- Auteur(s) : Catherine Merck, Annik Charnallet, Sophie Auriacombe, Serge Belliard, Valérie Hahn-Barma, Helgard Kremin, Béatrice Lemesle, Florence Mahieux, Olivier Moreaud, Danièle Perrier Palisson, Martine Roussel, François Sellal, Hervine Siegwart
- Mots-clés : semantic memory, normative data, cut-off, semantic dementia, Alzheimer disease
- Page(s) : 235-55
- DOI : 10.1684/nrp.2011.0194
- Année de parution : 2011
Résumé : Semantic memory is currently defined as a system that holds general knowledge about the world, including knowledge of biological entities and artefacts. To date, we are not aware of any standardized test available in French language, assessing semantic knowledge about the same set of items presented in different input modalities. However, evidence of a cross modal semantic breakdown is required to differentiate a degraded-store semantic impairment from a failure of access. The main goal of the GRESEM working group was to fill this gap by constructing a neuropsychological semantic battery (BECS-GRECO). This battery consists in evaluating a set of 40 items (20 living and 20 non living) and their attributes, in both verbal and visual conditions across 6 subtests. Normative data were derived from performance of 317 healthy volunteers (from 20 years to 75 years and more) and were presented through cut-off scores. This battery was also administered to two different patients groups, either fulfilling diagnostic criteria of semantic dementia or Alzheimer's disease at early stages. Except for the subtest “matching by identity”, this study emphasizes good agreements between sensibility and specificity for the 5 remaining subtests and, thus, the clinical relevance of this battery to detect semantic disorders.