Cross-corpus Feature Learning between Spontaneous Monologue and Dialogue for Automatic Classification of Alzheimer’s Dementia Speech

Abstract

Speech analysis could help develop clinical tools for automatic detection of Alzheimer's disease and monitoring of its progression. However, datasets containing both clinical information and spontaneous speech suitable for statistical learning are relatively scarce. In addition, speech data are often collected under different conditions, such as monologue and dialogue recording protocols. Therefore, there is a need for methods to allow the combination of these scarce resources. In this paper, we propose two feature extraction and representation models, based on neural networks and trained on monologue and dialogue data recorded in clinical settings. These models are evaluated not only for AD recognition, but also with respect to their potential to generalise across both datasets. They provide good results when trained and tested on the same data set (72.56% UAR for monologue data and 85.21% for dialogue). A decrease in UAR is observed in transfer training, where feature extraction models trained on dialogues provide better average UAR on monologues (63.72%) than the other way around (58.94%). When the choice of classifiers is independent of feature extraction, transfer from monologue models to dialogues result in a maximum UAR of 81.04% and transfer from dialogue features to monologue achieve a maximum UAR of 70.73%, evidencing the generalisability of the feature model. Clinical relevance We present a method for automatic screening of cognitive health in dementia risk settings. The method is based on spoken language, an ubiquitous source of data, therefore being cost-efficient, non-invasive and with little infrastructure required.

Identificação da Doença de Alzheimer Através da Fala Utilizando Reconhecimento de Emoções

“…The automatic recognition of emotions has become the focus of research related to technology and health. The identification of Alzheimer's disease using paralinguistic approaches has been less explored so far, however, acoustic analysis is able to assist in understanding the subtleties of speech that may indicate the presence of the disease (9) . This work aims to contribute to the research area of signal processing for Healthcare, with a multilingual approach based on machine learning for the identification of neurodegenerative diseases, evaluating the use of emotion recognition through speech as a biomarker for classification of Alzheimer's disease.…”

Dementia Prediction in Older People through Topic-cued Spontaneous Conversation

“…Considering the development of practical home use applications, several research groups have focused on predicting mild cognitive impairment (MCI) or mild dementia onset using machine learning methods trained on large-size language corpora (Hernández-Domínguez et al, 2018;Fraser et al, 2019;Asgari et al, 2017). Recently it has become more manageable to collect and analyze large-size language datasets, such as daily conversations (Stark et al, 2020;Power et al, 2020), due to the development of applications (Otake et al, 2009;Otake-Matsuura, 2018;Otake-Matsuura et al, 2021), acoustic feature processing approaches (Fu et al, 2020;d. la Fuente Garcia et al, 2020), and natural language processing (NLP) algorithms (Bojanowski et al, 2016;Joulin et al, 2016a,b).…”