Hi,
We do not have a reliable and robust way of segmenting WML in FSL.
We have found this to be a very difficult problem and are working on a method, but it is not yet ready.
Depending on your sequence, and exactly what you need to do, it might be possible to get a reasonable result by using a multi-channel segmentation (giving some combination of T1w, T2w, PD, and FLAIR images, all registered, as inputs to FAST).
All the best,
Mark
On 21 Sep 2012, at 02:39, Ai Edward <[log in to unmask]> wrote:
> Hi,
>
> I am using FSL-FAST to segmentate GM and WM in patients with white matter lesions(WMLs). But I found FAST segmentation generally mistake most of WMLs as GM. I don not know how to solve this problem. Anyone know how to solve this problem? Thanks very much!
>
>
>
|