Skip to Main content Skip to Navigation
Theses

Indexation guidée par les connaissances en imagerie médicale

Abstract : In this research, we focus on the exploitation of medical knowledge sources in order to support the diagnostic decision making. Our goal is to define a medical diagnostic decision support system capable to handle medical data, information and knowledge in an efficient manner to provide the physician, in the system¿s output, information permitting to facilitate diagnostic decision-making. Initially, we address the two mains issues in a medical diagnostic decision support system: medical knowledge representation and reasoning mechanisms adapted to the considered representation model. In this work, two modes of reasoning are particularly considered: Classification based reasoning and Similarity based reasoning (Case based reasoning). Knowledge representation and reasoning are discussed in the light of two characteristics of medical information: the heterogeneity and imperfection. As a general framework of the proposed system, we considered the application of the possibility theory thanks to its advantages compared to other decision support theories in terms of capability of handling and processing of heterogeneous and imperfect information. Based on this theory, we proposed two possibilistic models for medical knowledge representation, and for each proposed possibilistic model, we have discussed the two adopted types of reasoning by classification and by similarity. The performance of the proposed system is evaluated by considering an endoscopic medical application with two bases: a knowledge base consisting of a set of diagnostics and lesions (case) base. The results are very promising and show the effectiveness of possibility theory as a framework for knowledge representation as well as a tool for medical and diagnostic reasoning. In addition, the proposed approach has been very effective in the integration of multiple sources of knowledge, in order to define the similarity between cases and the use of the confidence index as a decision criterion (in terms of quality of information provided to the physician).
Document type :
Theses
Complete list of metadata

Cited literature [81 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00719587
Contributor : Bibliothèque Télécom Bretagne Connect in order to contact the contributor
Submitted on : Friday, July 20, 2012 - 11:23:04 AM
Last modification on : Saturday, June 25, 2022 - 9:04:18 PM
Long-term archiving on: : Friday, December 16, 2016 - 1:20:53 AM

File

2012telb0226_Al_Sun.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : tel-00719587, version 1

Citation

Mohammad Homam Al Sun. Indexation guidée par les connaissances en imagerie médicale. Traitement des images [eess.IV]. Télécom Bretagne, Université de Bretagne Occidentale, 2012. Français. ⟨tel-00719587⟩

Share

Metrics

Record views

317

Files downloads

584