Medical image retrieval using ROI extraction and hybrid bag-of-features model
Abstract Medical image processing is a vital component of modern healthcare, with content-based medical image retrieval (CBMIR) playing an increasingly important role. However, existing CBMIR methods still face challenges in handling multimodal datasets and large-scale image collections effectively....
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2025-07-01
|
| Series: | Journal of Electrical Systems and Information Technology |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s43067-025-00228-7 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Medical image processing is a vital component of modern healthcare, with content-based medical image retrieval (CBMIR) playing an increasingly important role. However, existing CBMIR methods still face challenges in handling multimodal datasets and large-scale image collections effectively. This paper proposes a novel CBMIR framework based on a hybrid bag-of-features model designed for improved scalability, interpretability, and retrieval performance. The approach integrates Bag-of-Visual-Words (BoVW) with shape and texture descriptors to enhance feature richness. Region of interest extraction is performed using the dynamic delaunay triangulation method to eliminate irrelevant regions and focus on diagnostically significant areas. Binary robust invariant scalable keypoints and K-means clustering are used to generate the BoVW, while edge histogram descriptors and gray-level co-occurrence matrix are employed to extract shape and texture features. The final retrieval is performed using a Euclidean distance similarity measure. Experimental evaluation on the IRMA 2009 and Kvasir datasets demonstrates that the proposed framework achieves competitive accuracy and robustness compared to existing methods. These results support the framework’s potential to address key limitations in CBMIR, especially in clinical environments where interpretability and computational efficiency are essential. |
|---|---|
| ISSN: | 2314-7172 |