A new approach for case acquisition in CBR based on multilabel text categorization: a case study in child’s traumatic brain injuries

Abstract

Case-based reasoning (CBR) is an approach to solving new problems based on those already solved in the past. This means searching in previous cases for one that is similar to the new one and reusing it in this new problem situation. In the literature, there are several CBR developments that have paid particular attention to the stages of the process without paying as much attention to the Case Acquisition (CA) stage. This paper focuses on this task through the use of a Multi- Label Text Categorization (MLTC) approach. The objective of this work, is to automatically complete additional information on cases that were obtained from the Magnetic Resonance Imaging (MRI) scan reports provided by the pediatric intensive care unit of Oran hospital -Algeria. The results suggest that the methodology we have proposed and which we call Multi-Label Text Categorization for Cases Acquisition (MLTC4CA) is a promising way to add automatically values' labels to the case that represents a medical situation related to a child victim of Traumatic Brain Injuries (TBIs).