David Atkinson

It looks like it is still there. When this line is called `subplot_([2,3,1],FDG_arr [100:-100, 100:-100],'FDG' ,cmap="hot")` `FDG_arr` will be (1,344,344) (I don't know how the above subplot copes without a...

Resolving lead to the expand_dims fix for a new bug. Original aim of removing the [100:-100 … code was not implemented. Not essential, just for clarity.

The patient is a 3D object, but they can change in time (for example beating heart), or you can inject a contrast agent that changes the image intensities, or you...

But when you unravel, you need to know the size of each dimension and that is not in the intent code? I'm not sure what is the point of forcing...

> > But when you unravel, you need to know the size of each dimension and that is not in the intent code? > > Yes, that's why it would...

Just a guess, but have you tried scaling to 1 below 2^16, i.e. `img_data *= (2^16 - 1)` and then rounding to integer values? The range of uint16 is 0...

If you multiply by 1000 do you get the same effect of everything after the decimal point being zero? In that case we know it is wrong because we have...

The Reg case makes most sense to me - the 3rd array dimension has shape value of 15 which I assume corresponds to slice, and the corresponding 3rd column of...

From the NIfTI header: ``` N.B.: The i index varies most rapidly, j index next, k index slowest. Thus, voxel (i,j,k) is stored starting at location (i + j*dim[1] +...

I don't understand these comments fully, but I just want to clarify that the data has pixel widths and heights that are the same, but the slice thickness is different....