 |
 |
Hi - You get FA values that are higher than 1 when you have negative eigenvalues. In dtifit, there is no positivity constraint on the tensor (or equivalently on the eigenvalues). So it may happen that you get some, especially in low SNR data. You will notice however that not many voxels will have negative tensors, and most of them will be outside the brain, in which case you can exclude them from the brain mask.. Saad. On 30 Jul 2008, at 04:16, Yujin Zhang wrote: > Dear FSL users: > > I am a question about the range of calculated FA value using > dtifit: should it be less than 1 determinately? By the definition > of FA, I think the answer is yes. and when I use dtifit on example > data (fdt_subj1), I got very reasonable results, with FA ranging > from 0 to 0.770387. But when I calculated FA maps using dtifit on > our data (256*256*45, 13 directions, eddy current corrected), FA > maps were ranging from 0 to 1.224745. The answer maps looked like > reasonable, however, with high FA values where they should be > (e.g., corpus callosum and CST). I tested L1, L2 and L3 value > (using fslstats) at the same time. Their range were as follow: > L1: -0.002313 0.004803 > L2: -0.002548 0.004386 > L3: -0.002707 0.004128 > I have a question why negative value came out? when I compute FA > maps using another program (DTIstudio) on the same data, the FA is > from 0 to 1 with positive L1, L2 and L3. > In the mannul of TBSS, it said that: " If you created your FA data > using dtifit in FDT then the data will indeed be roughly in the > range 0:1. " I am wondering why FA were bigger than 1 if it is > correct. Were there something wrong in my progress? > > Thank you very much! > > > -- > Zhang Yujin > National Key Laboratory of Cognitive Neuroscience and Learning > Beijing Normal University ------------------------------------------------------------------------ --- Saad Jbabdi, Postdoctoral Research Assistant, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222545 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~saad ------------------------------------------------------------------------ ---