<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=US-ASCII">
<META content="MSHTML 6.00.2800.1458" name=GENERATOR></HEAD>
<BODY>
<DIV><FONT face=Verdana color=#0000ff size=2><SPAN
class=882584315-29102004>Neha, </SPAN></FONT></DIV>
<DIV><FONT face=Verdana color=#0000ff size=2><SPAN
class=882584315-29102004></SPAN></FONT> </DIV>
<DIV><FONT face=Verdana color=#0000ff size=2><SPAN
class=882584315-29102004>GrayscaleFillHoleImageFilter is implemented using
geodesic morphology. In particular, GrayscaleGeodesicErodeImageFilter. In
geodesic erosion, the structuring element is prescribed as a radius of 1 kernel.
There are no options for changing the kernel.</SPAN></FONT></DIV>
<DIV><FONT face=Verdana color=#0000ff size=2><SPAN
class=882584315-29102004></SPAN></FONT> </DIV>
<DIV><FONT face=Verdana color=#0000ff size=2><SPAN
class=882584315-29102004>Geodesic erosion runs with a kernel of radius 1 and
runs until convergence. Geodesic morphology uses a "marker" image and a
"mask" image. One of these images is the one that is "operated on".
The other serves as a "constraint".</SPAN></FONT></DIV>
<DIV><FONT face=Verdana color=#0000ff size=2><SPAN
class=882584315-29102004></SPAN></FONT> </DIV>
<DIV><FONT face=Verdana color=#0000ff size=2><SPAN class=882584315-29102004>Like
Luis mentioned, Running FillHole/GrindPeak followed by an image substraction
will highlight "holes" or "peaks". </SPAN></FONT></DIV>
<DIV><FONT face=Verdana color=#0000ff size=2><SPAN
class=882584315-29102004></SPAN></FONT> </DIV>
<DIV><FONT face=Verdana color=#0000ff size=2><SPAN
class=882584315-29102004>There are other geodesic morphology classes, HMinimum,
HMaximum, HConcave, HConvex, DoubleThreshold, etc.</SPAN></FONT></DIV>
<DIV><FONT face=Verdana color=#0000ff size=2><SPAN
class=882584315-29102004></SPAN></FONT> </DIV>
<DIV><FONT face=Verdana color=#0000ff size=2><SPAN
class=882584315-29102004>Jim</SPAN></FONT></DIV>
<DIV><FONT face=Verdana color=#0000ff size=2><SPAN
class=882584315-29102004></SPAN></FONT> </DIV>
<BLOCKQUOTE>
<DIV class=OutlookMessageHeader dir=ltr align=left><FONT face=Tahoma
size=2>-----Original Message-----<BR><B>From:</B> neha k
[mailto:itkneha1@yahoo.com]<BR><B>Sent:</B> Friday, October 29, 2004 10:13
AM<BR><B>To:</B> Miller, James V (Research); Luis Ibanez; Insight-users
(E-mail)<BR><B>Subject:</B> RE: Examples on GrayScaleFillHoleImageFiler? And
using breast ima ges on mypacs.net<BR><BR></FONT></DIV>
<DIV>Hello Jim,</DIV>
<DIV> </DIV>
<DIV>Thanks for your reply. I just checked implementation of
GrayscaleFilleholeImageFilter<FONT face=Verdana color=#0000ff>, and it uses
Erosion operation that uses some structuring element. So, I am little
confused. </FONT></DIV>
<DIV><FONT face=Verdana color=#0000ff></FONT> </DIV>
<DIV><FONT face=Verdana color=#0000ff>Thanks,</FONT><FONT face=Verdana
color=#0000ff>Neha</FONT></DIV>
<DIV><BR><B><I>"Miller, James V (Research)"
<millerjv@crd.ge.com></I></B> wrote:</DIV>
<BLOCKQUOTE class=replbq
style="PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #1010ff 2px solid">
<META content="MSHTML 6.00.2800.1458" name=GENERATOR>
<DIV><SPAN class=026104913-29102004><FONT face=Verdana color=#0000ff
size=2>A reference to the FillHole technique is mentioned in the
documentation. The excerpt is below.</FONT></SPAN></DIV>
<DIV><SPAN class=026104913-29102004><FONT face=Verdana color=#0000ff
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=026104913-29102004><FONT face=Verdana color=#0000ff
size=2> * Geodesic morphology and the Fillhole algorithm is described
in<BR> * Chapter 6 of Pierre Soille's book "Morphological Image
Analysis:<BR> * Principles and Applications", Second Edition, Springer,
2003.<BR></FONT></SPAN></DIV>
<DIV><SPAN class=026104913-29102004><FONT face=Verdana color=#0000ff
size=2>The GrindPeak algorithm is essentially the dual of this algorithm.
(Though not mentioned in the above text).</FONT></SPAN></DIV>
<DIV><SPAN class=026104913-29102004><FONT face=Verdana color=#0000ff
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=026104913-29102004><FONT face=Verdana color=#0000ff
size=2>These techniques use geodesic morphology. There is no
structuring element to change. They operate on the 4/8 connected
neighbors and run iteratively until convergence.</FONT></SPAN></DIV>
<DIV><SPAN class=026104913-29102004><FONT face=Verdana color=#0000ff
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=026104913-29102004><FONT face=Verdana color=#0000ff
size=2>Jim</FONT></SPAN></DIV>
<DIV><SPAN class=026104913-29102004><FONT face=Verdana color=#0000ff
size=2></FONT></SPAN> </DIV>
<DIV><SPAN class=026104913-29102004><FONT face=Verdana color=#0000ff
size=2> </DIV></FONT></SPAN>
<BLOCKQUOTE>
<DIV class=OutlookMessageHeader dir=ltr align=left><FONT face=Tahoma
size=2>-----Original Message-----<BR><B>From:</B> neha k
[mailto:itkneha1@yahoo.com]<BR><B>Sent:</B> Thursday, October 28, 2004
4:11 PM<BR><B>To:</B> Luis Ibanez; itkneha1@yahoo.com<BR><B>Subject:</B>
Examples on GrayScaleFillHoleImageFiler? And using breast images on
mypacs.net<BR><BR></FONT></DIV>
<DIV>Hello Luis,</DIV>
<DIV>Thanks for your reply. Is there any example on
GrayScaleFillHoleImageFiler or GrayScaleGrindPeakImageFilter?
What book can I refer to understand these filters?</DIV>
<DIV> </DIV>
<DIV>Also, can you please let me know what you mean by "fine
tune the structuring<BR>element for matching the relative size of
calcifications in image". </DIV>
<DIV> </DIV>
<DIV>Since I need pair of breast MRIs (Which unfortunately are not
avilable to me) I checked mypacs.net, where I found one set of images
GRE-images at following link - <A
href="http://www.mypacs.net/cgi-bin/repos/mpv3_repo/wrm/repo-view.pl?cx_subject=28078&cx_image_only_mode=off&cx_repo=mpv3_repo&cx_from_folder">http://www.mypacs.net/cgi-bin/repos/mpv3_repo/wrm/repo-view.pl?cx_subject=28078&cx_image_only_mode=off&cx_repo=mpv3_repo&cx_from_folder</A>=</DIV>
<DIV>Fil. 9,10 and 11. Does it make sense to use these 2
images, subtracting them and applying Watershed on the result. If
so, how can I decide by checking these images if that needs deformable
registration?</DIV>
<DIV>Any input on this is most appreciated. </DIV>
<DIV>Thanks,</DIV>
<DIV>Neha</DIV>
<DIV><BR><B><I>Luis Ibanez <luis.ibanez@kitware.com></I></B>
wrote:</DIV>
<BLOCKQUOTE class=replbq
style="PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #1010ff 2px solid"><BR>HI
Neha,<BR><BR>You can find Breat images
at<BR><BR>www.mypacs.net<BR><BR>you can create an account for free
and<BR>download real datasets.<BR><BR>Look for the title "Shared Cases"
and<BR>click in "Breast".<BR><BR>--<BR><BR>You may want try the advanced
Mathematical Morphology<BR>methods such as: FillHole and
GrindPeak.<BR><BR>http://www..itk.org/Insight/Doxygen/html/classitk_1_1GrayscaleFillholeImageFilter.html<BR>http://www.itk.org/Insight/Doxygen/html/classitk_1_1GrayscaleGrindPeakImageFilter.html<BR><BR>Taking
an input imag, applying one of these filter and<BR>then subtracting from
the original may enhance spiculations<BR>and calcificaitions. You will
have to fine tune the structuring<BR>element for matching the relative
size of calcifications in
your<BR>image.<BR><BR><BR>Regards,<BR><BR><BR>Luis<BR><BR>----------------------------<BR>neha
k wrote:<BR><BR>> Hello All,<BR>> <BR>> I am working on Breast
Image Segmention for deleneating lesions, ducts <BR>> and fat
tissues. Problem is that I don't have Breast Images with <BR>>
contrast agent used. Hence I can't have pre and post contrast agent
<BR>> to subtract them and locate tumour etc. What will be the best
way to <BR>> segment such breast images (in RAW format). I am
currently trying to <BR>> use Watershed segmentation and not getting
any good results with it. <BR>> It causes lot of oversegmentation. I
have not used Level Set Seg. <BR>> method yet.<BR>> Any input is
appreciated. <BR>> <BR>> Thanks,<BR>>
Neha<BR>><BR>><BR><BR><BR><BR></BLOCKQUOTE>
<P>
<HR SIZE=1>
Do you Yahoo!?<BR>Yahoo! Mail - CNET Editors' Choice 2004. <A
href="http://reviews.cnet.com/Yahoo_Mail/4852-9236_7-30980704.html?part=editchoice">Tell
them what you think</A>. </BLOCKQUOTE></BLOCKQUOTE>
<P>
<HR SIZE=1>
Do you Yahoo!?<BR><A
href="http://us.rd.yahoo.com/mail_us/taglines/aac/*http://promotions.yahoo.com/new_mail/static/ease.html">Yahoo!
Mail Address AutoComplete</A> - You start. We finish.</BLOCKQUOTE></BODY></HTML>