Predicting Intelligibility Deficits in Parkinson's Disease With Perceptual Speech Ratings

Abstract

Purpose This study investigated whether perceptual ratings of speech parameters were predictive of transcription intelligibility in quiet and in noise for speakers with Parkinson's disease (PD). Method Ten speakers with PD and five healthy controls read 56 sentences. One group of 60 listeners orthographically transcribed the sentences in quiet, and another group of 60 listeners transcribed in noise. An additional 23 listeners judged a variety of speech parameters, including articulation, prosody, resonance, voice quality, and ease of understanding on a visual analog scale. Scores of the visual analog scale ratings were regressed against transcription intelligibility in quiet and in noise. Results Perceptual ratings of all the speech parameters were lower for speakers with PD. Global speech understanding, indexed by ease of understanding ratings, was associated with transcription intelligibility in quiet and in noise with a stronger effect in noise. Among the rated speech parameters, ease of understanding and voice quality ratings were significant predictors of speech intelligibility in noise. Conclusions Speech in individuals with PD was more difficult for listeners to understand and was characterized by deficits in articulation, prosody, resonance, and voice quality compared to normal speech produced by healthy older adults. Ease of understanding ratings, even when performed in quiet, predicted intelligibility in noise. Poor voice quality ratings in PD, a sign of phonatory deficit, had a negative impact on intelligibility in noise for speakers with PD.

Automatic Speech Recognition in Noise for Parkinson's Disease: A Pilot Study

“…Prolonged silences (i.e., over 500 ms) were trimmed and equalization of the audio spectrum in a moving window was subsequently performed. The equalized audios were combined to create the final audio file, which contained an equal number of male and female speakers (Moya-Galé et al, 2018 ), which resulted in 10-talker babble (5 males, 5 females; Chiu and Neel, 2020 ).…”

“…Intelligibility in individuals with PD is particularly affected in noisy environments, such as dining out at a restaurant or in social gatherings. What is more, a recent study showed that even those individuals with a mild speech disorder may experience a reduction in their intelligibility in the presence of background noise (Chiu and Neel, 2020 ). Multi-talker babble noise is a form of structured background noise that may mask a target speaker's voice.…”

“…The goal for this pilot study was to measure speech intelligibility in individuals with Parkinson's Disease using ASR in noise. To this end we report the sensitivity of Google Cloud speech-to-text API, a prominent provider of ASR, to a specific type of background noise, multi-talker babble, which is commonly implemented in the study of dysarthria (Moya-Galé et al, 2018 ; Chiu and Neel, 2020 ).…”

Automatic Assessment of Intelligibility in Noise in Parkinson Disease: Validation Study

“…For each recorded sentence, the app automatically embedded the speakers' voice signal into +6-dB signal-to-noise multitalker babble noise [33] to provide an intelligibility score, defined as the percentage of words accurately understood by the ASR system. Automatic feedback on performance was provided at the end of the recording session and not after each sentence to avoid any potential priming effects that could influence sentence production on subsequent items [34].…”

Single Word Intelligibility of Individuals with Parkinson’s Disease in Noise: Pre-Specified Secondary Outcome Variables from a Randomized Control Trial (RCT) Comparing Two Intensive Speech Treatments (LSVT LOUD vs. LSVT ARTIC)

“…These speech changes in prosody (loudness) and articulation have been reported to lead to significant declines in functional communication, communicative participation, and quality of life [6,22,[46][47][48][49][50][51][52][53][54][55]. Neither medical (neuropharmacological or neurosurgical) nor traditional speech treatments, which focus on a range of motor speech symptoms (e.g., respiration, articulation, speech rate, loudness, intonation) in a low dosage mode, have proved consistently or significantly beneficial in improving the degenerative speech or intelligibility of people with PD [56][57][58][59].…”