DI-UMONS : Dépôt institutionnel de l’université de Mons

Recherche transversale
Rechercher
(titres de publication, de périodique et noms de colloque inclus)
2017-10-13 - Colloque/Présentation - poster - Anglais - 1 page(s)

Itani Sarah , Rossignol Mandy , Lecron Fabian , Fortemps Philippe , "A Tree-Based Decision Support System For Attention Deficit Hyperactivity Disorder Diagnosis" in 57th Annual Meeting of the Society for Psychophysiological Research, Vienne , Autriche, 2017

  • Codes CREF : Intelligence artificielle (DI1180), Modèles mathématiques d'aide à la décision (DI1151), Informatique médicale (DI3314), Technologies de l'information et de la communication (TIC) (DI4730), Neurosciences cognitives (DI4296)
  • Unités de recherche UMONS : Psychologie cognitive et Neuropsychologie (P325), Management de l'Innovation Technologique (F113), Mathématique et Recherche opérationnelle (F151)
  • Instituts UMONS : Institut de Recherche en Technologies de l’Information et Sciences de l’Informatique (InforTech), Institut des Sciences et Technologies de la Santé (Santé)

Abstract(s) :

(Anglais) Despite considerable progress in the medical field, the pathophysiology of Attention Deficit Hyperactivity Disorder (ADHD) remains at the core of a controversy, which makes it difficult to diagnose the trouble objectively. To address this issue, we propose a decision support system providing an assessment based on the brain activity. To implement this system, we lead an analysis on the ADHD-200 sample, including resting state fMRI of ADHD and typically developing subjects aged from 7 to 21 years (n = 776). The sample provides notably, for every patient, Blood Level Oxygen Dependent (BOLD) signals for all the cerebral zones, given a brain parcellation in 116 regions. We investigated the dynamism of the brain activity through the energy of the BOLD signals. We extracted the most relevant cerebral zones as presenting significant differences in energy between both control and ADHD groups. This was achieved thanks to a correlation-based selection of zones that weakly correlate in terms of energy but are highly correlated with the diagnosis. Then follows the question of the interactions that exist between these zones. To solve this issue, we resorted to an algorithmic approach that develops decision trees. They establish a sequence of questions about the patient to result in a final diagnosis. Thus, through readability, the system provides a result open to interpretation. Moreover, according to the geographical origin of the subjects, we acquired different diagnosis systems, which reveals the possible influence of social and cultural factors to explain ADHD.


Mots-clés :
  • (Anglais) Children
  • (Anglais) Infants
  • (Anglais) Attention
  • (Anglais) Adolescents