Related Experiment Video
Updated: May 5, 2026
14:27
Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
Published on: June 26, 2013
15.2K
Local mesh patterns versus local binary patterns: biomedical image indexing and retrieval
IEEE Journal of Biomedical and Health Informatics
|November 16, 2013
Summary
A novel algorithm enhances biomedical image retrieval by analyzing relationships among neighboring pixels, outperforming standard methods. This approach improves accuracy for medical imaging applications like CT and MR scans.
Area of Science:
- Medical Imaging
- Computer Vision
- Biomedical Engineering
Background:
- Standard local binary patterns (LBP) analyze pixel-neighbor relationships.
- Biomedical image retrieval requires robust and accurate indexing methods.
- Existing spatial and transform domain methods have limitations.
Purpose of the Study:
- To propose a new image indexing and retrieval algorithm for biomedical applications.
- To introduce a method analyzing relationships among surrounding neighbors.
- To enhance the accuracy of medical image retrieval.
Main Methods:
- Developed a novel algorithm using local mesh patterns for image indexing.
- Encoded relationships among surrounding neighbors, dependent on the number of neighbors (P).
- Validated the algorithm by combining it with the Gabor transform.
Main Results:
- The proposed algorithm demonstrated significant improvements in evaluation measures.
- Experiments were conducted on three biomedical image databases (OASIS-MRI, NEMA-CT, VIA/I-ELCAP).
- Outperformed standard LBP, LBP with Gabor transform, and other domain methods.
Conclusions:
- The local mesh pattern algorithm offers superior performance for biomedical image retrieval.
- The method shows promise for applications involving CT and MR imaging.
- This technique provides a significant advancement in medical image analysis and retrieval.
