>> 3) A new 3D phase unwrapping algorithm had been published in NeuroImage,
>> http://www.idealibrary.com/links/doi/10.1006/nimg.2002.1092
>> Can anyone tell how this compares the PRELUDE?
>
>I know of Rhodri's work on phase unwrapping and it is a nice algorithm.
Thanks Mark!
>We had a copy of the code, but unfortunately it would crash on some of our
>test images, so I was never able to conduct a comparison.
>
The first version didn't run on the Alpha platform - but this has since
been fixed, and it does now compile and run. I've got binaries for Solaris,
Linux, Windows & Alpha... if anyone one does want to perform that
comparison!
>I suspect, based on
>previous discussions, that both methods will work in the vast majority of
>cases.
I'm sure this is true - particularly for field maps, which usually have
reasonably slow variation over space, and so less wrapping. However, given
that our algorithms use very different approaches, they might have
different strengths & weaknesses, and would perform differently depending
on the degree of noise, true shears in the image and extent of phase
wrapping.
Rhodri
|