Existing approaches to diagnosis of Parkinson’s disease are based on clinical features, which are often very subjective (e.g. UPDRS; the Unified Parkinson’s Disease Rating Scale). A NICE report in 2006 revealed that these existing methods have poor sensitivity, with up to 25% error in diagnosis. An alternative diagnostic approach, using PET scans to detect decreased dopamine activity is not 100% reliable and is impractical on a large scale due to cost.
The University of York is undertaking research to develop a novel way to diagnose this condition, using a non-invasive task-based approach. The aim is to identify artefacts by observing particular behaviour in velocity profiles that are unique to Parkinson’s disease patients. One such artefact indicates a two-stage acceleration in the drawing activity consistent with the hesitation associated with Bradykinesia. Previously, this would be achieved through visual inspection, but such an inspection is less quantitative and more subjective (and therefore less repeatable) than our proposed approach.
An evolutionary algorithm developed at York is used to perform the identification of artefacts, and patients are then classified according to the number of such artefacts observed in their tests.
Current tests (including ‘leave out one’ analysis) with a small group (12 patients with idiopathic Parkinson’s disease and 10 controls) indicate very encouraging results, and research work continues to evaluate and refine the technique.
Diagnosis of idiopathic Parkinson’s Disease
Conversations with potential collaborators or licencees are welcomed
IP is in place to protect the in-house developed software