Interactive fuzzy connectedness image segmentation for neonatal brain MR image segmentation

被引：0

作者：

Kobashi, Syoji ^{[1
]}

Kuramoto, Kei ^{[1
]}

Hata, Yutaka ^{[1
]}

机构：

[1] Univ Hyogo, Grad Sch Engn, Himeji Initiat Computat Med & Hlth Technol, Himeji, Hyogo 6712201, Japan

来源：

2013 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN, AND CYBERNETICS (SMC 2013) | 2013年

关键词：

interactive image segmentation; fuzzy connectedness image segmentation; radial-basis-function network; neonatal brain; magnetic resoance images; ALGORITHMS;

D O I：

10.1109/SMC.2013.311

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

Image segmentation plays a fundamental work to analyze medical images. Although many literatures studied automated image segmentation, it is still difficult to segment region-of-interest in any kind of images. Thus, manual delineation is important yet. In order to shorten the processing time and to decrease the effort of users, this paper introduces two approaches of interactive image segmentation method based on fuzzy connectedness image segmentation (FCIS). The first approach interactively updates object affinity of FCIS according to users' additional seed voxels. The second approach models the profile of the object affinity using radial-basis function network (RBFN), and applies online training for users' additional seed voxels. The proposed methods updates segmentation results for not only the seed voxels but also the other miss-classified voxels. The methods had been applied to neonatal brain magnetic resonance (MR) images. The experimental results showed the second approach produced the best results.

引用

页码：1799 / 1804

页数：6

共 20 条

[1] SEEDED REGION GROWING [J].