Dear Jan,
as you know, periventricular plaques of multiple sclerosis may be
indistinguishable from CSF on long TR / TE- such as T2w-images whereas the
lesions are going to be bright in contrast to dark CSF on long TR/ short TE-
such as proton-density-weighted images. However, the may be quite isointense
to gray matter, and particularily those related to cortical U-fibres may
mess up your segmaenation a bit. Nevertheless, I would feel more inclined to
use the latter for segmentation purposes in MS. Inhomogenity corrections
will largely depend on the quality & the spectrum of images you enter, I
guess.
Regards-
Andreas
----- Original Message -----
From: "Jan Versijpt" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Wednesday, August 22, 2001 1:02 PM
Subject: Segmentation
> Dear All,
>
> We are currently segmenting an MR-series of Multiple sclerosis
> patients. Previously we used T1 images and got fairly good results with
> this. In a current project, we only have T2- and PD-weighted images at
> our disposal.
>
> Can anyone advise which of these two MR-sequences we should use and the
> most
> appropriate level of inhomogeneity correction? As these are multiple
> sclerosis patients with a significant amount of periventricular white
> matter abnormalities, particularly the CSF-segmentation does not seem to
> work very well all the time on the T1 images.
>
> Thanks in advance and kind regards,
>
> Jan
>
|