|
KMID : 1146720140010020075
|
|
국제컴퓨터가상수술학회 학술지
2014 Volume.1 No. 2 p.75 ~ p.79
|
|
|
Segmentation of Arterial Vascular Anatomy around the Stomach based on the Region Growing Based Method
|
|
Kang Ji-Woo
Kim Do-Young
Lee Sang-Hoon
|
|
|
Abstract
|
|
|
Purpose: The region growing has a critical problem that it often extract vessels with unexpected objects such as a bone which has a similar intensity characteristics to the vessel. We propose the new method to extract arterial vascular anatomy around the stomach from the CTA volume without the post-processing.
Materials and Methods: Our method, which is also based on the region growing, requires the two seed points from the use. I automatically extracts perigastric arteries using the adaptive region growing method and it does not need any post-processing.
Results: The three region growing based methods are used to extract perigastric arteries - the conventional region growings with restrict and loose thresholds each and the proposed method. The 3D visualization from the result of our method shows our method extracted the all required arteries for gastric surgery.
Conclusion: By extracting perigastric arteries using the proposed method, over-segmentation problem that unexpected anatomical objects such as a rib or backbone are also segmented does not occurs anymore. The proposed method does not need to sensitively determine the thresholds of the similarity function. By visualizing the result, the preoperative simulation of arterial vascular anatomy around the stomach can be possible.
|
|
|
KEYWORD
|
|
|
Medical image processing, Segmentation, Arterial vascular anatomy, Perigastric vascular anatomy, Computed tomography angiography
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|
|