Segmentation of vascular intima ultrasound images based on an improved C-V model.
- Author:
Yujie HU
1
;
Ling JIANG
;
Haibo SONG
;
Qi LIU
Author Information
1. Electrical Information School Sichuan University, Chengdu 610065, China.
- Publication Type:Journal Article
- MeSH:
Algorithms;
Artifacts;
Humans;
Image Processing, Computer-Assisted;
methods;
Models, Biological;
Tunica Intima;
diagnostic imaging;
Ultrasonography
- From:
Journal of Biomedical Engineering
2010;27(5):1115-1119
- CountryChina
- Language:Chinese
-
Abstract:
In view of the fact that blood vessels can reflect, to a certain degree, the pathologic changes of cardiovascular diseases, research workers count it necessary to take quick and accurate segmentation of neointimal ultrasound images. The segmentation is faster owing to the reconciled signed distance function added into the classical C-V mode, thus avoiding the step of re-initialization. Moreover, the curve evolution is controlled by external energy item. The displacement of boundary points is calculated. The result shows that the improved model can effectively and accurately extract the contour of neointimal ultrasound images. The curve can not only assess the extracted boundary but also reflect the abnormality of vascular or cardiac rhythm.
