My professional life is firmly focussed on improving speech and language testing and management in progressive neurological diseases. This work is underpinned by my role as Head of Speech Pathology at The University of Melbourne, educating the next generation of clinicians, the clinical research our research group – NEUS, and our industry partner, Redenlab Inc, who collaborate closely with the pharmaceutical industry to improve drug testing protocols.

My career has been devoted to improving speech, language and swallowing function in people with progressive and acquired neurological conditions. Multiple sclerosis (MS) is a very important part of that work. The challenges faced by people with MS, their families and the medical community serving them are huge. The variability in function in MS brings challenging clinical questions that need attention.

New treatments are always needed, and this space continues to grow. Our own focus is on improving how we measure meaningful change in health and function in MS. The explosion of AI and large language models (think ChatGPT) is at the bleeding edge of clinical science – making this an exciting frontier for clinical outcome development.

MS is a chronic neurological disorder that often impairs communication abilities. Language function in individuals with MS can be impaired as a result of cognitive deficits and processing speed. This impacts quality of life. Individuals with MS often present with word-finding difficulties, leading to hesitations and non-specific language use. Verbal fluency may decline, and some people have impaired syntax and grammar, leading to simplified sentences and grammatical errors. While comprehension is generally less affected, understanding complex sentences or following lengthy conversations can be problematic, especially during fatigue. These language changes are often linked to broader cognitive deficits, such as memory, attention, and executive functions. They can also co-occur with motor speech impairments like dysarthria, that result in slower rate of speech, imprecise articulation and reduced voice quality.

We are bringing together a large established cohort of MS, The ACTIVE-MS study, "Active self-monitoring to prospectively detect treatment failure and define subclinical progression trajectories in MS”. ACTIVE-MS is a research initiative established by Professor Anneke van der Walt and Professor Helmet Butzkueven to detect change from disease and treatment in MS. Speech recordings are available in this study. They will be analysed using large language models (LLM) and Natural Language Processing (NLP), two methods for acquiring objective data on language function. This information will be used to build a model of language function designed to detect subtle differences in language performance that correlate with the severity of MS, potentially enabling earlier and more precise identification of meaningful change in performance.

No validated, objective monitoring tools for MS language function exist. Advances in AI and large language models offer promise for a new way of assessing communication. Data can be acquired efficiently, either through clinical interviews or set tasks, and outcomes can be directly linked to quality of life.

I love collaborating with colleagues with diverse skill sets from around the world. This project brings together expertise in AI, engineering, neurology and speech pathology, making it an exciting cross-disciplinary endeavour.