 |
 |
Hi, Thank you for that great suggestion. Jeremy ------------------------------------ Hi I was just thinking that if you wanted to extract a regional summary measure (e.g. regionally masking the final 4D FSLVBM image that gets fed into randomise, and extracting the mean value for each subject with fslmeants) then that would give you a single regressor to put into the TBSS randomise - and more easily done than with SIENAx. p.s. - please include previous emails in replies to the list so I (and others) can see what has gone before in the thread. Thanks. Cheers, Steve. On 16 Oct 2012, at 00:00, Jeremy Strain wrote: > Hi, > Ok, I apologize but I am still confused about one thing. Are you suggesting that I can use the FSLVBM output as a voxel-dependent covariate? If thats the case then I am confused on how that can be incorporated with a TBSS design that only looks at the white matter skeleton. > > Thanks, > Jeremy ------------------------------- Hi - I still think you're better off using FSLVBM outputs for this - e.g. masking the 4D GM image output by FSLVBM to give you the covariates you need? Cheers. On 13 Oct 2012, at 00:48, Jeremy Strain wrote: > Hi, > First, thank you for responding. > Second there is some method to my madness. The reason I propose this question is in a recent paper I was arguing that the significant differences I found in TBSS between two groups was in more part due to white matter changes than gray matter changes. To support this claim I performed sienax to get the gray matter volume estimates and used it as a covariate. However, the reviewers thought this was a poor strategy and nitpicked that my changes could be due to regional gray matter changes as opposed to whole brain changes. Therefore, I wanted to use the results from fslvbm as a mask so that I could create a new gray matter volume that only consists of the atrophic regions between the two groups. Hopefully addressing the reviewers point and still bolstering my claim. > > Thanks, > Jeremy ------------------------------- Hi - if you're running a full FSLVBM Then I don't think there's too much point in also running SIENAX? Cheers -------------------- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford. OX3 9 DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ---------------------- On 8 Oct 2012, at 18:38, Jeremy Strain <[log in to unmask]> wrote: > Dear FSL list, > I have a question that regards both fslvbm and sienax. I have two populations one patient and the other control. I performed fslvbm on both groups and then did randomise to find group differences. I was wondering is it possible to limit the sienax normalized calculations to only the significant regions that were revealed from the fslvbm? In other words I binarized the fslvbm results and I only want to include those voxels instead of the entire brain. > > Thank you, > Jeremy >