean here computed across all time pts, per voxel? (or across space?) And then subtracted off of each vol/time pt in series? And is this done just to filtered data or also to DM in some form?
Yes - the demeaning is across time, on a per-voxel basis. The design should already be demeaned ( this is carried out by the feat_model command ).
>
> 2) I might want/need to implement something similar to FSL's pre-whitening routine myself. Is there any reference you can give me with pseudocode for this algorithm? I tried reading the featmodel.cc and prewritten.cc scripts to get a sense for it, but am a bit rusty on my C/C++ and would really appreciate something like a pseudocode for it that I could use as base to implement it in other languages. I also read the Woolrich 2001 paper, which talks about the pre-whitening used in FSL but I don't think it goes into details of the algorithm itself.
I’m afraid we don’t really have any pseudocode/documentation beyond the C++ code itself and the original paper/report, so some C++ translation will be needed to reimplement.
>
> 3)And what exactly (math wise) does FSL do under the label of ``temporal filtering"? I originally assumed this referred to removing temporal autocorrelation using pre-whitening, but based on what you said, seems like these are different steps?
The temporal filtering is Gaussian based ( Gaussian-weighted least-squares straight line fitting for the highpass, simple Gaussian convolution for the low pass ) and occurs as part of pre-processing before the FILM call.
>
> Thank you so much for the help and again sorry for the re-posting!
> Daniela
>
> ########################################################################
>
> To unsubscribe from the FSL list, click the following link:
> https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=FSL&A=1
>
> This message was issued to members of www.jiscmail.ac.uk/FSL, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
Hope this helps,
Kind Regards
Matthew
--------------------------------
Dr Matthew Webster
FMRIB Centre
John Radcliffe Hospital
University of Oxford
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=FSL&A=1
This message was issued to members of www.jiscmail.ac.uk/FSL, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
|