Segmentation of cerebrovascular mra image based on the optimal sub-block images

Magnetic resonance angiography (MRA) is the primary choice for vascular imaging in clinical practice. The precision of MRA vessel segmentation is the key to study the characteristics of the cerebral vascular anatomical structure, diagnosis, treatment and evaluation of vascular disease. In this paper, we propose a method for cerebrovascular segmentation based on the Otsu algorithm used to MRA sub-block images. First, original image is preprocessed with anisotropic diffusion filtering. Second, the optimal size of the sub-block image is determined by calculating the change rate of the image internal standard deviation between two images of different sub-block size. Next, all the sub-block images are segmented by using the Otsu algorithm, and then all segmented images are merged into a final result. Finally, the final result is reconstructed by visualization toolkit (VTK) and projected to an image for quantitative evaluation. The experimental results show that the proposed method can obtain the better segmentation effect with the more cerebral vascular details than the Otsu algorithm directly used for segmentation on the none sub-block images. © 2015 ISSN.