Print

Print
Hi Joeri,

Please see below:


On Thu, 7 Nov 2019 at 07:41, SUBSCRIBE FSL Joeri van Helden <[log in to unmask]> wrote:
Dear all,

I have a rather complex design and wanted to know whether my reasoning and design are right.

DTI data of a patient group (25 subjects), before and after surgery (removal of tumour). I also obtained cognitive test scores from pre- and post surgery. They will be inserted separately and also together.

Question: Relation between pre- and post surgical changes in FA and cognitive test scores

Sounds like this could be rephrased as: are changes in FA associated with changes in cognitive test scores?

If this is right, i.e., it is what you'd like to investigate, then the design below is fine, but not the contrasts. See below...
 

Followed this link: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/GLM#Single-Group_Paired_Difference_.28Paired_T-Test.29
I explicitly set the exchangability group values for randomise.

5 example patients:

Number of EVs: 8 ( last two are cognitive test scores)
Number of additional, voxel-dependent EVs: 0

Score 1: a transformed index score
Score 2: amount of omitted targets.

Group        EV1       EV2      EV3     EV4     EV5     EV6    score1        score2

1                  1            1           0          0          0         0          -0.23           0     
2                  1            0           1          0          0         0           0.37           7
3                  1            0           0          1          0         0           0.21           2
4                  1            0           0          0          1         0          -0.01           0
5                  1            0           0          0          0         1          -0.03           10
1                 -1            1           0          0          0         0           0.27           4
2                 -1            0           1          0          0         0           0.02           7
3                 -1            0           0          1          0         0          -0.36           0
4                 -1            0           0          0          1         0          -0.28           2
5                 -1            0           0          0          0         1           0.02           9

Contrasts: 3
F-tests: 0

            Title            EV1       EV2     EV3    EV4   EV5   EV6   score1    score2
C1     Pre-post1         1            0          0         0       0        0          1            0
C2     Pre-post2         1            0          0         0       0        0          0            1
C2     Pre-post1&2    1             0          0         0       0        0          1            1


Consider instead:

C1     Pre-post1         0            0          0         0       0        0          1            0
C2     Pre-post2         0            0          0         0       0        0          0            1

Don't run C3 (which btw was type as C2) as it mixes different things and isn't meaningful. Instead, consider an F-test of C1 and C2.

You can also test the negative versions of C1 and C2 (the F-test is always two-tailed).
 
Then run:

randomise -i PairedT4D -o PairedT -d design.mat -t design.con -e design.grp -m mask -n 10000 -T -x

This is fine, though maybe some extra power can be obtained by using permutations not only within-block but also between-blocks. This isn't yet available in randomise but can be done with PALM (option -within together with option -whole).

 

Am I on the right path with regard to my reasoning and design?

Yes!

All the best,

Anderson

 

Best wishes,

Joeri

########################################################################

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1