دانلود رایگان مقاله یک روش تجربی به منظور بهبود تصویربرداری MRA

Abstract

In the Region Growing Algorithm (RGA) results of segmentation are totally dependent on the selection of seed point, as an inappropriate seed point may lead to poor segmentation. However, the majority of MRA (Magnetic Resonance Angiography) datasets do not contain required region (vessels) in starting slices. An Enhanced Region Growing Algorithm (ERGA) is proposed for blood vessel segmentation. The ERGA automatically calculates the threshold value on the basis of maximum intensity values of all the slices and selects an appropriate starting slice of the image which has a appropriate seed point. We applied our proposed technique on different patients of MRA datasets of different resolutions and have got improved segmented images with reduction of noise as compared to tradition RGA.

5. Conclusion

The segmentation of blood vessels is an active research area which plays a significant role in many medical applications including diagnosis, surgery planning and radiation treatment. The quality of segmentation in the case of the region growing algorithm completely relies on the selection of seed point. In cases where the selected seed point does not belong to the region of interest (vessels), the whole region will be grown incorrectly. MRA, which is used specifically for images of blood vessels, does not contain the required region in start of slices. Applying a region growing algorithm directly from the first slice will result in an inappropriate seed point, thereby leading to poor segmentation. In order to overcome this problem, an ERGA has been proposed. The vessel images generated by ERGA are improved quality as compared to generic RGA.