Hi,
Yes, you are right.
If you keep all the parameters of the gradient-echo sequence the
same as the spin-echo (except of course for the 180 pulse) then you
get a field map image with very little distortion. Technically all
images suffer from distortion but normally on modern machines
with one k-space line per shot then this distortion is negligible and
I have been calling it undistorted.
Practically, gradient-echo sequences tend to be worse because of the
reduced signal to noise due to through voxel dephasing, and
EPI-based sequences are very bad indeed as these are also distorted.
Like most acquisition issues, there are a number of different
trade-offs involved and our practical experience is that better
results are obtained using spin-echo based sequences.
However, any field-mapping technique can be used with Fugue
as it has a --loadfmap option which allows a previously generated
field map (in rad/s) to be loaded in. Similarly, Prelude can be
used for phase unwrapping of any general images. It is only the
-p option to Fugue which is specific to the spin-echo based field maps.
I hope this helps make things a little clearer and more precise.
Mark
David Lythgoe wrote:
>>From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On
>>
>>
>Behalf
>
>
>>Of Mark Jenkinson
>>Sent: 25 April 2004 23:07
>>To: [log in to unmask]
>>Subject: Re: [FSL] Fugue and Prelude applied on Siemens 3T Allegra
>>
>>As for gradient-echo vs spin-echo, at the moment we only recommend
>>using spin-echo because the field maps are then undistorted and do
>>not suffer from signal loss.
>>
>>
>
>It is not true to say that spin-echo images are undistorted. Assuming
>the same sequence parameters (except of course replacing the spin-echo
>with a gradient-echo), spin-echo and gradient-echo images should have
>the same distortions. The gradient-echo image would have signal losses
>due to phase dispersion over the voxel, whereas the integrated intensity
>of a spin-echo image should be independent of background gradients. In
>other words, integrated over the whole image, a gradient-echo would give
>reduced intensity, but a spin-echo image should give a constant
>intensity, but with increased or decreased signal in some voxels.
>
>
>Dave
>
>
|