Background/Purpose: Early diagnosis of speech disorders, combined with intensive treatment, significantly accelerates recovery. However, a shortage of speech-language pathologists (SLPs) limits the availability of adequate therapy services. Advances in AI technology offer potential solutions to this challenge by providing automated tools that assist SLPs in managing and monitoring clients remotely. This proposal introduces a remote therapy tool designed to support SLPs in delivering client exercises, tracking progress, and generating reports on pronunciation quality.
Methods: The core of the tool is an AI-driven model that analyzes speech sounds and generates assessment reports at the phoneme feature level, evaluating placement and manner of articulation. The system models typical articulatory features and detects any deviations or missing features in the client's pronunciation. It provides SLPs with detailed feedback that identifies specific mispronunciations and articulatory errors, allowing for targeted intervention.
Results: The tool will deliver highly accurate assessments of client speech, providing insight into the articulatory configurations responsible for pronunciation errors. By automating the tracking of progress and providing detailed reports, the system allows SLPs to manage a larger number of clients while still delivering personalized feedback.
Conclusions: This AI-driven remote therapy tool addresses the shortage of SLPs by augmenting their ability to diagnose and treat speech disorders. The system's detailed, automated reports enhance the quality of therapy services, enabling earlier diagnosis and more focused treatment. This could significantly accelerate recovery times and improve the overall effectiveness of speech therapy."