Abstract
Computer analysis of images of anatomical structures can benefit from the use of a priori knowledge about the inspected domain. Even though the anatomy structure of humans is variable, it is far more organized than other domains which are commonly used for image understanding (e.g., outdoor scenes or even images of a boxes-and-cylinders world). This paper discusses an organization scheme for modeling the 3-D structure of the abdominal anatomy and its use for analyzing 3-D CAT (Computed Axial Tomography) scans of the abdomen. The discussion is divided into two major portions of the knowledge organization: a relational database for gross anatomy and a 3-D shape model for individual organs using generalized cylinders. The paper also includes an example of 3-D image analysis for the detection of the kidneys' 3-D shape from abdominal CAT scans.
[1]
J. Petasnick,et al.
Attenuation values in computed tomography of the abdomen.
,
1977,
Radiology.
[2]
Gerald J. Agin.
Representation and description of curved objects
,
1972
.
[3]
Uri Shani.
Understanding three-dimensional images: the recognition of abdominal anatomy from computed axial tomograms (cat)
,
1980
.
[4]
Norman I. Badler,et al.
Three-dimensional representations for computer graphics and computer vision
,
1978,
SIGGRAPH.
[5]
Dana H. Ballard,et al.
An Approach to Knowledge-Directed Image Analysis
,
1977,
IJCAI.