Hello FSL users

Can someone please help me with this error message:
Problem with processing the model: Warning: at least one EV is (close
to) a linear combination of the others. You probably need to alter your
design. (Design matrix is rank deficient - ratio of min:max eigenvalues
is SVD of matrix is 8.88555e-34)

I got this error message when I tried to View Design after setting up my
model of 3 EVs, for each EV I entered the following values:
Off = 16, On = 40, Phase = 16. I chose a square basic shape and left
convolution as it is in the default settings (Gamma). Where did I go
wrong?

Many thanks in advance for your help.

Kind regards
Reem

On Wed, 2010-03-31 at 19:11 +0100, Stephen Smith wrote:
> Hi
> 
> On 29 Mar 2010, at 02:18, Reem Jan wrote:
> 
> > Dear FSL experts
> > 
> > I am a brand new FSL user. I run a blocked Stroop task with fMRI,
> > and I
> > have previously analysed some of my data using SPM. I am trying to
> > learn
> > using FEAT and unfortunately I'm getting stuck at very preliminary
> > stages.. Can someone please help?
> > I'll explain my paradigm and what I have been entering in the GUI so
> > far..
> > 
> > My TR is 3 seconds
> > I have 118 functional images per subject.
> > The blocked Stroop task I use is set up as follows:
> > Rest (16sec) - Neutral condition (40sec) - Rest (16sec) - Congruent
> > condition (40sec) - Rest (16sec) - Incongruent condition (40sec) -
> > Rest
> > (16sec) - Incongruent condition (40sec) - Congruent condition
> > (40sec) -
> > Rest (16sec) - Neutral condition (40sec)
> > 
> > I have several questions:
> > 1) Does that make my design rA-rB-rC-rC-rB-rA?
> > 
> 
> 
> Yes; you won't explicitly model the rest conditions - you will have
> EV1=A EV2=B EV3=C
> 
> > 
> > 2) Is my 'high pass filter cutoff' = r+A+r+B+r+C (16+40+16+40+16+40)
> > =
> > 168sec?
> > 
> 
> 
> It depends on your contrasts - have a read of the Smith NeuroImage
> paper on efficiency, and a play with the efficiency estimator in FEAT.
> 
> > 3) My data is acquired with interleaved order but for a blocked
> > design
> > like mine, do I need to use slice time correction? (I know I don't
> > tick
> > that option in SPM).
> > 
> 
> 
> No, we generally recommend against that.
> 
> > 4) I struggled with setting up a full model despite reading the
> > instructions on FSL webpage. According to the information I supplied
> > above, do I have 3 or 6 EVs? as I have a repetition of each
> > condition as
> > described above.
> > 
> 
> 
> You have 3 "original" EVs, which by default will get turned into 6
> "real" EVs once the temporal derivatives are added by FEAT for you.
> 
> 
> Cheers.
> 
> 
> 
> 
> 
> > 
> > I greatly appreciate your time and advice.
> > 
> > Kind regards
> > Reem
> > 
> > 
> 
> 
> ---------------------------------------------------------------------------
> 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
> ---------------------------------------------------------------------------
> 
> 
> 
> 
> 
> 
>