Abstract
Early identification of the onset of neurological disease is critical for testing drugs or interventions to halt or slow progression. Speech production has been proposed as an early indicator of neurological impairment. However, for speech to be useful for early detection, speech changes should be measurable from uncontrolled conversational speech collected passively in natural recording environments over extended periods of time. Such longitudinal speech data sets for testing the robustness of algorithms are difficult to acquire. In this paper, we exploit YouTube interviews from Muhammad Ali from 1968 to 1981, before his 1984 diagnosis of parkinsonism1. The interviews are unscripted, conversational in nature, and of varying fidelity. We measured changes in speech production from the Ali interviews and analyzed these changes relative to a coded registry of blows Mr. Ali received in each of his boxing matches over time. This provided a rich and unique opportunity to evaluate speech change as both a function of disease progression and as a function of fight history. Multivariate analyses revealed changes in prosody and articulation consistent with hypokinetic dysarthria over time, and a relationship between reduced speech intonation and the amount of time elapsed since the most recent fight preceding the interview.
Original language | English (US) |
---|---|
Pages (from-to) | 1809-1813 |
Number of pages | 5 |
Journal | Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH |
Volume | 2017-August |
DOIs | |
State | Published - 2017 |
Event | 18th Annual Conference of the International Speech Communication Association, INTERSPEECH 2017 - Stockholm, Sweden Duration: Aug 20 2017 → Aug 24 2017 |
Keywords
- Articulation
- Muhammad Ali
- Parkinson's disease
- Parkinsonism speech production
- Prosody
ASJC Scopus subject areas
- Language and Linguistics
- Human-Computer Interaction
- Signal Processing
- Software
- Modeling and Simulation