Print

Print


Hi,

> 1- Is the order of input DTI in TBSS analysis as follows?
> patient_1_exam_number_one
> patient_2_exam_number_one
> patient_3_exam_number_one
> patient_4_exam_number_one
> patient_5_exam_number_one
> patient_6_exam_number_one
> patient_7_exam_number_one
> patient_8_exam_number_one
> 
> patient_1_exam_number_two
> patient_2_exam_number_two
> patient_3_exam_number_two
> patient_4_exam_number_two
> patient_5_exam_number_two
> patient_6_exam_number_two
> patient_7_exam_number_two
> patient_8_exam_number_two


No, not really... The order in TBSS is following the alphabetical order, so in this case you would have:
patient_1_exam_number_one
patient_1_exam_number_two
patient_2_exam_number_one
patient_2_exam_number_two
etc.

You can either adapt your matrix to this order, or put a prefix to your scans so that they will be concatenated in the order you want.

> 2- In order to do longitudinal analysis I have to use
> randomise with a .mat and .con created with glm: open Glm,
> select "Higher-level / non-timeseries design",
> input is number of exams or number of subjects (in my case
> is it 8 or 16?) 

16 (2 groups of 8 scans)

> click "Wizard", select "2
> group paired" click "Process" then I save the
> output.
> 
> Is the output as follows correct?

Hmm, no... If you specify 16 inputs + the two groups paired wizard, you should end up with 9 EVs (1 for the two groups, then one for each subject), not 13...

Hope this helps,
Gwenaelle