Parkinson's disease is a progressive neurodegenerative disorder. The biggest risk factor for developing Parkinson's disease is age and so prevalence is increasing in countries where the average age of the population is rising. Cognitive problems are common in Parkinson's disease and identifying those with the condition who are most at risk of developing such issues is an important area of research. In this work, we explore the potential for using objective, automated methods based around a simple figure copying exercise administered on a graphics tablet to people with Parkinson's disease. In particular, we use a multi-objective evolutionary algorithm to explore a space of regression models, where each model represents a combination of features extracted from a patient's digitised drawing. The objectives are to accurately predict clinical measures of the patient's motor and cognitive deficit. Our results show that both of these can be predicted, to a degree, and that certain sub-sets of features are particularly relevant in each case.
|Title of host publication||2016 IEEE Symposium Series on Computational Intelligence|
|Publication status||Published - Dec 2016|
|Event||2016 IEEE Symposium on Computational Intelligence in Healthcare and e-health - Athens, Greece|
Duration: 6 Dec 2016 → 9 Dec 2016
|Conference||2016 IEEE Symposium on Computational Intelligence in Healthcare and e-health|
|Abbreviated title||IEEE SSCI 2016|
|Period||6/12/16 → 9/12/16|