Advanced Auditory Brain–Computer Interfaces for Enhanced Speech Therapy in Children with Developmental Language Disorders

Developmental language disorders (DLDs) are among the most common early-childhood disorders, affecting between 8-12% of children worldwide. If left untreated, DLD can lead to significant professional and interpersonal difficulties later in life. While speech therapy is highly effective, the barrier-to-access is high with regards to cost, availability of speech therapists, and the delay between detection and intervention. There is therefore a significant unmet clinical need for the development of novel, easy-to-deploy interventions to alleviate the disease burden of DLD on healthcare and educational institutions, while maintaining a high standard of compassionate and personalized care. Leveraging advances in large language models (LLMs), it is possible to partially automate the delivery of auditory stimuli that caters to the individual deficits for each child, in the form of custom-made audiobooks. The efficacy of this proposed therapy will be examined by tracking the neural signatures of good speech perception using a non-invasive magnetoencephalography-based brain-computer interface. Furthermore, subjects will receive real-time neurofeedback in the form of gamified rewards on their performance (animated gold stars, points counter) while listening to these audiobooks, to enhance motivation and task engagement, with the goal of improving the efficacy of the therapy.

Faculty Supervisor:

Luka Milosevic;Andrew Dimitrijevic

Student:

Partner:

Max Planck Institute for Human Cognitive and Brain Sciences

Discipline:

Computer science

Sector:

Education

University:

University of Toronto

Program: