[Insight-users] Bone Segmentation from MR Volume

[Taru S]

*Dear ITK developers,
*
I  am working on the segmentation of the anatomical structures from the knee
MR 3D volume. I tried some of the region growing segmentation algorithms
(confidence connected, connected threshold, neighborhood connected etc) and
got some success with the segmentation of the Tibia bone but the results
aren't good for the Femur.

The intensity based thresholds don't work well with the MR so it is hard to
define the anatomical boundaries.

I tried  ITK-SNAP for the same. The Region Competition approach works well
for Tibia but not for others because of the intensity issue.  The Edge
Feature approach results in leaks ( I tried different advection and
curvature velocities wrt propagation vel ).

I will really appreciate your advise on the approach I should follow for
this segmentation.

Thank You,
Taru
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20090723/0eac8043/attachment.htm>