Hi NBS community


I wonder whether it is possible to somehow correlate (or use) behavioral measures into NBS to determine strength of structural connectivity between a set of self defined nodes.


Design:

- 2 Timepoints (Pre vs. Post) Therapy

- 14 Subjects with psychometric data

- DTI FA data processed into connectivity matrixes


Goal:

Somehow correlate (or test the influence of) the difference in a certain psychometrical scale to determine the differences in structural connectivity within a self-choosen 10 nodes based network between two time points.



Can one of you community members has a possible solution within NBS to test such an influence of psychometrical difference measure in a repeated measures design?


Thank you very much for your support

Hi,

You can specify any statistical design with the NBS that can be formulated in terms of the general linear model. This is the same as SPM, FSL, etc.

As a start, you might just want to identify connections for which your measure of connectvity is altered at follow-up, relative to baseline.

The following design will do that for you:


Example of a paired t-test (repeated measures) with 4 observations/subjects. The first four columns of the design matrix each model a within-subject mean. The final column models the between-group difference.
____________________________________
DESIGN MATRIX:
1 0 0 0 1
0 1 0 0 1
0 0 1 0 1
0 0 0 1 1
1 0 0 0 -1
0 1 0 0 -1
0 0 1 0 -1
0 0 0 1 -1

CONTRAST (Group 1 > Group 2):
0 0 0 0 1
STATISTICAL TEST:
t-test
EXCHANGE BLOCKS:
1 2 3 4 1 2 3 4



If you want to test with a behavioral scale. You probably want to provide NBS with the difference in connectivity between pre and post, and then simply test if this difference is correlated with you behavioral scale. The design matrix in this case is simply a columns of ones, followed by a column with your behavioural scale.

Andrew


Originally posted by Vi St:

Hi Andrew,

Thank you very much for your help.

I've applied the design that you mentioned for the correlation but a further question arised: 

I use difference correlation matrices (TP1-TP2) and a design matrix with 3 columns: 1.) Constant 2.) Age 3.) Difference in a behavioral measure (TP1-TP2)

The contrast file would be [0 0 1] for testing a positive correlation between the connectivity and the behavioral measure, would that be correct? For testing a negative correlation do I have to use -1 instead of 1 or do I have to multiple all correlation matrices with -1 and again use the same contrast file?

Beste wishes

Vivian

Please see next post....
Originally posted by Vi St: