Feasibility of Technology Enabled Speech Disorder Screening

Andreas Duenser,L. Ward,A. Stefani,Daniel V. Smith,J. Freyne,A. Morgan,B. Dodd
DOI: https://doi.org/10.3233/978-1-61499-666-8-21
Abstract:One in twenty Australian children suffers from a speech disorder. Early detection of such problems can significantly improve literacy and academic outcomes for these children, reduce health and educational burden and ongoing social costs. Here we present the development of a prototype and feasibility tests of a screening and decision support tool to assess speech disorders in young children. The prototype incorporates speech signal processing, machine learning and expert knowledge to automatically classify phonemes of normal and disordered speech. We discuss these results and our future work towards the development of a mobile tool to facilitate broad, early speech disorder screening by non-experts.
What problem does this paper attempt to address?