Dear Ping,
Regarding your question on which input features to use for robust
segmentation of WM lesions I would recommend you to read our related work:
"Segmentation of age-related white matter changes in a clinical
multi-center study" published in Neuroimage;41(2):335-45, 2008.
Best,
Tim
Matt Glasser wrote:
>
> If you want to get only the lesions I would suggest just using the
> FLAIR. You might need to play with the number of classes (default is
> 3) to get the lesions as their own tissue class. This is the –n
> option (e.g. –n 3 means to divide the image into three classes. It is
> sufficient to type fast –n 3 FLAIRImageName.nii.gz to segment an
> image. Also, you should brain extract the image first. Have a look
> at the FAST manual page for more detailed instructions:
> http://www.fmrib.ox.ac.uk/fsl/fast4/index.html
>
>
>
> Peace,
>
>
> Matt.
>
>
>
> ------------------------------------------------------------------------
>
> *From:* FSL - FMRIB's Software Library [mailto:[log in to unmask]] *On
> Behalf Of *WangPing
> *Sent:* Tuesday, December 16, 2008 4:28 PM
> *To:* [log in to unmask]
> *Subject:* Re: [FSL] white matter lesion segmentation on FLAIR image
> *Importance:* High
>
>
>
> Thanks Matt, I would try. In the past I only know that FAST can
> segment brain tissues into WM, GM, CSF. The lesion areas are more
> clear on Flair modality, should I only use Flair, or need to combine
> other modalities?
> BTW, what the type command line of FAST for this case (lesion
> segmentation), thanks ahead for your help.
> Ping
>
> ------------------------------------------------------------------------
>
> Date: Tue, 16 Dec 2008 16:19:51 -0600
> From: [log in to unmask]
> Subject: Re: [FSL] white matter lesion segmentation on FLAIR image
> To: [log in to unmask]
>
> If they are of significantly different intensity from other tissues in
> a particular modality, yes.
>
>
>
> Peace,
>
>
>
> Matt.
>
>
>
> ------------------------------------------------------------------------
>
> *From:* FSL - FMRIB's Software Library [mailto:[log in to unmask]] *On
> Behalf Of *WangPing
> *Sent:* Tuesday, December 16, 2008 4:14 PM
> *To:* [log in to unmask]
> *Subject:* Re: [FSL] white matter lesion segmentation on FLAIR image
> *Importance:* High
>
>
>
> Matt, you mean FAST could segment the lesions? My Flair shows some
> white matter lesions due to some diseases. Best, ping
>
> ------------------------------------------------------------------------
>
> Date: Tue, 16 Dec 2008 16:00:35 -0600
> From: [log in to unmask]
> Subject: Re: [FSL] white matter lesion segmentation on FLAIR image
> To: [log in to unmask]
>
> Have you tried using FAST (FMRIB's Automated Segmentation Tool)? It
> will work with multiple image channels too.
>
>
>
> Peace,
>
>
>
> Matt.
>
>
>
> ------------------------------------------------------------------------
>
> *From:* FSL - FMRIB's Software Library [mailto:[log in to unmask]] *On
> Behalf Of *WangPing
> *Sent:* Tuesday, December 16, 2008 3:33 PM
> *To:* [log in to unmask]
> *Subject:* [FSL] white matter lesion segmentation on FLAIR image
> *Importance:* High
>
>
>
> Dear FSL list,
>
> Sorry for bothering, I want to get help from the list regarding while
> matter lesion segmentation on FLAIR image. Seems FSL has no this
> tool, I am wondering if anyone could recommend one tool or give some
> suggestions?
> P.S. Besides Flair, I do have T1, T2, PD, T2 for each subject.
>
> Best Regards.
> Ping Wang
>
> ------------------------------------------------------------------------
>
> 更多热辣资讯尽在新版MSN首页! 立刻访问! <http://cn.msn.com/>
>
>
>
> ------------------------------------------------------------------------
>
> MSN保护盾,十八般武艺保障MSN安全! 现在就下载! <http://im.live.cn/safe/>
>
>
>
> ------------------------------------------------------------------------
>
> MSN热搜榜全新升级,更多排行榜等着你! 立即查看!
> <%20http://top.msn.com.cn>
>
|