From ADNI data to 2D matrix data

[Alberto N]

Hi all,

I would like to obtain from ADNI data (fMRI or PET) a 2D matrix where each row indicates the brain region of interest, while each column indicates the time instant, so that in each cell I have the quantity of interest. In this way I can run my algorithms for further analysis. The problem is HOW to obtain the 2D matrix: I don't have a biomedical background, thus I don't know ho to perform the pre-processing step but I have seen that in ADNI there are pre-processed data. 

Can you give me some advices on how to achieve the above task?

PS: I tried to download the NIFTI pre-processed MRI image, but I have seen they are in 3D and not 4D.... should not the data be 4D since the 4th dimension is the time?

Thank you

[Nicolò]

Hi Alberto,

If I understand correctly you want to extract the 2D image matrix from the volume.

If you use python it is very simple since the information is encoded in x, y, z.

You have to imagine the volume of ADNI's brain as a cube where each slice is xy and scroll in z (a bit when you integrate a double integral).

So in pseudocode:

vol = reader ("path")

silce1 = vol [:,:, 1]

in this case I have extracted the first slice xy

This can also be repeated at a temporal level, whereas you have (x,y,z,t)

Then the tensor now is a 3d matrix: 

sile2_3d = vol[:,:,1,:]

fmri they should be codify in 4d  matrix the rest MRI, PET not.

Best,

Nicolò

[Alberto N]

Hi Nicolò,

thank you for your reply. Maybe I'm confused on what is econded on the NIFTI file then. If I have a 3D matrix, should not it be a cube of voxel representing only 1 time instant? Or it represent a 2D slices over time in the 3rd dimension? And by the way, I would like to have pre-processed data so that I don't have to perform operations I never did with possibly methodical errors. There are also DICOM files in ADNI, but I imagine that are the raw files to be pre-processed, right?

Could you give me a brielfy explanation?

Thank you,

Alberto