you can try Otsu thresholding

http://rsb.info.nih.gov/ij/plugins/otsu-thresholding.html

or k-means clustering

http://ij-plugins.sourceforge.net/plugins/segmentation/k-means.html

or watershed segmentation

http://bigwww.epfl.ch/sage/soft/watershed/

For complicated organ structures it is rare that structure independent methods such as the above pixel/gradient methods are robust.  

A reasonable compromise for CT is using an augmented manual technique.  A good  plugin for ImageJ is Yawi-3D ( http://yawi3d.sourceforge.net ) in reaching this compromise.  It is a bit older but can still work with ImageJ.

Good Luck.

Why not try MITK or 3D Slicer?

An another option would be to try ITK-Snap. It has a number of nice functions for segmenting 3D images.  You can find it at http:/www./itksnap.org 

http://www.itksnap.org