Evaluation of the Speech Viewer Computer-based Speech Training System with Neurologically Impaired Individuals Evaluation du systeme informatise de la parole Speech Viewer avec des personnes souffrant de troubles neurologiques

There are few research studies documenting the effectiveness of Computer-Based Speech Training (CBST) systems. This paper evaluates the effectiveness of the IBM SpeechViewer CBST combined with a newly developed rate control program (The Stepping Stones Game) for improving speech production in individuals with dysarthria. Three subjects were studied using a single-subject multiple baseline design. Voice timing, speaking rate, and vowel accuracy were trained sequentially. Dependent measures included: speech inteJligibiIity, speaking rate, and acoustic measures obtained on the selected speech parameters. Four of the six multiple baseline treatment phases analyzed statistically revealed significant training effects. The other two phases revealed trends towards normal values. Results indicate that for the three subjects studied the Speech Viewer and the Stepping Stones Game were effective tools for modifying speech attributes and improving intelligibility. Resume 11 y a eu peu d' etudes de l' efficaciti des systemes informatises d' education de la parole. L' article evalue l' efficacite du sysfeme SpeechViewer d'/BM combine avec un nouveau programme de control du debit (Stepping Stones Game) pour ameliorer la production verbale chez les personnes souffrant de dysarthrie. On a etudie trois sujets a /' aide d' un modele de base multiple pour un sujet unique. On a fait consecutivement I' education du minutage de la parole, du debit de la parole et de I' exactitude des voyelles. On a notamment utilise les mesures dependantes suivantes: intelligibilite de la parole, debit de la parole et mesures acoustiques obtenues sur certains para metres de la parole. Quatre des six phases du traitemeiit de base multiple ont produit des effets significatifs de l' education sur le plan statistique. Les deux autres phases ont revile des tendances vel's les valeurs normales. Les nJsultats indiquent que, pour les trois sujets itudies, le Speech Viewer et le Stepping Stones Game ont ete des instruments efficaces pour modifier les attributs de la parole et ameliorer l'intelligibilite. JSLPA Vol, 15, No. 4. December 1991 I ROA Vol. 15. N0 4, decembre 1991 Michael McClean Graduate Department of Speech Pathology Dysarthria is a neuromuscular motor speech disorder that can range in severity from a disorder so mild that it is just noticeable during connected speech, to a disorder so severe that no functional speech is present. Dysarthria is present in approximately 33% of all patients with traumatic brain injuries (Sarno, Buonagura, & Levita, 1986); 68% of individuals with cerebral palsy (Hopkins, Bice, & Colton, 1954); and varies from 19% to 100% in individuals with degenerative neurological diseases, such as Multiple Sclerosis, Parkinson's Disease, and Amyotrophic Lateral Sclerosis (Darley, Brown, & Goldstein, 1972; Logemann & Fisher, 1981; Darley 1978). Dysarthria often results in reduced speech intelligibility. Individuals with reduced speech intelligibility experience difficulties with social interaction, academic performance, and vocational placement (Simon, 1985). Improving speech intelligibility is the primary clinical goal with dysarthric speakers (Yorkston, Beukelman, & Traynor, 1984). For mild dysarthric speakers, the goal may shift to maximizing speech naturalness, while maintaining intelligibility (Y orkston, Beukelman, & Bell 1988). Computer-Based Speech Training (CBST) has considerable potential for enhancing clinical practice in this area. These systems provide visual feedback on aspects of speech, such as pitch, loudness, voicing, and vowel productions. They also allow the clinician to obtain objective measures of speech production. The visual feedback helps patients develop insight into their speech difficulties and may increase the efficiency of the remediation process. There are few studies documenting the effectiveness of CBST (Bernstein, Goldstein, & Mahshie, 1988). Only 5-10% of CBST systems have been tested in controlled experiments, and few of these systems have been tested by investigators other than their developers (Watson & Kewley-Port, 1989). One reason for the scarcity of such research is that it is often the most time-consuming and labour intensive aspect of CBST development. Unless carefully collected data support the