Hi Nima
the text files t??.txt should contain #vols many values. There was a
buglet in melodic 2.0 beta which saved the FFT ot the timecourses twice
(as t??.txt and f??.txt) but that should be fixed FSL 3.1
ta
christian
On Monday, Oct 20, 2003, at 19:29 Europe/London, nima dehghani wrote:
> Hi Christine
> thanks for your reply and comments!
> one more question:
> a friend develeoped a new PCA analysis..we are trying to compare the
> results of this new method with PICA!
> in a part of such comparison, we need the "numerical data" of "t*.txt
> files" (numerical values of timecourse diagrams)! the problem is that
> if we have 224 scans, there are only 112 numbers! could you help me to
> understand why there are values only about "half of the scans"?
> best
>
> Nima
> Schizophrenia, Cognition and Imaging Labaratoy; Department of
> Psychiatry; UBC; CA
> Email:
> [log in to unmask], [log in to unmask]
>
> ----- Original Message -----
> From: "Christian F. Beckmann" <[log in to unmask]>
> Date: Wed, 15 Oct 2003 10:08:00 +0100
> To: [log in to unmask]
> Subject: Re: [FSL] ICA...Spatial smoothing
>
> Re: Hi
> Re:
> Re: On Wednesday, Oct 15, 2003, at 03:02 Europe/London, nima dehghani
> wrote:
> Re:
> Re: > Hi Christian!
> Re: > back to you again..and sorry for the delay!
> Re: >
> Re: > in reply to your reply to my second question:
> Re: >
> Re: > the best smooth is a filter that is of matched size to the
> activation
> Re: > we wish to detect. The anatomical variability between subjects
> will
> Re: > mean that the signal across subjects may be expected to be rather
> Re: > widely distributed over the cortical surface. In such a case it
> may be
> Re: > wiser to use a wide smoothing to detect this signal. In
> contrast, for
> Re: > a single subject experiment, it would be wiser to use a very
> narrow
> Re: > smoothing, or even no smoothing.....
> Re: >
> Re: > THIS IS FOR GLM ANALYSIS!..THE SAME LOGIC COULD ALSO BE ASCRIBED
> TO
> Re: > ICA (i think so,...maybe you do not agree with me)
> Re: >
> Re: > if we are using ICA:
> Re: > & if we are analyzing data of ONE subject, there is no prob; as
> we can
> Re: > do it with least gaussian (or even no guassian..as you pointed)
> Re: > smoothing!
> Re: > but if we need data across subjects; and we do the same, we will
> be in
> Re: > danger of lossing significant points!
> Re: > what is your idea?
> Re: >
> Re:
> Re: Again, the optimal procedure will depend on what you propose to do
> for
> Re: a multi-subject study; both w.r.t. to width of a smoothing kernel
> and
> Re: w.r.t. the question of individual vs. unique smoothing kernel.
> Re:
> Re: Let's assume for the moment that you would like to combine IC maps
> from
> Re: individual subjects. Then at the individual subject level the
> optimal
> Re: filter should match the expected activation size. Ideally, one
> could
> Re: employ some form of non-linear filter in order to preserve
> intensity
> Re: edges and to avoid 'blurring' across boundaries. As the 'true'
> Re: activation size is unknown it makes perfectly sense to use the same
> Re: smoothing width for each subject and match the width to one single
> Re: expected activation size. If you now want to combine individual
> maps at
> Re: some kind of group level, registration issues come into play. At
> this
> Re: point the 'optimal' thing to do is to get a really good
> registration in
> Re: order to avoid any kind of additional blurring. If registration is
> not
> Re: perfect, some smoothing could be used where now the smoothing
> kernel
> Re: width should best match the expected displacement of an individual
> Re: voxel across subjects.
> Re:
> Re: > and... what are the practical steps for what you called
> Re: > "concatenating"?
> Re: >
> Re: >
> Re:
> Re: the command line tool 'avwmerge' allows you to concatenate data in
> Re: x,y,z or t.
> Re:
> Re: ta
> Re: christian
>
>
>
> --
> __________________________________________________________
> Sign-up for your own personalized E-mail at Mail.com
> http://www.mail.com/?sr=signup
>
> CareerBuilder.com has over 400,000 jobs. Be smarter about your job
> search
> http://corp.mail.com/careers
|