Digital App for Speech and Health Monitoring Study (DASH): protocol for a prospective longitudinal case-control observational study for developing speech datasets in neurodegenerative disorders and dementia.

INTRODUCTION: Neurodegenerative disorders (NDDs) represent an unprecedented public health burden. These disorders are clinically heterogeneous and therapeutically challenging, but advances in discovery science and trial methodology offer hope for translation to new treatments. Against this background, there is an urgent unmet need for biomarkers to aid with early and accurate diagnosis, prognosis and monitoring throughout the care pathway and in clinical trials.Investigations routinely used in clinical care and trials are often invasive, expensive, time-consuming, subjective and ordinal. Speech data represent a potentially scalable, non-invasive, objective and quantifiable digital biomarker that can be acquired remotely and cost-efficiently using mobile devices, and analysed using state-of-the-art speech signal processing and machine learning approaches. This prospective case-control observational study of multiple NDDs aims to deliver a deeply clinically phenotyped longitudinal speech dataset to facilitate development and evaluation of speech biomarkers. METHODS AND ANALYSIS: People living with dementia, motor neuron disease, multiple sclerosis and Parkinson's disease are eligible to participate. Healthy individuals (including relatives or carers of participants with neurological disease) are also eligible to participate as controls. Participants complete a study app with standardised speech recording tasks (including reading, free speech, picture description and verbal fluency tasks) and patient-reported outcome measures of quality of life and mood (EuroQol-5 Dimension-5 Level, Patient Health Questionnaire 2) every 2 months at home or in clinic. Participants also complete disease severity scales, cognitive screening tests and provide optional samples for blood-based biomarkers at baseline and then 6-monthly. Follow-up is scheduled for up to 24 months. Initially, 30 participants will be recruited to each group. Speech recordings and contemporaneous clinical data will be used to create a dataset for development and evaluation of novel speech-based diagnosis and monitoring algorithms. ETHICS AND DISSEMINATION: Digital App for Speech and Health Monitoring Study was approved by the South Central-Hampshire B Ethics Committee (REC ref. 24/SC/0067), NHS Lothian (R&D ref. 2024/0034) and NHS Forth Valley (R&D ref. FV1494). Results of the study will be submitted for publication in peer-reviewed journals and conferences. Data from the study will be shared with other researchers and used to facilitate speech processing challenges for neurological disorders. Regular updates will be provided on the Anne Rowling Regenerative Neurology Clinic web page and social media platforms. TRIAL REGISTRATION: ClinicalTrials.gov NCT06450418 (pre-results).