Ikram CHOURIB Thesis defense

In the last few decades, decision support has significantly evolved and has been widely adopted in the medical field. This has led to the emergence of a large number of Medical Decision Support Systems (MDSS), some of which have been easily adopted in medicine while others have remained at the experimental stage.
Due to the complexity of clinical situations, clinicians need assistance in making critical and effective decisions. In this context, the Case Based Reasoning (CBR) methodology has been widely adopted to help experts perform their tasks. Its use has led to significant progress in solving problems related to the diagnosis, therapy and prognosis of diseases. However, this methodology has shown some limitations that have led researchers to consider its use in conjunction with other solution methods, such as data mining and machine learning methods.
In this perspective, we have worked on three axes: medical reasoning, decision support and design of the medical decision process. Our contribution concerns the design of a medical decision support system containing three main modules allowing the structuring of the bases (case base and expert knowledge base), the extraction of knowledge and the integration of knowledge, with the objective of providing assistance to the expert in his decision making process (and not replacing the expert) and this by processing massive data from different sources. This is achieved by jointly using a case base and an expert knowledge base while integrating the expert into the case-based reasoning system to ensure continuous and incremental learning. The expected output of our approach will be in the form of a decision processing report containing interpretable results that prove whether the target case is worthy of consideration or not.
The obtained results are encouraging and highlight interesting perspectives to be explored.