Issue with image orientation

[sunly92]

I am trying to understand and check the CT geometry in DICOM and TIGRE.
I plot my phantom using volshow in Matlab and it gives me a display of standard DICOM orientation looks like:

Then I define my geometry (pretty normal one) and plot it:

Supposedly the source should be on the left hand side based on the two figures? If so, the first projection should be from the side, but the first projection shows

Could you kindly inform me of what I am missing?

Thanks

[AnderBiguri]

Just that dicom contains orientation information that has not been used to flip the volume. This is not very important, but of course you can just use the matlab function "flip" to flip the volume upside down if this is something you prefer.

Note that almost surely the dicom also shows that the pixels are anisotropic, ie the size in cráneo caudal direction of the pixels is much bigger than in the other two. You are also not using this in Tigre (geo.dVoxel) which is why it looks like a "squeezed" human.

[sunly92]

Thank you for replying. My question is if the source is placed at the positive X direction, and the X direction is towards the left hand side in DICOM as shown in my first figure, the starting projection should be from the side and not from the front?

[AnderBiguri]

@sunly92 that would be true if MATLABs volshow used the same coordinate system as Tigre, but it seems that it doesn't.

Not super inportant anyway, just rotate (permute) your volume, and you'll be fine. Or add +pi/2 to the angle definition