Chrystalina Antoniades

AI digital biomarkers A number of brain diseases affect various types of bodily movements, either producing unwanted movements such as tremors or reducing movement by making the patient stiff and slow, or sometimes both. One such condition is Parkinson’s disease (PD), a common neurodegenerative disorder affecting over 10 million people worldwide. Another is Progressive Supranuclear Palsy (PSP), a much rarer condition with more rapid progression. Despite these conditions having a very different pathological basis they can appear very similar in their early stages. In our research clinics, we use a variety of quantitative experimental methods, based on precise measurement of subtle abnormalities of the speed and coordination of various movements such as saccades (fast eye movements), hand movements, and gait. This yields a large volume of data which we analyse using both classical machine learning techniques (ML) and deep neural networks (DL). The aim of this work is to improve our ability to accurately diagnose conditions like PD and PSP at the earliest possible stage, and then monitor their progression. At present, all treatments for PD and PSP are symptomatic only, but multiple potentially disease-modifying drugs are in development. The tools we are developing to monitor progression will be critical in their evaluation in trials, and our diagnostics will then facilitate their use clinically at as early a disease stage as possible. DPhil AND MSC Projects I'm happy to consider applications from prospective DPhil and MSc applicants. Please visit our group website (NeuroMetrology Group) for further information about the projects we are carrying out currently.

[2026] Multimodal mapping of balance dysfunction in Parkinson's disease: a consensus roadmap for research and intervention, Shaikh AG. et al, Current Opinion in Neurology, 39, 403 - 414 [2026] Cognitive function is associated with the progression of non-tremor motor symptoms in Parkinson’s disease, Gibbs MC. et al, Frontiers in Aging Neuroscience, 18 [2025] Daily steps are a predictor of, but perhaps not a risk factor for Parkinson’s disease: findings from the UK Biobank, Acquah A. et al, npj Parkinson's Disease, 11 [2025] Volumetric spline-based Kolmogorov-Arnold architectures surpass CNNs, vision transformers, and graph networks for Parkinson’s disease detection, Patel SB. et al, Communications Medicine, 5 [2025] Comparative analysis of trunk sensor placement in people with and without P