An automatic kidney segmentation from abdominal CT images
The problem we address in this paper is the automatic kidney segmentation from abdominal 2D computed tomography images (CT). Accurately identifying kidney in medical image is a key step relating to the study of various kidney diseases. However, kidney segmentation from CT images is generally perform...
Saved in:
Published in | 2010 IEEE International Conference on Intelligent Computing and Intelligent Systems Vol. 1; pp. 280 - 284 |
---|---|
Main Authors | , |
Format | Conference Proceeding |
Language | English |
Published |
IEEE
01.10.2010
|
Subjects | |
Online Access | Get full text |
ISBN | 9781424465828 1424465826 |
DOI | 10.1109/ICICISYS.2010.5658676 |
Cover
Summary: | The problem we address in this paper is the automatic kidney segmentation from abdominal 2D computed tomography images (CT). Accurately identifying kidney in medical image is a key step relating to the study of various kidney diseases. However, kidney segmentation from CT images is generally performed manually or semi-automatically because of gray levels similarities of adjacent organs, contrast media effect and relatively high variation of organ's positions and shapes in abdominal CT images. The proposed method in this paper is the combination of medical anatomic knowledge and image processing methods a which can be divided into two stages. First, an improved connected component labeling algorithm based on intensity value is proposed to extract estimated kidney position (EKP). This algorithm is applicable to abdominal CT images of different sizes by using the position of kidney and spine. In the second stage, a novel region growing approach based on multi-scale mathematical morphology and labeling algorithm is used to extract the fine kidney regions. The method is tested on over 200 clinically acquired images and very promising results were obtained. |
---|---|
ISBN: | 9781424465828 1424465826 |
DOI: | 10.1109/ICICISYS.2010.5658676 |