Hi André,
> I am still struggling with this issue. You're right in some subjects there are black holes in the mask.img. What is the easiest way to detect the relevant volume and how can I fix it?
In general, spm does not consider any voxel for which there is no data
for all timepoints/subjects. Hence, I have seen many second-level models
go bad due to one subject being off. The screenshot you sent me
off-list, though, only has a tiny hole where the large draining interior
veins converge, which is not that uncommon and should not be a problem.
Again, the mask.img in your second level directory tells you which
voxels were included.
> It is strange that the estimation error only occurred for some contrasts but not others.
As far as I remember, spm does an omnibus f-test with a weak cutoff to
see if it is worthwhile going ahead in the first place. I was under the
impression that the f-test, being omnibus ;) is calculated across all
columns but I may be wrong here. If this is done per contrast, then
contrasts with no effect might fail while others might work. Otherwise,
I have no further ideas on what might cause this to fail.
Cheers,
Marko
> ________________________________________
> Von: Marko Wilke [[log in to unmask]]
> Gesendet: Mittwoch, 30. September 2015 12:48
> An: André Schmidt; [log in to unmask]
> Betreff: Re: [SPM] estimation of second level batch
>
> Hi André,
>
> one common, and potentially fatal, reason is that not all voxels are
> available for the whole group. Have you checked your mask.img image from
> the group study to see which voxels actually enter the analysis?
>
> Cheers,
> Marko
>
> André Schmidt wrote:
>> Dear experts
>>
>> I try to estimate a second level batch with a 2-sample test design and
>> always receive the error message:
>>
>> Error using spm_est_non_sphericity (line 196)
>> Please check your data: There are no significant voxels.
>>
>> Error in spm_spm (line 418)
>> [xVi, am] = spm_est_non_sphericity(SPM);
>>
>> Error in spm_getSPM (line 234)
>> SPM = spm_spm(SPM);
>>
>> Error in spm_results_ui (line 261)
>> [SPM,xSPM] = spm_getSPM;
>> Error while evaluating uicontrol Callback
>>
>> I have checked the contrast images for all included subjects and all of
>> them show activation. Do you know what went wrong here?
>>
>> Thank you very much
>> André
>
> --
> ____________________________________________________
> PD Dr. med. Marko Wilke
> Facharzt für Kinder- und Jugendmedizin
> Leiter, Experimentelle Pädiatrische Neurobildgebung
> Universitäts-Kinderklinik
> Abt. III (Neuropädiatrie)
>
> Marko Wilke, MD, PhD
> Pediatrician
> Head, Experimental Pediatric Neuroimaging
> University Children's Hospital
> Dept. III (Pediatric Neurology)
>
> Hoppe-Seyler-Str. 1
> D - 72076 Tübingen, Germany
> Tel. +49 7071 29-83416
> Fax +49 7071 29-5473
> [log in to unmask]
>
> http://www.medizin.uni-tuebingen.de/kinder/epn/
> ____________________________________________________
>
>
>
--
____________________________________________________
PD Dr. med. Marko Wilke
Facharzt für Kinder- und Jugendmedizin
Leiter, Experimentelle Pädiatrische Neurobildgebung
Universitäts-Kinderklinik
Abt. III (Neuropädiatrie)
Marko Wilke, MD, PhD
Pediatrician
Head, Experimental Pediatric Neuroimaging
University Children's Hospital
Dept. III (Pediatric Neurology)
Hoppe-Seyler-Str. 1
D - 72076 Tübingen, Germany
Tel. +49 7071 29-83416
Fax +49 7071 29-5473
[log in to unmask]
http://www.medizin.uni-tuebingen.de/kinder/epn/
____________________________________________________
|