Detecting cognitive decline using speech only: The ADReSS O Challenge
Abstract Building on the success of the ADReSS Challenge at Inter-speech 2020, which attracted the participation of 34 teams from across the world, the ADReSS o Challenge targets three difficult automatic prediction problems of societal and medical relevance, namely: detection of Alzheimer’s Dementia, inference of cognitive testing scores, and prediction of cognitive decline. This paper presents these prediction tasks in detail, describes the datasets used, and reports the results of the baseline classification and regression models we developed for each task. A combination of acoustic and linguistic features extracted directly from audio recordings, without human intervention, yielded a baseline accuracy of 78.87% for the AD classification task, a root mean squared (RMSE) error of 5.28 for prediction of cognitive scores, and 68.75% accuracy for the cognitive decline prediction task.