Abstract
The objective of extracting bone structure boundaries from computed tomography (CT) data is to serve as input for three-dimensional model reconstruction. Various boundary detection algorithms1 for CT images have been developed in the last few decades but very few of them are completely automatic. In addition, most of them were not capable of dealing with critical locations such as narrow gaps. A novel intelligence-based approach is proposed for automatic extraction of bone boundaries. This approach involves three major steps: gap detection, fuzzy multilevel thresholding and boundary transform (thinning method). Results of applying the approach on CT images are presented, with a discussion of the potential for clinical application.