Thanks a lot for message, Doug!
With 4mm slice and no gap, I cover basically the whole brain from top all the
way to cerebellum. After co-registration to T1 coordinates, it was too many
slices to be shown, so I left the most inferior ones out.
Anyhow, I rather see signal-drop out than distortions in the most lower slices
and only minimal artifacts near the frontal sinuses. I guess, I have to scan
more subjects to get better idea, what's going on.
Appreciate your help.
Martin
Quoting Doug Greve <[log in to unmask]>:
> Martin Kavec wrote:
>
> >Hello everybody,
> >
> >as a novice in B0 correction of EPI, I would like to know opinions of
> >experienced fugue-ers of my setup and results. Particularly, I would
> like to
> >hear from those using similar acquisition setup.
> >
> >I have set up B0 mapping sequence on 3T Philips Achieva by following
> Doug
> >Greve's instructions available on web (many thanks for that). I use
> >double-echo gradient echo, with echos separated by 2.5ms. Comparison
> with
> >high-res inversion-recovery spin echo doesn't show any deformation of
> >phase-mapping abs images.
> >
> >For activation studies, I have set up gradient echo EPI 128x128 matrix
> for
> >230mm FOV, 4mm slice, 0 inter-slice gap, and SENSE acceleration factor
> of 2.
> >If I am not mistaken, I have also non-linear shimming of the measured
> volume.
> >FEAT results are nice.
> >
> >After studying FSL prelude-fugue tutorials and Doug's script, I was
> able to
> >correct the EPI images. You can see the results at:
> >http://www.uku.fi/~kavec/B0_correction.gif
> >
> >It is an animated gif, where yellow lines outline brain (anatomical
> T1W,
> >registered with EPI) and the two frames show original and corrected
> gradient
> >echo EPIs.
> >
> >The correction, to me, seems to do the job, by stretching left frontal
> lobe
> >and correcting a skew-like deformation. There are two things, which bug
> me.
> >
> >1. Frontal lobes in lower slices are not well recovered. Is it due to a
> signal
> >drop-out caused by strong dephasing?
> >
> >
> Yes, I think so.
>
> >2. More importantly, the original gradient-echo images show minimal
> >distortions compare to those in, say FEEDS, or experienced near frontal
>
> >sinuses at 1.5T. I don't quite believe that the subject was so great (
> me :)
> >I would expect to see much more distortions at 3T, in gradient-echo
> EPI, 128
> >phase encoding, even with acceleration factor of 2 and non-linear
> shims.
> >
> >
> I think your slices are just superior to the real dropout regions.
> You're just starting to see it in your most inferior slice, and if you
> kept collecting at lower slices, it would likely be much more evident.
>
> doug
>
> >Thanks for your thoughts.
> >
> >Martin
> >
> >PS: Sorry for such a long email.
> >
> >
> >
> >
>
>
> --
> Douglas N. Greve, Ph.D.
> MGH-NMR Center
> [log in to unmask]
> Phone Number: 617-724-2358
> Fax: 617-726-7422
>
|