 |
 |
Hi Have you tried coregistering the brain to the template first? I have found that segmentation fails for images that start a long way from the template space - in fact the documentation mentions this, and suggests that brains be within ~30 mm of the template as I recall. Best wishes, Paul Fumiko Hoeft wrote: > Dear Christian and Marko (and others that I have thanked already), > > Thank you very much for the reply. These images were collected at a 1.5T. > So far I tested some poor images using 1. custom templates created in > SPM2 and using a 5-10 yo CCHMC template, 2. reorienting to AC, and > many images look great now. > > Choosing 'No Affine Registration' in Affine Regularisation helped in > many tough cases. Can anyone tell me why this may be the case? I did > not quite understand (or find that information) in the manual. > > I am now trying Chritian's vbm5 beta version but I am running into > some problems. > I get the following error right away. Do you know what the problem is? > ---------- > Warning: Input should be a string. >> In spm_maff>priors at 314 > In spm_maff>affreg at 115 > In spm_maff at 45 > In cg_vbm at 182 > In cg_config_vbm>execute_estwrite at 723 > In spm_jobman>run_struct1 at 1384 > In spm_jobman>run_struct1 at 1392 > In spm_jobman>run_struct1 at 1392 > In spm_jobman>run_struct1 at 1392 > In spm_jobman>run_struct at 1351 > ??? Error using ==> spm_maff>priors > "ƪ" not recognised as type of regularisation. > > ??? Error while evaluating uicontrol Callback. > ----------- > Thanks very much for all your help. > Best, > Fumiko > > > ----- Original Message ----- From: "Christian Gaser" > <[log in to unmask]> > To: <[log in to unmask]> > Sent: Wednesday, April 25, 2007 12:32 AM > Subject: Re: [SPM] VBM5 problem in young children > > > Dear Fumiko, > > On Wed, 25 Apr 2007 08:45:17 +0200, Marko Wilke > <[log in to unmask]> > wrote: > >> Dear All, >> >>> I知 not convinced that this outcome is primarily due to the mismatch >>> between the characteristics of your study population and the template >>> (although it might be aggravated by it): >> >> I fully agree. Your imanges look like they were not normalized correctly >> (remember that even for native-space segmentation, the priors are in >> efect inversely normalized to the input images). Your results look like >> they are mainly the result of the prior probability maps. This can >> happen when matching the priors goes completely wrong which again can >> happen when you have very inhomogenous input data. This is not >> high-field data by any chance? >> >>> Is it possible that you didn稚 set the origin before preprocessing the >>> data? In our experience the unified segmentation approach is pretty >>> vulnerable to end up with such deformed results if you don稚 give the >>> program a reasonable starting point. Assuming that the data quality is >>> o.k., my best guess would be: reset the origin to AC, rerun data >>> preprocessing for these subjects, and you池e done. >> >> This may help, too. Christian had implemented an automated determination >> of the center of mass of an image in order to improve the starting >> estimates in a beta-version of the 5.1 toolbox but I don't know if it is >> out yet. > > I also assume that spatial registration was not correct. You may try > the new VBM5.1 Toolbox, > which is still a beta version with many new (and not yet documented) > features (e.g. correction for > non-isotropic smoothness, segmentation without priors, > pre-registration using center of mass): > > http://dbm.neuro.uni-jena.de/vbm/download/ > >> >> Also, >> >>> The ages of the subjects are 1-3 years old. >> >> there is no reference data for such subjects yet that I am aware of. We >> (Christian, the Cincinnati IRC group and myself) have two abstracts at >> HBM where we investigate a prior-less segmentation for datasets from >> infants. In effect, the segmentation is done as in spm5 but for writing >> out the results only tissue intensity information is used, disregarding >> the influence from the priors. It does make the segmentation somewhat >> more vulnerable, especially w.r.t. inhomogeneity, but it seems to >> considerably improve segmentation results on "unusual" datasets. Not >> sure if the updated 5.1 toolbox is already publicly available but you >> could ask Christian. > > Thanks for the push Marko! As he already mentioned, it might be > helpful for children/infant data > to try the new prior-free segmentation approach, which can be found in > the extended options. > > Best, > > Christian >