Detection of bulbar amyotrophic lateral sclerosis based on speech analysis

Abstract

The approach to analysis of speech signal in order to amyotrophic lateral sclerosis diagnosis is proposed. As a features of pathological changes in the voice the results of a format analysis of vowel /aa/ and /iy/, the distance between the envelopes of these vowel, and information about pauses between words in the test speech signal were used. To perform the classification, the linear discriminant analysis was used.

Keywords

• speech signal
• amyotrophic lateral sclerosis
• diagnosis
• classification