Many advances have been made in the early diagnosis of Alzheimer’s Disease (AD) using 
connected speech elicited from a picture description task. The use of hand built linguistic 
and acoustic features as well as Deep Learning (DL) approaches have shown promising results in the classification of AD patients. In this research, we compare both approaches on the Cookie Theft scene from the Boston Aphasia Examination with models trained with features derived from the text and audio extracts, as well as a DL approach using BERT. We 
train and evaluate our models on the new ADReSS challenge dataset as well as on the CCNA dataset in order to assess the generalization of the trained model on from a different dataset. A thorough evaluation of the interpretability of the models is performed to see how well each of the models learn the representations related to the disease. 


(The presentation will be in French)

Enregistrement de la présentation : https://drive.google.com/file/d/1yycrOgLFHG5YuvHHhPyJDXU4maKYmNGM/view?usp=sharing