Dysarthric speech evaluation: automatic and perceptual approaches

17705

Authors

nan

Abstract

Perceptual evaluation is still the most common method in clinical practice for the diagnosis and monitoring of the condition progression of people suffering from dysarthria (or speech disorders more generally). Such evaluations are frequently described as non-trivial, subjective and highly time-consuming (depending on the evaluation level). Clinicians have, therefore, expressed their need for new objective evaluation tools more adapted to longitudinal studies or rehabilitation context. We proposed earlier an automatic approach for the anomaly detection at the phone level for dysarthric speech. The system behavior was studied and validated on different corpora and speech styles. Nonetheless, the lack of annotated French dysarthric speech corpora has limited our capacity to analyze some aspects of its behavior, and notably,its severity (more anomalies detected automatically compared with human expert). To overcome this limitation, we proposed an original perceptual evaluation protocol applied to a limited set of decisions made by the automatic system, related to the presence of anomalies. This evaluation was carried out by a jury of 29 non-naive individuals. In addition to interesting information related to the system behavior, the evaluation protocol highlighted the difficulty for a human, even expert, to apprehend and detect deviations at the word level in dysarthric speech. Keywords: Dysarthria, speech disorders, automatic speech processing, perceptual evaluation 1.

Keywords

nan

Full Content

Dysarthric speech evaluation: automatic and perceptual approaches Imed Laaridh1,2, Christine Meunier2, Corinne Fredouille1 1University of Avignon, CERI/LIA, France 2Aix Marseille Universit´e, CNRS, LPL UMR 7309, 13100, Aix-en-Provence, France imed.laaridh@alumni.univ-avignon.fr, christine.meunier@lpl-aix.fr, corinne.fredouille@univ-avignon.fr Abstract Perceptual evaluation is still the most common method in clinical practice for the diagnosis and monitoring of the condition progression of people suffering from dysarthria (or speech disorders more generally). Such evaluations are frequently described as non-trivial, subjective and highly time-consuming (depending on the evaluation level). Clinicians have, therefore, expressed their need for new objective evaluation tools more adapted to longitudinal studies or rehabilitation context. We proposed earlier an automatic approach for the anomaly detection at the phone level for dysarthric speech. The system behavior was studied and validated on different corpora and speech styles. Nonetheless, the lack of annotated French dysarthric speech corpora has limited our capacity to analyze some aspects of its behavior, and notably,its severity (more anomalies detected automatically compared with human expert). To overcome this limitation, we proposed an original perceptual evaluation protocol applied to a limited set of decisions made by the automatic system, related to the presence of anomalies. This evaluation was carried out by a jury of 29 non-naive individuals. In addition to interesting information related to the system behavior, the evaluation protocol highlighted the difficulty for a human, even expert, to apprehend and detect deviations at the word level in dysarthric speech. Keywords: Dysarthria, speech disorders, automatic speech processing, perceptual evaluation 1. Introduction Dysarthria is a motor speech disorder that is a consequence of neurological damage located in the central or/and the pe- ripheral nervous system. This may result in disturbances in any of the components involved in speech production and may be reflected by weakness, spasticity, incoordina- tion, involuntary movements, or abnormal muscle tone de- pending on the location of the neurological damage (Duffy, 2005). Perceptual evaluation by a set of listeners is the most common paradigm used to evaluate the characteristics and severity of impairment in speech pathologies. Furthermore, the current classification of dysarthria into six categories is based on a perceptual description of deviant speech dimen- sions. ((Darley et al., 1969b; Darley et al., 1969a; Dar- ley et al., 1975). The clinical evaluation of patients is also based on several batteries of tests in which the production of dysarthric speakers is coded perceptually by clinicians. These batteries evaluate the vocal quality, phonetic realiza- tions, prosody, respiration and intelligibility. The BECD (Batterie d’Evaluation Clinique de la Dysarthrie) (Auzou and Rolland-Monnoury, 2006) is the most commonly used test by clinicians for French speech. This test differentiates 35 items in order to characterize dysarthria. Consequently, the use of perception for the evaluation of dysarthric speech is frequent and well documented. And the clinicians eval- uating the speech of patients are very well trained to the phonetic characteristics associated with the physiopathol- ogy of dysarthria. However, a frequent criticism to percep- tual evaluation is the subjectivity of the listeners (both naive and expert). In some previous work, the authors have proposed …

Content truncated to 500 words

Labeled Text

Perceptual evaluation is still the most common method in clinical practice for the diagnosis and monitoring of the condition progression of people suffering from dysarthria (or speech disorders more generally). Such evaluations are frequently described as non-trivial, subjective and highly time-consuming (depending on the evaluation level). Clinicians have, therefore, expressed their need for new objective evaluation tools more adapted to longitudinal studies or rehabilitation context. We proposed earlier an automatic approach for the anomaly detection at the phone level for dysarthric speech. The system behavior was studied and validated on different corpora and speech styles. Nonetheless, the lack of annotated French dysarthric speech corpora has limited our capacity to analyze some aspects of its behavior, and notably,its severity (more anomalies detected automatically compared with human expert). To overcome this limitation, we proposed an original perceptual evaluation protocol applied to a limited set of decisions made by the automatic system, related to the presence of anomalies. This evaluation was carried out by a jury of 29 non-naive individuals. In addition to interesting information related to the system behavior, the evaluation protocol highlighted the difficulty for a human, even expert, to apprehend and detect deviations at the word level in dysarthric speech. Keywords: Dysarthria, speech disorders, automatic speech processing, perceptual evaluation 1.

Publication Details

Paper ID:
17705
Year:
2018
Publication:
TIPSTER
Publication Type:
conf
DOI:
nan