Share Email Print

Proceedings Paper

Automated segmentation of middle hepatic vein in non-contrast x-ray CT images based on an atlas-driven approach
Author(s): Teruhiko Kitagawa; Xiangrong Zhou; Takeshi Hara; Hiroshi Fujita; Ryujiro Yokoyama; Hiroshi Kondo; Masayuki Kanematsu; Hiroaki Hoshi
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

In order to support the diagnosis of hepatic diseases, understanding the anatomical structures of hepatic lobes and hepatic vessels is necessary. Although viewing and understanding the hepatic vessels in contrast media-enhanced CT images is easy, the observation of the hepatic vessels in non-contrast X-ray CT images that are widely used for the screening purpose is difficult. We are developing a computer-aided diagnosis (CAD) system to support the liver diagnosis based on non-contrast X-ray CT images. This paper proposes a new approach to segment the middle hepatic vein (MHV), a key structure (landmark) for separating the liver region into left and right lobes. Extraction and classification of hepatic vessels are difficult in non-contrast X-ray CT images because the contrast between hepatic vessels and other liver tissues is low. Our approach uses an atlas-driven method by the following three stages. (1) Construction of liver atlases of left and right hepatic lobes using a learning datasets. (2) Fully-automated enhancement and extraction of hepatic vessels in liver regions. (3) Extraction of MHV based on the results of (1) and (2). The proposed approach was applied to 22 normal liver cases of non-contrast X-ray CT images. The preliminary results show that the proposed approach achieves the success in 14 cases for MHV extraction.

Paper Details

Date Published: 11 March 2008
PDF: 10 pages
Proc. SPIE 6914, Medical Imaging 2008: Image Processing, 69144E (11 March 2008); doi: 10.1117/12.771612
Show Author Affiliations
Teruhiko Kitagawa, Gifu Univ. Graduate School of Medicine (Japan)
Xiangrong Zhou, Gifu Univ. Graduate School of Medicine (Japan)
Takeshi Hara, Gifu Univ. Graduate School of Medicine (Japan)
Hiroshi Fujita, Gifu Univ. Graduate School of Medicine (Japan)
Ryujiro Yokoyama, Gifu Univ. School of Medicine (Japan)
Hiroshi Kondo, Gifu Univ. Hospital (Japan)
Masayuki Kanematsu, Gifu Univ. School of Medicine (Japan)
Hiroaki Hoshi, Gifu Univ. Graduate School of Medicine (Japan)

Published in SPIE Proceedings Vol. 6914:
Medical Imaging 2008: Image Processing
Joseph M. Reinhardt; Josien P. W. Pluim, Editor(s)

© SPIE. Terms of Use
Back to Top