On 25/01/2010 09:24, "Guido Biele" <[log in to unmask]> wrote:
> Hello,
>
> I am using feat/fugue to unwarp EPI images and I am wondering what the
> correct input for the echo spacing (dwell time) field is when EPI images
> (resolution = 64x64 voxel/slice) were acquired with an accelerated
> sequence.
>
> Specifically, I have images acquired with an acceleration factor of 2 and a
> TE of 30ms. My (naive?) first thought was to enter 30ms/2 = 15ms as the
> echo spacing.
> However, the sequence does not only acquire 64/2 = 32 lines but an
> additional 24 reference lines.
> Does the number of reference lines influence what I should enter as echo
> spacing time?
> [When I account for the reference lines by calculating either
> (a) unwarp_echo_spacing =
> echo_spacing*(lines_per_slice/acceleration_factor +
> reference_lines)/lines_per_slice as opp) or
> (b) unwarp_echo_spacing = lines_per_slice/acceleration_factor)
> results with (a) look often better, but I might of course simply shift voxels
> to location that actually suffer from signal drop outS]
Hi Guido,
Unfortunately EPI, like life, isn't that simple. The echo spacing you need
is the time between acquisition between successive lines of k-space. There
isn't a simple relationship between the echo spacing and TE.
Your first place to start would be to ask the
physicist/technologist/radiographer at the site your data was acquired for
the echo spacing. If that isn't possible, and you have DICOM images, the
echo spacing is probably in one of the DICOM tags. For example, on GE
scanners it's in the private tag (0x0043, 0x102c) EffectiveEchoSpacing. If
you have another make of scanner, perhaps someone else on the list can tell
you where to look.
Dave
>
> Iąd be thankful for any tip.
>
> Regards - Guido
>
|