Dear all,
I am attempting to use a single subject: conditions and covariates model to look at sex or diagnosis (control or schizophrenic) as conditions and neuropsychological measures (e.g. handedness scores) as covariates in structural MRI analyses of explicitly masked data, but something is going awry.
Upon computing the model, the spmT files appear to be written without problem, look fine, and show max T’s of 4.9, 6.6, etc. which I would expect to be significant. After running a contrast and hitting ‘volume’ however, there is a Matlab error beep (see output below) and the SPM clockface appears indefinitely as though it were computing rather than showing stats or stating ‘no suprathreshold clusters.’ The SPM display shows 'Height threshold T = NaN' and 'Extent threshold k = 0 voxels.' Any suggestions?
Also, can we test for a sex by diagnosis (conditions) interaction still using the neuropsych covariates?
Thank you,
Gina Clark
Matlab output:
SPM99: spm_results_ui (v2.31) 11:34:47 - 03/04/2003
========================================================================
SPM99: spm_getSPM (v2.35) 11:34:47 - 03/04/2003
----------------------------------------------------------------
contrast image 2 : ...written con_0002.img
spm{T} image 2 : ...written spmT_0002.img
SPM computation : ...computing
xNum =
0 0 0 0 0
...done
contrast structure : ...saved to xCon.mat
Warning: Returning NaN for out of range arguments
> In /sun1/jupiter/software/SPM/spm99/spm_invTcdf.m at line 89
In /sun1/jupiter/software/SPM/spm99/spm_u.m at line 25
In /home/gina/spm99_anat/spm_getSPM.m at line 659
In /sun1/jupiter/software/SPM/spm99/spm_results_ui.m at line 261
Warning: No voxels survive height threshold u=NaN
> In /home/gina/spm99_anat/spm_getSPM.m at line 673
In /sun1/jupiter/software/SPM/spm99/spm_results_ui.m at line 261
edf =
1.0000 -2.0000
FWHMmm =
8.6181 8.8695 8.4521
>> ??? Error while evaluating uicontrol Callback.
??? Error using ==> gammaln
Input arguments must be real and nonnegative.
Error in ==> /home/gina/spm99_anat/spm_P.m (spm_ECdensity)
On line 152 ==> b = exp(gammaln((v+1)/2) - gammaln(v/2));
Error in ==> /home/gina/spm99_anat/spm_P.m
On line 56 ==> EC = spm_ECdensity(STAT,Z,df);
Error in ==> /home/gina/spm99_anat/spm_list.m
On line 258 ==> Pz = spm_P(1,0,u,df,STAT,1,n);
________________________
Gina Clark
Department of Psychiatry
Warneford Hospital
Oxford University
Oxford OX3 7JX
UK
Tel: +44 (0)1865 455914
|