Content based medical image retrieval using multi-feature extraction and patch Sorensen similarity indexing technique Online publication date: Tue, 01-Oct-2024
by K. Saminathan; S. Amsavalli; M. Chithra Devi
International Journal of Bioinformatics Research and Applications (IJBRA), Vol. 20, No. 5, 2024
Abstract: In the intricate field of medical imaging, the analysis of image content plays a pivotal role in classification, retrieval, and indexing tasks, as well as in recognising objects and different settings within the image. While innovative, traditional methods typically fail to efficiently and accurately process medical image databases' massive and complicated data. Due to this shortcoming, discrete wavelet coefficients-bag of visual words-contour-local binary pattern (DWC-BoVW-Contour-LBP) relevance fusion was developed. A trimmed mean filter and contrast limited adaptive histogram equalisation (CLAHE) remove noise and boost contrast to optimise the image for feature extraction in this novel method. The system carefully extracts low-level frequency features using discrete wavelet transform (DWT), textural features using local binary pattern (LBP), shape features using contour analysis, and visual features using bag of visual words (BoVW). Pixel image fusion is used to combine various features into a complete picture. Patch Sorensen similarity measurement ranks database photos by query resemblance and selects the top 10 most similar images. The algorithm's precision, F-score, and recall were superior in the TCIA-CT database, showing a substantial progress in content-based medical image retrieval (CBMIR).
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Bioinformatics Research and Applications (IJBRA):
Login with your Inderscience username and password:
Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.
If you still need assistance, please email subs@inderscience.com