Project description

The principal objective of this research project is to bring out the subjective phenotypic signs of autism that are implicitly used by diagnosticians during their decision-making process but are not included as standardized signs described in the international classifications, or in the scoring of standardized instruments. This project will use language processing tools in diagnostic reports of children assessed in autism diagnostic clinic. The secondary objective is to verify the relevance of these signs, in the form of words or combinations of words, in the decision-making process of autism diagnosis by machine-learning models.