[Insight-users] Confidence connected (IKT) irregularity?

[MacDuff, Charles]

Hi everyone,

 

I notice that when I'm using the plug-in: Confidence connected of
VolView 3.2, there's something strange happening for certain set of
parameters.

 

When I set the number of iteration to 1 and the variance multiplier to 1
(other parameters by default), the final segmented volume is bigger than
when I put the number of iteration to 20 (with the same variance
multiplier=1).

Logically, at 20 iterations, the final segmented volume should be bigger
because de growth has processed several times.

 

 

Also, after a variance multiplier of about 1.5, the algorithm begins to
work normally (more iterations we have, bigger the final segmented
volume will be).

 

So, I wonder what's happening.

 

Thank you very much in advance for your precious help, 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20110401/1fc6621e/attachment.htm>