Hi All,
I have MRI images that have unconfirmed left/right orientation. The
images could have been flipped during conversion.
I also have PET images that are always in confirmed radiological
orientation.
I do two things:
1. Coregister PET to MRI
2. Coregister PET to the same MRI in (1) but flipped Left/Right
I now have to choose (i.e. by looking or examining statistics)
whether using the flipped or non-flipped MRI was better (correct).
I had an idea of choosing the right oriented MRI by using the one
with the lowest cost function when being coregistered to its
corresponding PET.
Does anyone have any information about whether this is a
scientifically correct thing to do?
Could a cost function discriminate between left/right?
Has anyone had similar problems? Where they were not sure if their
images were in correct orientations?
Thanks,
Arthur
|