More than Words: Cross-Linguistic Exploration of Parkinson’s Disease Identification from Speech

Vasilisa Verkhodanova*, Dominika Trčková, Matt Coler, Wander Lowie

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

4 Citations (Scopus)

Abstract

This study investigates the effect of listeners’ first language and expertise on their perception of speech produced by people with Parkinson’s Disease (PD). We compared assessment scores and identification accuracy of expert and non-expert Czech and Dutch listeners on two tasks: perception of speech healthiness and recognition of sentence type intonation. We collected speech data from 30 Dutch speakers diagnosed with PD and 30 Dutch speaking healthy controls. Short phrases from intonation tasks and spontaneous monologues were used as stimuli in an online perception experiment. 40 people (20 expert and 20 non-expert listeners) participated. Results show that both expertise and language familiarity are important factors in perception of speech of people with PD, however differences in identification accuracy depend on the task type. In recognition of PD speech, there are prominent acoustic cues that trigger perception of “unhealthiness” in the non-expert listeners, while experience with phonetics may lead to a different focus in their perception of such cues. In intonation accuracy recognition, both Czech and Dutch expert groups outperform both non-expert groups, indicating the added value of phonetic experience for the prosodic task. Yet, there is a clear benefit for having Dutch as the first language for both tasks, as Dutch listeners performed more accurately in both healthiness and sentence type recognition tasks .
Original languageEnglish
Title of host publicationSpeech and Computer
EditorsAlexey Karpov, Rodmonga Potapova
Place of PublicationCham
PublisherSpringer International Publishing, Cham, Switzerland
Pages613-623
Number of pages11
ISBN (Electronic)9783030602765
ISBN (Print)9783030602758
DOIs
Publication statusPublished - 29-Sept-2020

Publication series

NameLecture Notes in Computer Science
PublisherSpringer Nature
Volume12335

Fingerprint

Dive into the research topics of 'More than Words: Cross-Linguistic Exploration of Parkinson’s Disease Identification from Speech'. Together they form a unique fingerprint.

Cite this