Dear Sabrina,
Are you really trying to do a vertex analysis between two subjects?
This is really not possible and I would not trust any output that you got.
If you meant something else then please let us know.
You should be able to include the design matrix in the email if you
just copy it as text into the email itself.
All the best,
Mark
On 20 Jul 2011, at 20:42, diagnostico imagenes wrote:
> Dear Mark,
> Now I am using only two subjects to try the vertex analisys and I want to now if there is a difference in the left hippocampus.
> I checked the vtk files in the 3D viewer and it looks fine.
> I tried to attach the design matrix and the vtk files from two left hippocampus but the files are too heavy, how can I send you those files?
> Thank you in advance,
> Best regards,
> Sabrina
>
>
> > Date: Tue, 14 Jun 2011 19:33:48 +0100
> > From: [log in to unmask]
> > Subject: Re: [FSL] design matrix for vertex analysis
> > To: [log in to unmask]
> >
> > Dear Sabrina,
> >
> > By original analyses I mean the individual subject segmentations.
> > Have you checked these, and do they look correct?
> >
> > I suspect that the error you see here has something to do with
> > your design matrix. Can you show me the contents of the
> > file con0_dis1.mat ? Also, let me know how many subjects
> > you have and what sort of question(s) you are trying to ask
> > with this design matrix.
> >
> > All the best,
> > Mark
> >
> >
> > On 14 Jun 2011, at 19:20, diagnostico imagenes wrote:
> >
> > > Dear Mark,
> > >
> > > We use the command:
> > > first_utils --vertexAnalysis --usebvars -i con_to_dis_L_Hipp.bvars -d con0_dis1.mat -o shape_analysis/con0_dis1_L_Hipp --useReconNative --useRigidAlign -v >& shape_analysis/con0_dis1_L_Hipp_output.txt
> > >
> > > It only creates de .txt file inside the shape_analysis folder. An it says "abortado"
> > > and the text in the .txt is
> > >
> > > read model
> > > done reading model
> > > setting up shape/appearance model
> > > The shape has 732 vertices.
> > > 336 modes of variation are retained.
> > > The model was constructed from 336 training subjects.
> > > model loaded
> > > read model
> > > done reading model
> > > setting up shape/appearance model
> > > The shape has 732 vertices.
> > > 336 modes of variation are retained.
> > > The model was constructed from 336 training subjects.
> > > 1.12962 0.0220672 0.03358 -4.52313
> > > -0.0133035 1.08221 -0.0541649 8.58251
> > > -0.0318979 0.0605173 1.15819 -16.5332
> > > 0 0 0 1
> > > 1.12962 0.0220672 0.03358 -4.52313
> > > -0.0133035 1.08221 -0.0541649 8.58251
> > > -0.0318979 0.0605173 1.15819 -16.5332
> > > 0 0 0 1
> > > done recon
> > > done recon and reg
> > > new design matrix
> > > 1 0.5 -0.5
> > > 1 -0.5 0.5
> > > evs 1
> > > set hsape Mesh
> > > 1 0 0
> > > 0 0 1
> > > do mvglm
> > > terminate called after throwing an instance of 'NEWMAT::SingularException'
> > >
> > > What do you mean by original analyses?
> > >
> > > Thank you in advance,
> > > Best regards,
> > > Sabrina
> > >
> > > > Date: Mon, 6 Jun 2011 21:49:45 +0100
> > > > From: [log in to unmask]
> > > > Subject: Re: [FSL] design matrix for vertex analysis
> > > > To: [log in to unmask]
> > > >
> > > > Hi,
> > > >
> > > > Can you please send us the exact commands you tried to run
> > > > and the exact error messages they gave.
> > > >
> > > > Also, have you checked that all the original analyses were correct?
> > > >
> > > > If it creates a log directory, please also look at the contents of
> > > > files there too and let us know of any errors.
> > > >
> > > > All the best,
> > > > Mark
> > > >
> > > >
> > > > On 6 Jun 2011, at 20:53, diagnostico imagenes wrote:
> > > >
> > > > > Thanks Mark,
> > > > > I could create the matrix but now I have a new problem.
> > > > > When I wrote the command to do the vertex analysis, it starts to run but the it says "abortado". It only creates the txt file inside the shape_analysis directory.
> > > > > I created the concatenated bvars using: concat_bvars and then I wrote all the filenames I wanted to concatenate (patients and controls).
> > > > > I hope you can help me
> > > > > Thanks a lot
> > > > > Best regards
> > > > > Sabrina
> > > > >
> > > > >
> > > > > > Date: Fri, 3 Jun 2011 21:05:19 +0100
> > > > > > From: [log in to unmask]
> > > > > > Subject: Re: [FSL] design matrix for vertex analysis
> > > > > > To: [log in to unmask]
> > > > > >
> > > > > > Dear Sabrina,
> > > > > >
> > > > > > Open the Glm GUI.
> > > > > > Choose the higher-level design.
> > > > > > Set the number of inputs to the total number of subjects.
> > > > > > Select the Wizard.
> > > > > > Choose two groups, unpaired.
> > > > > > Set the number of subjects in the first group.
> > > > > > Press Process.
> > > > > > Check the image looks OK.
> > > > > > Go back to the original (small) window.
> > > > > > Select Save.
> > > > > > Choose the appropriate directory.
> > > > > > Type in a filename.
> > > > > >
> > > > > > And that's it.
> > > > > >
> > > > > > All the best,
> > > > > > Mark
> > > > > >
> > > > > >
> > > > > >
> > > > > > On 3 Jun 2011, at 20:55, diagnostico imagenes wrote:
> > > > > >
> > > > > > > Hi,
> > > > > > > I am trying to use the vertex analysis for the hippocampus in FSL-FIRST but I am having problems to generate de design matrix. The manual says that I should use the glm in the FSL program with 1 EV and the matrix should be 0 for one group and 1 for the other. I tried to create the matrix by changing the number of inputs but then I canīt save it.
> > > > > > > Can you please give me more detailed instructions to create the matrix???
> > > > > > > Thank you very much in advance,
> > > > > > > Best regards,
> > > > > > > Sabrina
|