Computer mediated speech assessment in aphasia and AOS

University Of North Carolina Chapel Hill, Chapel Hill NC

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): Few tools are available for objective and sensitive quantification of speech involvement in persons with aphasia and other language disorders. This situation limits both the pursuit of theoretically motivated research on speech production and the clinical management of speech disorders associated with left hemisphere pathology. The assessment of perceptual consequences and magnitude of involvement are particularly challenging in individuals who present with coexisting apraxia of speech (AOS). The purpose of the proposed research is to develop interactive and customized elicitation, recording, and testing procedures for perceptual and basic acoustic assessment of speech in this population. Useful applications are anticipated to the evaluation of speech in other communication disorders as well. The first year will be devoted to software development. The program will include five components: (1) selection and preparation of target utterances; (2) elicitation and recording of a speech sample; (3) duration measurement and editing of the recorded signal; (4) perceptual testing; and (5) sample storage and data management. Initial programming will target maximal flexibility and efficiency of the system. Preliminary testing will identify areas for improvement and the software will be modified accordingly. During the second and third years, a series of experiments will assess the suitability of the procedures for persons with aphasia and AOS. Speakers with a range of speech and language severity will be included. Data will be collected for critical components of the assessment protocol with regard to procedural efficiency, successful step completion, observer and test-retest reliability, and participants' perception of the protocol. Comparisons will be made between and among alternate administration procedures to allow further development of the tool and specific recommendations to maximize reliability, efficiency, and sensitivity. Finally, the validity of the procedure as a severity indicator will be evaluated. The correspondence among the duration measures and speech intelligibility scores generated by the procedure and other measures of articulatory severity will be examined. These will include clinicians' and everyday listeners' ratings of speech severity, estimation of speaking rate, and frequency of errors identified on broad phonetic transcription.