CSUSM Part of Grant to Improve Speech Outcomes for Children

A child does speech therapy through a game made by Verboso. Photo courtesy Verboso

Cal State San Marcos is the co-recipient of a grant from the National Science Foundation through which it will work with an industry partner to improve outcomes for children with speech impairments.

Through the one-year, $250,293 grant from Small Business Innovation Research arm of the NSF, CSUSM will collaborate with Verboso, a Chicago-based company that uses state-of-the-art technology to create therapy video games with automated feedback.

The principal investigator for the grant from CSUSM is Alison Scheer-Cohen, an associate professor in the Department of Speech-Language Pathology who will conduct the research along with two student assistants. In partnership with San Diego Unified School District and Capistrano Unified School District, school-age students will be able to use automated technology to practice their speech at home. Scheer-Cohen and her team will collect participant recordings to produce a larger pediatric data set and refine the automated technology.

“I’m excited to be a part of a project that is building automatic technology for speech sound production practice,” Scheer-Cohen said. “Ultimately, this will provide resources for children and families to practice their speech production at home and potentially have an impact on treatment outcome.”

The target of the project is children with speech sound disorders (SSD), which means that they have trouble saying certain sounds and words past the expected age. Approximately 5% of children have an SSD, which places them at risk for reading difficulties and processing challenges into adulthood.

Less than 70% of affected children receive needed speech therapy, however, due to difficulty accessing care and cost. The NSF project will develop artificial intelligence that can recreate the feedback decisions that a speech-language pathologist makes in a live therapy setting. Because of the automation, the cost of receiving services can be drastically reduced and many more children can receive treatment.

The objectives of the study are to:

build a database of impaired and accurate productions of the eight English consonants most commonly in error;
use this database to develop and train algorithms for identifying specific errors of segmented target phonemes (units of sound that can distinguish one word from another);
complete reliability tests of the trained algorithm with decisions from trained speech-language pathologists to ensure adequate agreement of presence of error and error type.

“This grant is a great recognition of the potential of our technology to significantly improve access to, and impact the delivery of, speech therapy,” said Amy Linde, chief clinical officer for Verboso. “We are thrilled to have Dr. Scheer-Cohen and her research team at CSUSM bring their expertise in pediatric speech sound disorders and motor learning approaches to speech therapy to this project. Verboso is excited about this collaboration and working with all of our research partners on this important work.”