The value labeled "Intensity" in the lower right of the fslview window shows the value in the selected image under the cross hair. Use that to see what the tstat value (contrast_n_tstatn) is and to check if it is significant (contrast_n_tfce_corrp_tstatn, where you are looking for a 1-minus-P of 0.95 or greater to attain FWE significance at 0.05). Usually people load up their corrp image, change to a colorful lookup table (click '(i)'), and window the image 0.95 to 1.0 (Min and Max near the top of the window).
-To
On Wed, Feb 1, 2012 at 5:31 PM, Leslie Engineering
<[log in to unmask]> wrote:
I ran TBSS on my DTI data. For group comparisons I have implemented randomise. For each contrast in my design matrix, the following files are outputs:
contrast_n_tfce_corrp_tstatn
contrast_n_tstatn
If I open contrast_n_tfce_corrp_tstatn in fslview how do I determine the t-value (or p or z) of each white matter tract?
If it is a corrected p value, does that mean the highlighted tracts are significant? Or, do I need to threshold the file? If so, what is a standard threshold?
THanks so much experts!
--
__________________________________________________________
Thomas Nichols, PhD
Principal Research Fellow, Head of Neuroimaging Statistics
Department of Statistics & Warwick Manufacturing Group
University of Warwick, Coventry CV4 7AL, United Kingdom