A hybrid case-based rule-based system used in a content-based image retrieval system
- Resource Type
- Conference
- Authors
- Bradford, Ian; Evans, Carl; Winter, Robin
- Source
- 2008 7th IEEE International Conference on Cybernetic Intelligent Systems Cybernetic Intelligent Systems, 2008. CIS 2008. 7th IEEE International Conference on. :1-6 Sep, 2008
- Subject
- Computing and Processing
Robotics and Control Systems
Knowledge based systems
Image retrieval
Content based retrieval
X-ray imaging
Spine
Biomedical imaging
Medical diagnostic imaging
Diagnostic radiography
Shape
Approximation algorithms
Case based reasoning
content based image retrieval
dysmorphology
skeletal dysplasias
- Language
Images are an important resource in the diagnosis of skeletal dysplasias. A proof-of-concept system, X-Ray Specs, has been developed for evaluating and experimenting with content-based image retrieval on radiographs of skeletal dysplasias. Its domain-dependent sub-system uses a hybrid case-base/rule-base as an important part of its image classification and retrieval process. This proved more accurate than other classification/retrieval systems tested in X-Ray Specs. The case-base/rule-base sub-system is flexible and could function in any domain where suitable parameters can be extracted.