Quantifying perturbations in temporal dynamics for automated assessment of spastic dysarthric speech intelligibility

Abstract

Spastic dysarthric speech is often associated with imprecise placement of articulators which, in turn, cause perturbations in speech temporal dynamics, such as unclear distinctions between adjacent phonemes. While these perturbations can lead to a significant reduction in intelligibility, measures to objectively assess their detrimental effect on intelligibility are lacking. In this paper, short-and long-term temporal dynamics measures are proposed and evaluated as correlates of subjective intelligibility. The former is based on log-energy temporal dynamics information, whereas the latter is based on an auditory-inspired modulation spectral signal representation. A composite measure is also developed based on linearly combining the proposed measures with a tone-unit duration parameter. Experiments with the publicly-available 'Universal Access' database of spastic dysarthric speech show that the proposed composite measure can achieve rank correlations with subjective ratings as high as 0.87, thus providing a tool to automatically diagnose speech disorder severity and to evaluate dysarthria treatment outcomes.

Utterance Verification-Based Dysarthric Speech Intelligibility Assessment Using Phonetic Posterior Features

“…Previous work on objective dysarthric speech intelligibility assessment can be broadly grouped as: i) assessment without explicit use of linguistic information: Legendre et al proposed prediction of intelligibility using amplitude modulation spectra [3]. In [4], Falk et al investigated modeling of short-and long-term temporal dynamics information. In [5], inspired from the notion that intelligibility can be expressed as a linear combination of perceptual dimensions phonation, nasality, articulation and prosody [6], a signal processing-based composite measure was proposed.…”

Spectral Subspace Analysis for Automatic Assessment of Pathological Speech Intelligibility

“…Automatic pathological speech intelligibility assessment approaches can be broadly categorized into i) blind approaches which do not require any healthy (intelligible) speech signals [3][4][5][6][7][8][9][10] and ii) non-blind approaches which exploit information about intelligible speech from healthy speakers [11][12][13][14][15][16][17][18][19][20].…”

“…Blind intelligibility assessment approaches typically refer to extracting several acoustic features that are believed to be correlated with intelligibility, such as the range of the fundamental frequency or the low-to-high modulation energy ratio [6,7]. Intelligibility scores are then estimated by combining multiple features via feature selection and regression training [3][4][5][6][7][8][9][10]. Non-blind approaches encompass a wide range of approaches where healthy reference signals are exploited in different manners.…”

Pathological Speech Intelligibility Assessment Based on the Short-time Objective Intelligibility Measure

“…Finally, feature-based approaches typically refer to the blind assessment (not requiring any reference signal or other information such as phone boundaries) of speech intelligibility by extracting several acoustic features such as pitch range or voiced frames percentage. Using feature selection and regression training, an intelligibility measure is then derived [14][15][16][17][18][19]. In many of these approaches, rigorous validation strategies have not been followed.…”