La Trobe University News (2026) – Generative AI and aphasia
Research from La Trobe University has explored how generative artificial intelligence (AI) may support communication for people living with aphasia, a language disorder that can affect speaking, understanding, reading, and writing.
Aphasia can significantly impact a person’s ability to participate in everyday conversation and social interaction. While current augmentative and alternative communication (AAC) tools provide support, they may be limited when language output is incomplete, fragmented, or contains errors.
Researchers from La Trobe University’s Centre for Data Analytics and Cognition (CDAC) and the Centre of Research Excellence in Aphasia Recovery and Rehabilitation investigated how large language models (LLMs) could assist by interpreting incomplete speech and generating clearer reconstructions of intended meaning.
The study examined a framework in which AI systems are trained to process partial or disrupted language input and produce more coherent sentence outputs while maintaining the original communicative intent as closely as possible.
The research suggests that generative AI may offer a more flexible and context-sensitive approach than traditional assistive technologies, which often rely on fixed word prediction systems or pre-programmed phrases.
Rather than replacing existing communication supports, the approach is intended to complement them by improving the ability to interpret meaning from real-world, naturalistic speech patterns.
The researchers also highlight the potential for future development using multimodal inputs, such as combining speech fragments with visual cues or gesture-based information, to further improve interpretation and communication support in practical settings.
Ongoing work in this area is focused on refining model accuracy, improving clinical relevance, and exploring safe and effective applications in real-world communication environments.
This research contributes to the growing field of AI-supported communication technologies aimed at enhancing participation, independence, and quality of life for people living with aphasia.
Source: La Trobe University News (2026), “Generative AI and aphasia”
Written by Clare Jones (B Occ Thy) Director at MediRecruit