Dear all,
First I would like to thank again Simo for his answer. The method
described by Buchel et al. (1998) seems very useful.
But, secondly, help is needed on how to run the orthogonal polynomial
expansion described by Buchel. There has been several studies citing
the method but no further explanation is given. I thought the default
SPM modulation could use Buchel's method but after testing some
contrasts it doesn't seem so.
The method in a few words:
As I've understood it runs polynomials step by step starting from
zero-th order. Each higher order a covariate is used for the
variability already explained by the previous order. This way one can
find areas that has a general linear trend at the beginning but also a
nonlinear trend on the next order regressor. Apparently this is not
possible with the actual modulation in SPM which splits drastically
areas that are explained ONLY by linear or ONLY by nonlinear trends.
In case I am wrong please correct me.
Any help on how to do it is welcome.
Best Regards.
Dorian
2009/2/17 Simo Vanni <[log in to unmask]>:
> Dear Dorian,
>
> We had a similar case in Pihlaja et al. Human Brain Mapping 29 (2008)
> 1001-1014. We used the method of Buchel et al, Neuroimage 8 (1998) 140-148,
> and it worked fine.
>
> best regards,
> Simo Vanni
>
> Dorian P wrote:
>>
>> Is there a way to detect non-linear increasing in a parametric
>> modulation with 2nd order modulation?
>>
>> The contrast of the 2nd order regressor outputs indifferently
>> increases or decreases that have slight non-linearities in U shape (1
>> contrast) or inverted U shape (-1 contrast, examples attached). I need
>> only those areas that has a non-linear component, but also an overall
>> increase between 1 and 4 in 1-2-3-4 modulation. I am playing around
>> with F-contrasts but it doesn't seem to be possible.
>>
>> Thank you.
>>
>> P.s. The effects may be reversed in the graphs due to "negative"
>> activation.
>>
>> Dorian
>> Ruhr University Bochum
>> PhD Student
>>
>> ------------------------------------------------------------------------
>>
>>
>> ------------------------------------------------------------------------
>>
>
> --
> ************************************************************
> Simo Vanni, MD PhD,
> Academy Research Fellow
>
> Vision Systems Physiology
> Brain Research Unit/AMI centre
> Low Temperature Laboratory
> Helsinki University of Technology
> PO BOX 3000, 02015-TKK, Espoo
> Finland
>
> tel: +358-9-451 6159
> fax: +358-9-451 2969
> e-mail: [log in to unmask]
> http://ltl.tkk.fi/wiki/BRU/Vision_Systems_Physiology
> ************************************************************
>
>
|