Hi - it's easy to add confound EVs (covariates) in FEAT which would
be equivalent to what you're doing in SPM - does this not do what you
want?
Cheers.
On 18 May 2007, at 23:27, Rutger Goekoop wrote:
> Dear all,
>
> For some yet unexplained reason, a 0.5 Hz scanner artefact appears
> in our 800
> volume PRESTO SENSE timeseries (TR 0.609). This noise sums nearly
> linearly on
> top of existing BOLD signal intensity. Analysis shows that it can be
> eliminated during postprocessing using user-specified regressors in
> SPM5
> (with delta functions every 4th scan) as follows:
>
> Regressor1:
> 1 0 0 0 1 0 0 0 1 0 0 0 1 0 0 0 ... etc: complete 800 volumes.
> Regressor2:
> 0 1 0 0 0 1 0 0 0 1 0 0 0 1 0 0 ... etc: complete 800 volumes.
> Regressor3:
> 0 0 1 0 0 0 1 0 0 0 1 0 0 0 1 0 ... etc: complete 800 volumes.
>
> This yields 3 PE images for these 3 regressors with almost 100% noise
> elimination with good residual SD.
>
> Since I want to use MELODIC for further study of these data (and
> since I'm
> forever hooked to FSL), I'd like to do something similar in FSL. I
> tried
> MELODIC denoising, however MELODIC does not find the relevant
> components.
> Instead, nearly all components are contaminated with the 0.5 Hz
> artefact
> (possibly because of its regular nature?). Thus, I think I may
> first need to
> do a model-based analysis in all individuals (as in SPM5, see
> above) using
> custom EVs with delta functions, and then somehow filter out the
> beta images
> from the filtered_func_data (using avwmaths?). Or is there perhaps
> a more
> sound way of dealing with this problem?
>
> Any help would be greatly appreciated,
>
> Cheers,
>
> Rutger Goekoop.
------------------------------------------------------------------------
---
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
------------------------------------------------------------------------
---
|