Related Experiment Video
Updated: Jul 11, 2026
07:13
Multimodal Cross-Device and Marker-Free Co-Registration of Preclinical Imaging Modalities
Published on: October 27, 2023
Automatic multilevel medical image annotation and retrieval
A Mueen1, R Zainuddin, M Sapiyan Baba
1Faculty of Information Technology, Multimedia University, 63100, Cyberjaya, Selangor, Malaysia. ahmad-mueen@mmu.edu.my
Journal of Digital Imaging
|September 12, 2007
Summary
This study introduces a new method for semantic image retrieval using multilevel features and concept hierarchies. The approach enhances image annotation and classification, showing promising results in X-ray image analysis.
Area of Science:
- Computer Science
- Medical Imaging
- Artificial Intelligence
Background:
- Semantic image retrieval relies heavily on accurate image annotation and classification.
- Key challenges include automatic feature extraction, semantic concept modeling, and annotation algorithms.
Purpose of the Study:
- To develop an improved approach for semantic image retrieval.
- To address limitations in automatic feature extraction, concept modeling, and annotation algorithms.
Main Methods:
- Multilevel features were extracted to create comprehensive image content representations.
- A domain-specific concept hierarchy was developed for semantic interpretation.
- Automatic multilevel code generation was proposed for classification and annotation.
Main Results:
- The proposed methods demonstrated effectiveness in addressing the identified challenges.
- Experiments on X-ray images yielded encouraging results for semantic retrieval.
Conclusions:
- The integrated approach of multilevel features and concept hierarchies improves semantic image retrieval.
- The method shows potential for applications in medical image analysis and annotation.
