Abstract:A novel semi-automatic vessel segmentation technique is proposed that is based on a combined filter using wavelet transform, region-growing, threshold segmentation and morphological algorithms. Bone is the major barrier in visualizing the 3-D blood vessel tree from CT Angiography data, the technique can remove the bone region and retain most of the CTA data untouched. The accuracy of the method is tested and proved successfully by visual inspection of all segmented slices. The blood vessels are clearly visualized in 3-D with different ray casting volume rendering techniques without the obstruction of bone.

A Novel Method For Vessel Segmentation Based On Combined Filter