OATAO - Open Archive Toulouse Archive Ouverte Open Access Week

A New Combination Method Based on Adaptive Genetic Algorithm for Medical Image Retrieval

Gasmi, Karim and Torjmen-Khemakhem, Mouna and Tamine, Lynda and Ben Jemaa, Maher A New Combination Method Based on Adaptive Genetic Algorithm for Medical Image Retrieval. (2014) In: 10th Asia Information Retrieval Society Conference (AIRS 2014), 3 December 2014 - 5 December 2014 (Sarawak, Malaysia).

[img]
Preview
(Document in English)

PDF (Author's version) - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
505kB

Official URL: https://doi.org/10.1007/978-3-319-12844-3_25

Abstract

Medical image retrieval could be based on the text describing the image as the caption or the title. The use of text terms to retrieve images have several disadvantages such as term-disambiguation. Recent studies prove that representing text into semantic units (concepts) can improve the semantic representation of textual information. However, the use of conceptual representation has other problems as the miss or erroneous semantic relation between two concepts. Other studies show that combining textual and conceptual text representations leads to better accuracy. Popularly, a score for textual representation and a score for conceptual representation are computed and then a combination function is used to have one score. Although the existing of many combination methods of two scores, we propose in this paper a new combination method based on adaptive version of the genetic algorithm. Experiments are carried out on Medical Information Retrieval Task of the ImageCLEF 2009 and 2010. The results confirm that the combination of both textual and conceptual scores allows best accuracy. In addition, our approach outperforms the other combination methods.

Item Type:Conference or Workshop Item (Paper)
Additional Information:Thanks to Springer editor. This papers appears in Volume 8870 of Lecture Notes in Computer Science ISSN : 0302-9743 ISBN 978-3-319-12843-6 The original PDF is available at: https://link.springer.com/chapter/10.1007/978-3-319-12844-3_25
Audience (conference):International conference proceedings
Uncontrolled Keywords:
Institution:French research institutions > Centre National de la Recherche Scientifique - CNRS (FRANCE)
Université de Toulouse > Institut National Polytechnique de Toulouse - INPT (FRANCE)
Université de Toulouse > Université Toulouse III - Paul Sabatier - UPS (FRANCE)
Université de Toulouse > Université Toulouse - Jean Jaurès - UT2J (FRANCE)
Université de Toulouse > Université Toulouse 1 Capitole - UT1 (FRANCE)
Other partners > Université de Sfax (TUNISIA)
Laboratory name:
Statistics:download
Deposited By: IRIT IRIT
Deposited On:25 Sep 2017 14:48

Repository Staff Only: item control page