Title :
Echocardiographic image sequence segmentation by multiscale directional snake
Author :
Cheng, Jierong ; Foo, Say Wei ; Krishnan, Shankar M.
Author_Institution :
Sch. of Electr. & Electron. Eng., Nanyang Technol. Univ., Singapore
Abstract :
In this paper, a new algorithm named as ´multiscale directional snake´ is proposed for segmentation of echocardiographic images by computer. The method makes use of morphological operations to locate the left ventricular center point and region of interest. In addition, a multiscale directional edge map is introduced to improve the performance of snake. The snake is confined within the region of interest during deformation and does not require initialization like traditional snake. Results show that the boundaries obtained using the proposed method fit very closely to the manually traced boundaries.
Keywords :
echocardiography; edge detection; image segmentation; image sequences; medical image processing; boundary detection; echocardiographic image segmentation; image sequence; left ventricular center point; morphological operation; multiscale directional edge map; multiscale directional snake; region-of-interest; Active contours; Heart; Image edge detection; Image processing; Image segmentation; Image sequences; Morphological operations; Pixel; Smoothing methods; Ultrasonic imaging;
Conference_Titel :
Signal Processing and Information Technology, 2004. Proceedings of the Fourth IEEE International Symposium on
Print_ISBN :
0-7803-8689-2
DOI :
10.1109/ISSPIT.2004.1433680