Dear all,
I am working on a DTI project on cognitive dysfunction in Parkinson's disease. I have 48 subjects, no controls, and each of which has a set of cognitive tests done with z score and I have the following questions about the data analysis.
1. I am using TBSS to do the analysis and I am a bit confused with the design matrix/contrast set up. I used excel to save the design as text file and used Text2Vest to convert it into .mat and .con files as the input for randomise. The design matrix as below
1 [z score 1] [age 1] [gender 1] [medication dosage 1] [education level 1]
1 [z score 2] [age 2] [gender 2] [medication dosage 2] [education level 2]
1 [z score 3] [age 3] [gender 3] [medication dosage 3] [education level 3]
.
.
.
1 [z score 48] [age 48] [gender 48] [medication dosage 48] [education level 48]
The contrast file
0 1 0 0 0 0
I am expecting a decrease in FA values in the white matter tract if the cognitive scores are lower and therefore set the value 1 in the second column while controlling for the other confounding variables.
Am I right to put a 1 for each of the subject in the first column and what is the meaning for that? It's weird because when I try to remove this column and the corresponding 0 in the contrast file the output can be completely different which I don't really understand - how come controlling for the same variable of "1" for each subject would make the output be so different?
2. Besides running randomise on the skeleton, is it possible to run it on the "all_FA.nii.gz" or "all_MD"nii.gz" files? This is because I am also curious to know the correlation with some of the deep gray matter structures which is affected in Parkinson's disease.
3. Are there any good ways to the ROI analysis on the deep gray matters (e.g. caudate nucleus, thalamus) to extract the MD/FA values out of it, using standardised altas?
Many thanks for your help!!!!
Best,
Oscar
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
|