Table 1
Perceptual Features Rated for Different Speaking Tasks in Each Perceptual Domain.
| PERCEPTUAL DOMAIN | SPEAKING TASK | PERCEPTUAL FEATURES RATED |
|---|---|---|
| Speech timing | Sentences, spontaneous speech | Slow rate, fast rate, variable rate, short phrases, prolonged interword intervals, atypical pauses/silences, prolonged phonemes |
| Fluency | Sentences, spontaneous speech | Abnormal noises that interrupt speech or occur when patient is not speaking, stutter-like disfluencies, distorted substitutions or articulatory additions, syllable segregation |
| Prosody | Sentences, spontaneous speech | Reduced use of stress, prosodic excess or scanning, errors marking stress, monopitch, monoloudness |
| Articulation | Sentences, spontaneous speech | Imprecise consonants, distorted vowels, irregular articulatory breakdown, articulatory groping, telescoping, deterioration of speech during continuous speaking |
| Resonance | Sustained vowels, sentences, spontaneous speech | Audible nasal emission or nasal snorting, hyponasality, hypernasality |
| Voice Quality | Sustained vowels, sentences, spontaneous speech | Breathiness, aphonia, hoarseness, strained/strangled, diplophonia, harshness, voice stoppage, vocal tremor, rapid vocal flutter |
| Loudness | Sustained vowels, sentences, spontaneous speech | Reduced loudness, explosive loudness bursts, loudness decay, excess loudness variation |
Table 2
Acoustical Analyses of Speech and Voice Features.
| PERCEPTUAL DOMAIN | PERCEPTUAL FEATURE | ACOUSTICAL CORRELATE | SPEAKING TASK | ANALYSIS TOOL | ANALYSIS DESCRIPTION |
|---|---|---|---|---|---|
| Speech timing | Slow rate of speech | Speech rate | Spontaneous speech | Publicly-available Praat script (Praat Vocal Toolkit) [54] | Number of syllables divided by the entire utterance duration (including silent and filled pauses and other dysfluencies like syllable repetitions) [34] |
| Articulation rate | Spontaneous speech | Publicly-available Praat script (Praat Vocal Toolkit) [54] | The number of syllables divided by the phonation duration, which excluded silent pauses [34, 54] | ||
| Prolonged inter-word intervals | Silent pause duration | Spontaneous speech | Publicly-available Praat script (Praat Vocal Toolkit) [54] | Duration of silences at least 0.3 s long | |
| Filled pause duration | Spontaneous speech | Custom-written Python script | Number and duration of filled pauses (“uh”, “uhm”, etc.) obtained from transcribed .txt files and corresponding TextGrids generated using Montreal Forced Aligner | ||
| Prolonged phonemes | Syllable duration | Spontaneous speech | Publicly-available Praat script (Praat Vocal Toolkit) [54] | Mean duration of syllables | |
| Variable rate of speech, imprecise consonants and other articulatory inaccuracies | Variability of syllable duration | Spontaneous speech | Custom-written Praat script | SD of syllable duration obtained from a TextGrid generated using Corretge (2021–2024). | |
| Prosody | Prosodic excess | Variability of fo | Spontaneous speech | Custom-written Praat script.1 | SD of fo computed using autocorrelation |
| Voice Quality | Vocal tremor | fo modulation rate and extent, intensity modulation rate and extent | Sustained vowels | Custom-written Praat scripts [55] | Number of cycles of fo modulation per second and magnitude of fo modulation for the middle 1 s segment of each sustained /ɑ/ and /i/2 |
| Harshness, Breathiness, Hoarseness | Smoothed cepstral peak prominence (CPPS) | Sustained vowels | Publicly-available Praat script (Praat Vocal Toolkit) | A measure of periodic energy in the voice signal from the middle 1 s segment of the second sustained /ɑ/ produced by each participant [56] |
[i] 1 A template of this Praat script was generated using OpenAI (2025) using the prompt “generate a Praat script to calculate SD of fundamental frequency” and then modified by the first author.
2 Semi-automated analyses of vocal tremor were completed for participants who exhibited rhythmic modulation of fo or intensity based on visual inspection of the contours in Praat.
Figure 1
Perceptual Features of Dysarthria Across Speaking Tasks.
Figure 2
Speech Timing Measures in Spontaneous Speech Samples from Participants With ET.
A. Speech rate (syllables/s).
B. Articulation rate (syllables/s).
C. Silent pause duration (s).
D. Filled pause duration (s).
The solid blue box represents the interquartile range of the speech timing measure, with the black dots representing individual participants and the black horizontal lines representing the group mean. The dashed orange line marks the normative mean, and the shaded orange region indicates the normative range (±1 SD from the mean). Square brackets on the y-axis indicate patterns associated with ataxic or hyperkinetic dysarthria.
Figure 3
Syllable Duration Measures in Spontaneous Speech Samples from Participants with ET.
A. Mean syllable duration
B. SD of syllable duration
The solid blue boxes represent the interquartile range, with the black dots representing individual participants and the black horizontal lines representing the group mean. The dashed orange line in A. marks the normative mean, and the shaded orange region indicates the normative range (±1 SD from the mean). Because the mean SD of syllable duration was not reported in previous studies, the shaded orange region indicates the normative range only in B. Square brackets on the y-axis indicate patterns associated with ataxic or hyperkinetic dysarthria.
Figure 4
The fo Variability in Spontaneous Speech Samples of Female and Male Participants with ET.
The solid blue boxes represent the interquartile ranges, with the black dots representing individual participants and the black horizontal lines representing the group mean. The dashed orange line marks the normative mean, and the shaded orange region indicates the normative range (±1 SD from the mean). Square brackets on the y-axis indicate ranges associated with ataxic or hyperkinetic dysarthria.
Figure 5
CPPS for Female and Male Participants from Sustained Vowels.
The solid orange boxes represent the interquartile range, with the black dots representing individual participants and the black horizontal lines representing the group mean. The dashed blue line marks the normative mean, and the shaded blue region indicates the normative range (±2 SD from the mean).
Table 3
Participant-level means of the acoustical features illustrated in Figures 2, 3, 4, 5. Dysarthria feature classifications are indicated in parentheses: (H) hyperkinetic, (A) ataxic, and (B) both hyperkinetic and ataxic.
| PARTICIPANT | SEX | AGE (years) | SPEECH RATE (syllables/s) | ARTICULATION RATE (syllables/s) | AVG SILENT PAUSE DURATION (s) | AVG FILLED PAUSE DURATION (s) | AVG SYLLABLE DURATION (s) | SD OF SYLLABLE DURATION | SD OF fo | CPPS (dB) |
|---|---|---|---|---|---|---|---|---|---|---|
| P1 | F | 68 | 3.0 | 3.4 (B) | 0.4 | 0.2 | 0.3 (A) | 0.1 (A) | 35.8 | 10.9 |
| P2 | M | 62 | 4.9 | 4.9 | 0.0 | 0.4 | 0.2 | 0.1 (A) | 79.9 (A) | 12.8 |
| P3 | M | 65 | 3.9 | 4.5 | 0.5 | 0.4 | 0.2 (A) | 0.1 (A) | 18.2 | 9.6 (H) |
| P4 | F | 77 | 4.0 | 4.7 | 0.6 | 0.4 | 0.2 | 0.1 (A) | 26.2 | 11.6 |
| P5 | M | 80 | 3.3 | 3.3 (B) | 0.0 | 0.5 | 0.3 (A) | 0.3 (A) | 44.1 | 9.3 (H) |
| P6 | M | 69 | 4.1 | 4.1 (B) | 0.0 | 0.6 | 0.2 (A) | 0.1 (A) | 22.9 | 10.2 (H) |
| P7 | M | 82 | 2.8 (B) | 4.6 | 0.9 (H) | 0.0 | 0.2 | 0.1 (A) | 52.3 (A) | 9.8 (H) |
| P8 | F | 82 | 3.3 | 4.2 (B) | 0.7 | 0.7 | 0.2 (A) | 0.1 (A) | 48.7 (A) | 7.2 (H) |
| P9 | F | 76 | 3.0 | 3.0 (B) | 0.0 | 0.4 | 0.3 (A) | 0.2 (A) | 40.8 | 3.7 (H) |
| P10 | F | 83 | 4.1 | 4.8 | 0.7 | 0.0 | 0.2 | 0.1 (A) | 60.0 (A) | 9.1 (H) |
| P11 | F | 81 | 1.9 (B) | 3.3 (B) | 1.2 (H) | 0.4 | 0.3 (A) | 0.2 (A) | 30.2 | 11.9 |
| P12 | F | 83 | 2.9 (B) | 4.2 (B) | 0.5 | 0.0 | 0.2 (A) | 0.1 (A) | 41.5 | 5.7 (H) |
| P13 | F | 88 | 4.6 | 4.7 | 0.0 | 0.4 | 0.2 | 0.1 (A) | 99.9 (A) | 8.2 (H) |
| P14 | M | 67 | 3.4 | 4.5 | 0.3 | 0.4 | 0.2 (A) | 0.1 (A) | 23.2 | 6.5 (H) |
| P15 | M | 75 | 3.3 | 4.1 (B) | 0.2 | 0.4 | 0.2 (A) | 0.1 (A) | 33.6 (A) | 6.9 (H) |