<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta content="text/html; charset=ISO-8859-1"
 http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Just a comment regarding the kd-tree optimization, this will work only
if considered objects are really distinct. (One object must not contain
another one).<br>
But if I correctly understood what you want to do, this should not
apply to your case...<br>
<br>
On 03/26/2010 01:25 PM, Kishore Mosaliganti wrote:
<blockquote
 cite="mid:cd51098d1003261025v14c4ad08p92351662b4d3e35b@mail.gmail.com"
 type="cite">Hi Baoyun,<br>
  <br>
You are right that if you set 4 phases in the levelset, then no phase
will shrink to empty.<br>
  <br>
In ITK, if you use the classes on ScalarChanAndVeseLevelSetImageFilter,
you can separately encode the 3 regions in 3 separate level-set
functions and these functions compete with each other for the
segmentations at the boundaries. You can also control each level-set
parameters separately. <br>
  <br>
If you have N objects, then you can make use of a built-in kd-tree
optimization that speeds up the approach. Each levelset function is
only defined in a small subdomain containing your object instead of the
whole image thereby saving space.<br>
  <br>
  <br>
Kishore<br>
  <br>
  <div class="gmail_quote">On Fri, Mar 26, 2010 at 10:48 AM, Baoyun Li <span
 dir="ltr">&lt;<a moz-do-not-send="true"
 href="mailto:baoyun_li123@yahoo.com">baoyun_li123@yahoo.com</a>&gt;</span>
wrote:<br>
  <blockquote class="gmail_quote"
 style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
    <table border="0" cellpadding="0" cellspacing="0">
      <tbody>
        <tr>
          <td
 style="font-family: inherit; font-style: inherit; font-variant: inherit; font-weight: inherit; font-size: inherit; line-height: inherit; font-size-adjust: inherit; font-stretch: inherit;"
 valign="top">
          <div>Dear All:</div>
          <div>&nbsp;</div>
          <div>Thanks for your reply, now seems I undertand the
question more.</div>
          <div>&nbsp;</div>
          <div>According to Wanlin's paper, I need to encoding the 2
level set to&nbsp; 3 phase if I want to 3 segmenations. If I need 4&nbsp;
segmentation, I need to encode the levelset to 4 phases.</div>
          <div>&nbsp;</div>
          <div>So my origianl question is not correct, once I set 4
phases, no phase will shrink to empty, and I will get 4 segmentation
anyway. (Please correct me if I am wrong).</div>
          <div>&nbsp;</div>
          <div>Thanks and my best regards</div>
          <div>&nbsp;</div>
          <div>Baoyun</div>
          <div>&nbsp;</div>
          <div>Please correct me if I am wrong<br>
          <br>
--- On <b>Thu, 3/25/10, Luis Ibanez <i>&lt;<a moz-do-not-send="true"
 href="mailto:luis.ibanez@kitware.com" target="_blank">luis.ibanez@kitware.com</a>&gt;</i></b>
wrote:<br>
          </div>
          <blockquote
 style="border-left: 2px solid rgb(16, 16, 255); padding-left: 5px; margin-left: 5px;"><br>
From: Luis Ibanez &lt;<a moz-do-not-send="true"
 href="mailto:luis.ibanez@kitware.com" target="_blank">luis.ibanez@kitware.com</a>&gt;
            <div class="im"><br>
Subject: Re: [Insight-users] muliple pahse Chan Vese, one partition
totally shrink<br>
To: "Baoyun Li" &lt;<a moz-do-not-send="true"
 href="mailto:baoyun_li123@yahoo.com" target="_blank">baoyun_li123@yahoo.com</a>&gt;<br>
            </div>
Cc: <a moz-do-not-send="true" href="mailto:insight-users@itk.org"
 target="_blank">insight-users@itk.org</a>, "Mosaliganti, Kishore Rao"
&lt;<a moz-do-not-send="true"
 href="mailto:Kishore_Mosaliganti@hms.harvard.edu" target="_blank">Kishore_Mosaliganti@hms.harvard.edu</a>&gt;<br>
Date: Thursday, March 25, 2010, 10:43 PM
            <div>
            <div class="h5"><br>
            <br>
            <div><br>
Hi Baoyun,<br>
            <br>
            <br>
You seem to be mixing two different issues here.<br>
            <br>
One thing is to segment the image in different regions,<br>
and a separate issue is to find an efficient encoding <br>
for the resulting segmentations.<br>
            <br>
            <br>
If you want to segment three different anatomical structures<br>
from your image, then you want to use a level set with <br>
three phases.<br>
            <br>
That is, one phase per object to be segmented.<br>
            <br>
            <br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Regards,<br>
            <br>
            <br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Luis<br>
            <br>
            <br>
-----------------------------------------------------------------------------<br>
            <div class="gmail_quote">On Wed, Mar 24, 2010 at 5:52 PM,
Baoyun Li <span dir="ltr">&lt;<a moz-do-not-send="true"
 href="http://us.mc1110.mail.yahoo.com/mc/compose?to=baoyun_li123@yahoo.com"
 rel="nofollow" target="_blank">baoyun_li123@yahoo.com</a>&gt;</span>
wrote:<br>
            <blockquote class="gmail_quote"
 style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
              <table border="0" cellpadding="0" cellspacing="0">
                <tbody>
                  <tr>
                    <td
 style="font-family: inherit; font-size-adjust: inherit; font-stretch: inherit;"
 valign="top">
                    <div>Dear All:</div>
                    <div>&nbsp;</div>
                    <div>I have a theoritical question about Chen Vese
multiplse pahse leve set.</div>
                    <div>&nbsp;</div>
                    <div>I have two level set and the two level set can
segment the image to 4 segments, then I can calculate the mean for each
segments as following</div>
                    <div>&nbsp;</div>
                    <font size="2">
                    <div>%c11 = mean (phi1&gt;0 &amp; phi2&gt;0)</div>
                    <div>%c12 = mean (phi1&gt;0 &amp; phi2&lt;0)</div>
                    <div>%c21 = mean (phi1&lt;0 &amp; phi2&gt;0)</div>
                    <div>%c22 = mean (phi1&lt;0 &amp; phi2&lt;0)</div>
                    </font>
                    <div>&nbsp;</div>
                    <div>How about if I only have three segments, let
say segment (phi1&lt;0 &amp; phi2&lt;0) will finally shrink to zeros.</div>
                    <div>&nbsp;</div>
                    <div>Then how to I calcuate c22, and also what
value of c22&nbsp;I&nbsp;should use to&nbsp;update leve set equation as shown below:</div>
                    <div>&nbsp;</div>
                    <div>((u-c22)^2-(u-c12)^2)H(phi2).</div>
                    <div>&nbsp;</div>
                    <div>I saw somebody use the mininum number of the
machine to replace c22, but I could not imagine what will happen and
why to do that.</div>
                    <div>&nbsp;</div>
                    <div>Thanks</div>
                    <div>Baoyun</div>
                    <div>&nbsp;</div>
                    <div>&nbsp;</div>
                    <div>&nbsp;</div>
                    <div>&nbsp;</div>
                    <div>&nbsp;</div>
                    </td>
                  </tr>
                </tbody>
              </table>
              <br>
              <br>
_____________________________________<br>
Powered by <a moz-do-not-send="true" href="http://www.kitware.com/"
 rel="nofollow" target="_blank">www.kitware.com</a><br>
              <br>
Visit other Kitware open-source projects at<br>
              <a moz-do-not-send="true"
 href="http://www.kitware.com/opensource/opensource.html" rel="nofollow"
 target="_blank">http://www.kitware.com/opensource/opensource.html</a><br>
              <br>
Kitware offers ITK Training Courses, for more information visit:<br>
              <a moz-do-not-send="true"
 href="http://www.kitware.com/products/protraining.html" rel="nofollow"
 target="_blank">http://www.kitware.com/products/protraining.html</a><br>
              <br>
Please keep messages on-topic and check the ITK FAQ at:<br>
              <a moz-do-not-send="true"
 href="http://www.itk.org/Wiki/ITK_FAQ" rel="nofollow" target="_blank">http://www.itk.org/Wiki/ITK_FAQ</a><br>
              <br>
Follow this link to subscribe/unsubscribe:<br>
              <a moz-do-not-send="true"
 href="http://www.itk.org/mailman/listinfo/insight-users" rel="nofollow"
 target="_blank">http://www.itk.org/mailman/listinfo/insight-users</a><br>
              <br>
            </blockquote>
            </div>
            <br>
            </div>
            </div>
            </div>
          </blockquote>
          </td>
        </tr>
      </tbody>
    </table>
    <br>
    <br>
_____________________________________<br>
Powered by <a moz-do-not-send="true" href="http://www.kitware.com"
 target="_blank">www.kitware.com</a><br>
    <br>
Visit other Kitware open-source projects at<br>
    <a moz-do-not-send="true"
 href="http://www.kitware.com/opensource/opensource.html"
 target="_blank">http://www.kitware.com/opensource/opensource.html</a><br>
    <br>
Kitware offers ITK Training Courses, for more information visit:<br>
    <a moz-do-not-send="true"
 href="http://www.kitware.com/products/protraining.html" target="_blank">http://www.kitware.com/products/protraining.html</a><br>
    <br>
Please keep messages on-topic and check the ITK FAQ at:<br>
    <a moz-do-not-send="true" href="http://www.itk.org/Wiki/ITK_FAQ"
 target="_blank">http://www.itk.org/Wiki/ITK_FAQ</a><br>
    <br>
Follow this link to subscribe/unsubscribe:<br>
    <a moz-do-not-send="true"
 href="http://www.itk.org/mailman/listinfo/insight-users"
 target="_blank">http://www.itk.org/mailman/listinfo/insight-users</a><br>
    <br>
  </blockquote>
  </div>
  <br>
</blockquote>
<br>
</body>
</html>