Lukas:
A procedure used in FSL by some methods is to first register your MRI to a standard MRI; use the extracted brain from the standard MRI and transform this back into your subject space; perhaps dilate it; then mask your MRI with the inverse-transformed-dilated standard brain. This will eliminate the eyes, head, neck, etc.
I know structural MRIs better than fMRI; but it is my understanding that the BET procedure works the best on T1- (and T1/T2), for which the above procedure works very well (inverse mask from standard MRI, then BET). If you have a high-res anatomical T1 then do the BET on the T1, then use the T1-DTI registration to mask the DTI.
John
***********************************************
John E. Richards
Carolina Distinguished Professor
Department of Psychology
University of South Carolina
Columbia, SC 29208
Dept Phone: 803 777 2079
Fax: 803 777 9558
Email: [log in to unmask] <applewebdata:[log in to unmask]>
HTTP: jerlab.psych.sc.edu
*************************************************
On 10/22/15, 11:18 AM, "FSL - FMRIB's Software Library on behalf of Lukas A. Jung" <[log in to unmask] on behalf of [log in to unmask]> wrote:
>Dear FSL experts,
>
>as part of a dissertation I am doing some analysis of DTI data and am about to finish the preprocessing soon. For that I use BET (brain extraction using bet2) to create a mask by varying the -f and -g options which works well in most cases. But there are some data where I could not avoid that a part of the eyes or an artifact is included in the mask. Now I am wondering how far this could affect further analysis (e.g. if it provides more inaccuracy for tractography by including "false activation"). I attached two screenshots to give you examples:
>"Eye" - left eye is affected a bit (this is the most common problem) and right eye is affected a lot (only a very few cases like that but maybe the much bigger problem); I think the main reason for this problem is that the eyes themselves are obviously distorted due to head motion
>"Artifact" - there is an occipital artifact which is included by the mask
>
>Would you see problems with an brain extraction like that and if so, are there any opportunities to create a better fitting mask?
>
>
>Many thanks in advance!
>
>
>
>Best regards
>
>Lukas A. Jung
>Goethe-University Frankfurt, Germany
>Department of Psychiatry, Psychosomatic Medicine and Psychotherapy
>Laboratory of Neurophysiology and Neuroimaging
>
|