CAtCh: Cognitive Assessment through Cookie Thief

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Colonel, Joseph T, Hagler, Carolyn, Wismer, Guiselle, Curtis, Laura, Becker, Jacqueline, Wisnivesky, Juan, Federman, Alex, Pandey, Gaurav
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912418217590784
author Colonel, Joseph T
Hagler, Carolyn
Wismer, Guiselle
Curtis, Laura
Becker, Jacqueline
Wisnivesky, Juan
Federman, Alex
Pandey, Gaurav
author_facet Colonel, Joseph T
Hagler, Carolyn
Wismer, Guiselle
Curtis, Laura
Becker, Jacqueline
Wisnivesky, Juan
Federman, Alex
Pandey, Gaurav
contents Several machine learning algorithms have been developed for the prediction of Alzheimer's disease and related dementia (ADRD) from spontaneous speech. However, none of these algorithms have been translated for the prediction of broader cognitive impairment (CI), which in some cases is a precursor and risk factor of ADRD. In this paper, we evaluated several speech-based open-source methods originally proposed for the prediction of ADRD, as well as methods from multimodal sentiment analysis for the task of predicting CI from patient audio recordings. Results demonstrated that multimodal methods outperformed unimodal ones for CI prediction, and that acoustics-based approaches performed better than linguistics-based ones. Specifically, interpretable acoustic features relating to affect and prosody were found to significantly outperform BERT-based linguistic features and interpretable linguistic features, respectively. All the code developed for this study is available at https://github.com/JTColonel/catch.
format Preprint
id arxiv_https___arxiv_org_abs_2506_06603
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CAtCh: Cognitive Assessment through Cookie Thief
Colonel, Joseph T
Hagler, Carolyn
Wismer, Guiselle
Curtis, Laura
Becker, Jacqueline
Wisnivesky, Juan
Federman, Alex
Pandey, Gaurav
Machine Learning
Artificial Intelligence
Sound
Audio and Speech Processing
Several machine learning algorithms have been developed for the prediction of Alzheimer's disease and related dementia (ADRD) from spontaneous speech. However, none of these algorithms have been translated for the prediction of broader cognitive impairment (CI), which in some cases is a precursor and risk factor of ADRD. In this paper, we evaluated several speech-based open-source methods originally proposed for the prediction of ADRD, as well as methods from multimodal sentiment analysis for the task of predicting CI from patient audio recordings. Results demonstrated that multimodal methods outperformed unimodal ones for CI prediction, and that acoustics-based approaches performed better than linguistics-based ones. Specifically, interpretable acoustic features relating to affect and prosody were found to significantly outperform BERT-based linguistic features and interpretable linguistic features, respectively. All the code developed for this study is available at https://github.com/JTColonel/catch.
title CAtCh: Cognitive Assessment through Cookie Thief
topic Machine Learning
Artificial Intelligence
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2506.06603