 |
 |
Hi, On 1 Feb 2007, at 15:58, Bum Seok Jeong wrote: > Hi, > > In my case, neck area is always remained after BET during SIENA. > If I have segmented brain (WM, CSF, GM), is there any way to bypass > only the segmentation process of SIENA? Well, you could edit $FSLDIR/src/siena/siena_diff.cc and change if(1) // always done unless the above uncommented and used instead of this test to if(0) and it will use the pre-existing segmentations > In my knowledge, SIENA is a tool to compare between two time point > in same subject. > Is it possible to compare among three time point in same subjects > like repeated measure anova? SIENA can only take in two images, but you can run all 3 pairings and combine the results in a number of statistical ways. You might look at Frost et al Statist. Med. 2004; 23:3275–3286 Cheers, Steve. > > Bests, > > bsjeong > > > > Steve Smith wrote: >> Hi, >> >> At this point SIENAX does not use betsurf to generate the skull >> outline, it uses the simpler skull output by bet itself. Hence the >> brain extraction and the skull estimation are both estimated at >> the same time by the same call to bet at the start of SIENAX. >> Hence whilst it is easy to change SIENAX to allow you to hand edit >> the brain extraction, it's not so easy to feed that hand editing >> into the skull estimation. >> >> However, I'd be surprised if this made much difference anyway - >> the bet skull estimation has always been quite rough, but that is >> fine for the purposes of fixing the scaling constraint in SIENAX. >> If you use the -d option in SIENAX you can check the intermediate >> stages, but it's probably working fine. >> >> Cheers, Steve. >> >> >> On 1 Feb 2007, at 00:23, Russell Chu wrote: >> >>> Hi All, >>> Currently, I'm using BET to generate a brain mask, which I hand- >>> edit and use >>> to extract the brain from T1-weighted MR images. Just by visual >>> inspection, >>> it seems that brain masks requiring significant hand-editing also >>> have less >>> accurate skull outlines generated with BETSURF. If this is the >>> case, then >>> less accurate inner skull outlines may adversely affect the inverse >>> v-scaling values from SIENAX. Would it be possible to somehow take >>> advantage of the hand-edited masks by inputting them into SIENAX >>> to improve >>> the inverse v-scaling value? Any input would be greatly appreciated. >>> Thanks in advance. >>> >>> Cheers, >>> Russell Chu >> >> >> --------------------------------------------------------------------- >> ------ >> Stephen M. Smith, Professor of Biomedical Engineering >> Associate Director, Oxford University FMRIB Centre >> >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >> +44 (0) 1865 222726 (fax 222717) >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >> --------------------------------------------------------------------- >> ------ ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ ---