[Insight-users] The performance of level sets in segmenting bones

[wyao4 at uwo.ca]

Hi,

I found the performance of ITK embedded level sets methods was bad in segmenting
bones. The CT image I used was not so blurred. I worry about whether I did not
choose the parameter values properly. Have you used level sets to segment wrist
or other bones? What is your feeling about the performance? I appreciate your
response.

The CT is attached. The procedure I used is

GeodesicActiveContourImageFilter.exe

with parameters: 5.0 2.3 -0.3 2.0 10.0
and more than 10 seeds(I added some code to let the procedure accept and work
for multiple seeds, but the result from the multiple seeds is not improved
compared to the single seed case).

Weiguang

-------------- next part --------------
A non-text attachment was scrubbed...
Name: GeodesicActiveContourImageFilterOutput1.png
Type: image/x-png
Size: 14773 bytes
Desc: not available
Url : http://public.kitware.com/pipermail/insight-users/attachments/20040812/65e5cd1d/GeodesicActiveContourImageFilterOutput1-0001.bin