Hi everyone .

I work on istatistical shape analysis of hipocampi. I have 18 diseased and 18 control's data. After all operations, no deformation is seen in fdr statistic. 

I am writing steps which ı follow.

i)    ./convertITKformats 1.hdr 1.gipl

ii)   ./SegPostProcess 1.gipl  -o 01.gipl

iii)  ./GenParaMesh  01.gipl -iter 1000 -label 1

iv)  ./ParaToSPHARMMesh 01_surf.meta 01_para.meta -subdiLevel 10 -spharmDegree 12 \ -flipTemplate template.coef -regTemplate template.meta

v)   ./Meta2VTK 01_surfSPHARM_procalign.meta example1.vtk

   ı apply that prosedure for all data. And ı use spharm_mancova. 

ı downloaded http://www.nitrc.org/docman/index.php?group_id=308&selected_doc_group_id=760&language_id=1#folder Hipocampi example data. And ı use Hipo\comp\template template_surfSPHARM.coef and template_surfSPHARM.meta as template in " iv)  ./ParaToSPHARMMesh 01_surf.meta 01_para.meta -subdiLevel 10 -spharmDegree 12 \ -flipTemplate template.coef -regTemplate template.meta" .

if something wrong with prosedure, please inform me ... 

Thank you .

 Onur.

FDR is usually very restrictive correction method. Do you see something meaningful in the raw-p values?
You can always change your FDR correction threshold. The rest of the pipeline sounds good to me, except that you should be using the CLP versions of the commands, since are the ones that are being maintained and corrected.

Regards,
BEATRIZ

Thanks for quick response. 
 
Yes there is some meanningful in the raw-p values. 

But even ı chance the fdr correction threshold (0.05 to 1) , nothing chance on hipocampi. 

is clp versions usage is same with the old version ?



Onur...