Hi all,
I’am very new to the SPM world, so I hope someone can help me.
I’m reading manuals and articles but sometimes I found problems I can’t fix by myself.. so I’m still here asking for help.
I’m comparing two set of patients underwent to PET/CT.
Before designing the study I had preprocessed PET scans following these steps.
1) ConvertDICOM to NiFTI format
2) Open images and manually set origin in a well know, always the same, clear anatomical structure (between putamen nuclei, in my case)
3) Normalized Images using INLAB PET Dementia specific atlas (Medium Bias, 7th level of interpolation)
4) Smoothed normalized images (with 4-4-4 matrix, that is 2x voxel dimensions)
I’ hadn't realigned or co-registered images because I understood that this can deform anatomical structures of the PET examinations. Isn’t it ? Am I Wrong ?
I had also used the “check registration” feature... all my PET seems to be perfectly co-registered.
But I think there should be an error, infact my images are co-registered between them but spatial coordinates do not match with dementia specific atlas and this involves that resulting clusters coordinates do not match with atlases.
Infact my occipital lobe coordinates (visually interpreted) match to cerebellum.
Please look at attached screenshot: the first image is the dementia specific atlas, the second and the third two patients.
I think my images needs to be scaled at dementia specific atlas.
Is this correct ?
How to do this ?
Thank you in advance for any help
|