Print

Print


See below.

On Mon, Sep 7, 2015 at 3:57 AM, NUML engg <[log in to unmask]> wrote:
Dear all,

I have a question regarding basis function. Normally we take canonical HRF as a basis function and some times temporal derivative and dispersion derivative also.

If I am taking Canonical HRF and I have two condition with realignment parameters then I have 9 regressors and I can define contrast as {1 -1 0 0 0 0 0 0] or [-1 1 0 0 0 0 0 0]

But if I have also temporal and dispersion derivative what will be my contrast (either 1 or 2 or none of them) with 6 regressors for 2 conditions and 6 regressors for realignment parameters.

1)  [1 1 1 -1 -1 -1   0 0 0 0 0 0]         and     [-1 -1 -1 1 1 1   0 0 0 0 0 0]
2)  [1 0 0 -1 0 0   0 0 0 0 0 0]   and    [-1 0 0 1 0 0   0 0 0 0 0 0]

You want to use contrast #2. Alternatively, you could read Vince Calhoun's paper (http://www.ncbi.nlm.nih.gov/pubmed/15110015) or Jason Steffener's paper (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2818488/) on boosting the HRF estimate based on the derivatives.

Contrast #1 doesn't make any sense. It is testing the difference in the sum/average of the HRF, temporal derivative, and dispersion derivative for each condition.
 

Please guide me
Thanks

Regards

NUML