![]() | Only 14 pages are availabe for public view |
Abstract Blood vessel segmentation plays an important role in medical image analysis. Blood vessel segmentation of 2D or 3D images is a challenging problem due to the acquisition environment (contrast, different intensities), noise, complex structure of blood vessels, the variability in the blood vessel size and finally the overlap with other organs. The modern blood vessel segmentation algorithms try to increase patient safety by providing a better diagnosis and support more accuracy medical decision. The segmentation algorithm should be robust against various image modalities and acquisition environment and less user-interaction.This thesis aims to introduce two automatic segmentation algorithms from 3D medical images. An approach to robust 3D image segmentation, image method termed as automatic segmentation started from an idea from related work, but extended to further research to reach better results and satisfy needs of the 3D image applications. |