How to calculating fourier transform?

I research 3d volumetric image using deep learning. I am calculating the fourier transform of 3d volumetric image. When calculating that, do you have to calculate two kind of imagedastore and pixcellabeldata.?

[Rik]

There is nothing in the Fourier transform that inherently forces you to store your data in a different format. What is the background of this question? Why are you asking it? What are you trying (but failing) to do?

[Hla Hla]

I research 3 d volumetric brain tumor image classification using deep learning. As the first step, i will use deep learning. There, in preprocessing step, before i will crop the image, i am thinking of using fourier transform. Can i use that. Please advise. In below coding, Can you use fourier transform for outl and outV? I want to remove noise from the image. What another method would be convenient to use?

outL = readNumeric(pxds);
outV = read(volds);
temp = outL>0;
sz = size(outL);
reg = regionprops3(temp,'BoundingBox');
tol = 64;
ROI = ceil(reg.BoundingBox(1,:));
ROIst = ROI(1:3) - tol;
ROIend = ROI(1:3) + ROI(4:6) + tol;

ROIst(ROIst<1)=1;
ROIend(ROIend>sz)=sz(ROIend>sz);

tumorRows = ROIst(2):ROIend(2);
tumorCols = ROIst(1):ROIend(1);
tumorPlanes = ROIst(3):ROIend(3);

[Rik]

Wouldn't it make more sense to denoise the image and store the result in the imagedatastore? Otherwise you force your deep learning system to only work on Fourier-transformed images. I don't have much experience with deep learning, but that seems highly unconventional.

[Hla Hla]

Thanks Sayar