 |
 |
Hi, On 21 Sep 2007, at 15:36, Antonios - Constantine Thanellas wrote: > Dear fsl users, > > I want to evaluate Controls and Patients and i want to ask some > questions( > regarding the data acquisition) in order to have valid analysis > results > after the use of sienax, siena and randomise. > > As far as it concerns the data parameters the data should have > (among them) > the same: > pulse sequence > TE > TR > T1 > Matrix size > FOV > B field strength Yes, ideally all the above should be the same. > Acquisition plane This isn't crucial as long as you've got ok resolution. > for accurate analysis with the fsl tools? or it's only enough for > the data > just to have the same modality, resolution in-plane size which is > less than > 2mm and slice thickness that can vary among 1 and 6mm? > > As far as it concerns the relation between the data and the MNI152 > standard > image that sienax and randomise use, the data should use the same > parameters > as the MNI152? No, it's not important to match the acquisition to what's in the MNI152. > And if yes which ones of the previous mentioned should be the > same? If for instance we have sagittaly data acquisitions then if we > analyse them with sienax that registers them to the axially > oriented MNI152, > will this cause some trouble? No, though the registrations may be slightly more robust if you flip your data with fslswapdim so that it appears axial with the same axes as the MNI152. Note that such a flip does not involve interpolation and hence there's no loss of quality in the data. > The data should be as close as possible to the MNI152 standard image > resolution in both the in-slice resolution and slice thickness? In > FSL-FAQ > you mention that it's better to use in plane resolution of less > than 2mm and > slice thickness between 1 to 6 mm. So at least the slice thickness > in not an > important issue if it's between some limits (as it is also > mentioned in > Smith et. al 2002). Suppose that our data are of 1.7x1.7x1.7 mm. Is > it the > same to use the MNI152 of 1.5x1.5x1.5 or 2.0x2.0x2.0 or even the > 1x1x1 as > the standard image for our comparisons? Or we have to choose the > one that is > as close as possible to our data resolution (which is 1.5x1.5x1.5 > in the > previous example)? Not sure what you're asking here... > Generally is it better to use the MNI152 template as the standard > template > or to create our own template for analyzing our data? (since sienax > and > randomise need a standard space template). If you are using adult brains then the best thing is probably simply to use the standard MNI152 template suplied with FSL. If you are using (eg) non-humans or children then it's probably worth generating a study-specific template. > If the second case is the > preferred one then we can make our own template by just averaging our > controls, do cross sectional analysis (sienax) and then repeat the > same > step with a patients template? Is this the best way? Where can i > find info > on how to make my own template? > > What abour randomise? Since with randomise we register the flow > images of > our groups (in our case controls and Parkinsons patients) that are > derived > from siena in a standard space image (MNI152) is it better to use this > default randomise template as the standard one or our own one? And > in this > case if it's better to make our own template, this template should > be an avg > image of both our controls and patients? where can i also find some > more > info about this topic? In this case I would recommend using the standard template and derived brain edge map. > > Finally is the MNI (2x2x2) which is the default standard image > used by all > applications of siena: flirt siena and randomise etc in the fsl 4 > version? Some FSL4 tools use 2mm standard space and others (eg TBSS) use 1mm. Cheers. > > Thanks in advance > Antonios-Constantine Thanellas ------------------------------------------------------------------------ --- 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 ------------------------------------------------------------------------ ---