[MINC-users] CIVET classification problem

[Claude LEPAGE]

Hi Jeni,

Can you repost at civet-users and follow the thread from there? If you are not
a member, simply subscribe to the group. This is a CIVET related topic and its
answer will interest the CIVET community.

Yours,

Claude



> From minc-users-bounces at bic.mni.mcgill.ca  Thu Dec 10 18:21:14 2009
> X-Authentication-Warning: grumio.bic.mni.mcgill.ca: jchen owned process doing
> 	-bs
> Date: Thu, 10 Dec 2009 18:19:37 -0500 (EST)
> From: Jeni Chen <jchen at bic.mni.mcgill.ca>
> To: minc-users at bic.mni.mcgill.ca
> MIME-Version: 1.0
> X-Spam-Status: No, score=-4.9 required=5.0 tests=ALL_TRUSTED,AWL,BAYES_00 
> 	autolearn=ham version=3.0.3
> X-Spam-Checker-Version: SpamAssassin 3.0.3 (2005-04-27) on 
> 	mail.bic.mni.mcgill.ca
> X-Virus-Scanned: ClamAV 0.91.2/10149/Thu Dec 10 17:26:20 2009 on
> 	mail.bic.mni.mcgill.ca
> X-Virus-Status: Clean
> X-Greylist: Message whitelisted by Sendmail access database, not delayed by
> 	milter-greylist-4.0 (mail.bic.mni.mcgill.ca [132.206.178.34]);
> 	Thu, 10 Dec 2009 18:19:49 -0500 (EST)
> Subject: [MINC-users] CIVET classification problem
> X-BeenThere: minc-users at bic.mni.mcgill.ca
> X-Mailman-Version: 2.1.9
> Precedence: list
> Reply-To: MINC users mailing list <minc-users at bic.mni.mcgill.ca>
> List-Id: MINC users mailing list <minc-users.bic.mni.mcgill.ca>
> List-Unsubscribe: <http://www2.bic.mni.mcgill.ca/mailman/listinfo/minc-users>, 
> 	<mailto:minc-users-request at bic.mni.mcgill.ca?subject=unsubscribe>
> List-Archive: <http://www2.bic.mni.mcgill.ca/pipermail/minc-users>
> List-Post: <mailto:minc-users at bic.mni.mcgill.ca>
> List-Help: <mailto:minc-users-request at bic.mni.mcgill.ca?subject=help>
> List-Subscribe: <http://www2.bic.mni.mcgill.ca/mailman/listinfo/minc-users>,
> 	<mailto:minc-users-request at bic.mni.mcgill.ca?subject=subscribe>
> Content-Type: text/plain; charset="us-ascii"
> Content-Transfer-Encoding: 7bit
> Sender: minc-users-bounces at bic.mni.mcgill.ca
> Errors-To: minc-users-bounces at bic.mni.mcgill.ca
>
> Hi all,
>
> Upon checking the results of the classified images after running CIVET, I 
> notice that quite often, small portions of the grey matter in the frontal 
> pole or the orbitofrontal area gets cut off. I tried playing with the N3 
> parameters and re-ran CIVET but it didn't help. I was also told that the 
> problem might reside in the algorithm used to extract the brain to make 
> the mask (mincbet, I presume?).  Are there options/parameters that we can 
> specify to improve the result of classification?
>
> Also when looking at the verify.png file- is there a reference value to 
> the ones that are indicated. In other words, when I see for example,
>
> classified image  CSF 7.24%  GM 49.79%  WM 42.97%
> native skull mask in stx space (8.75%)
>
> what do I make of these % numbers, does that tell me whether the 
> classified images are within a standard range?
>
> Thanks!
>
> Jeni
>
> ***********
> Jen-I Chen, MSc.
> Agente de recherche
> Universite de Montreal
> 514.343.6111 x4359
> _______________________________________________
> MINC-users at bic.mni.mcgill.ca
> http://www2.bic.mni.mcgill.ca/mailman/listinfo/minc-users
>