A Speech Prosody-Based Approach to Early Detection of Cognitive Impairment in Elderly Subjects: A Preliminary Study

A Speech Prosody-Based Approach to Early Detection of Cognitive Impairment in Elderly Subjects: A Preliminary Study

Shohei Kato (Graduate School of Engineering, Department of Computer Science and Engineering, Nagoya Institute of Technology, Japan), Sachio Hanya (Graduate School of Engineering, Department of Computer Science and Engineering, Nagoya Institute of Technology, Japan), Akiko Kobayashi (Ifcom Co., Ltd., Japan), Toshiaki Kojima (Ifcom Co., Ltd., Japan), Hidenori Itoh (Graduate School of Engineering, Department of Computer Science and Engineering, Nagoya Institute of Technology, Japan) and Akira Homma (Tokyo Dementia Care Research and Training Center, Japan)
DOI: 10.4018/978-1-60960-559-9.ch024

Abstract

This chapter presents a novel approach for early detection of cognitive impairment in the elderly. The approach incorporates the use of speech sound analysis and multivariate statistical techniques. The focus is on the prosodic features of speech. One hundred and fifteen Japanese subjects (32 males and 83 females between the ages of 38 and 99 years) participated in this study. The authors collected speech sounds from segments of dialogue during an HDS-R examination. The segments correspond to speech sounds from answers to questions about time orientation and number counting. One hundred and thirty prosodic features were extracted from each of the speech sounds. These prosodic features consisted of spectral and pitch features (53), formant features (56), intensity features (19), and speech rate and response time (2). These features were refined by principal component analysis and/or feature selection. In addition, the authors calculated speech prosody-based cognitive impairment rating (SPCIR) by multiple linear regression analysis. The results indicate that a moderately significant correlation exists between the HDS-R score and the synthesis of several selected prosodic features. Consequently, the adjusted coefficient of determination (= 0.50) suggests that prosody-based speech sound analysis could potentially be used to detect cognitive impairment in elderly subjects.
Chapter Preview
Top

Method

Design

We recorded the speech sound of elderly patients while they provided answers for an HDS-R questionnaire test. We focused on questions about time orientation and numbering. In addition, we collected speech sounds while the patients were talking about the topics of hometown, childhood, and school.

Participants

One hundred and fifteen Japanese subjects (32 males and 83 females between the ages of 38 and 99) participated in this study. With some exceptions, we collected three samples of speech sound from each of the participants. The number of total sound data points was 319, as shown in Table 1. The sound data contained 205 samples of speech by elderly patients whose HDS-R score was 30-24 (NL) and 114 samples from patients with cognitive impairment whose HDS-R score was 23-11 (CI).

Table 1.
Category breakdown of the speech data (N=115)
Age30’s40’s50’s60’s70’s80’s90’sTotal
Male3 (1)0 (0)15 (5)32 (11)21 (7)12(5)7 (3)90 (32)
Female0 (0)20 (7)45 (15)24 (8)28 (10)87 (33)25 (10)229 (83)
Subtotal3 (1)20 (7)60 (20)56 (19)49 (17)99 (38)32 (13)319 (115)

Value in bracket means the number of subjects.

Complete Chapter List

Search this Book:
Reset