Automated analysis of verbal fluency and naming as a predictor of severity in Parkinson’s disease

Parkinson’s disease (PD) is a prevalent and disabling neurodegenerative disorder. Although incurable, early detection can improve the quality of life of patients and their caregivers, while reducing the financial burden on families and healthcare systems. Typical markers for early detection arise from clinical, neuroimaging and biological assessments, which are invasive, expensive or inaccurate. Promisingly, psycholinguistic tests circumvent such limitations and allow revealing early deficits that discriminate between this and other diseases, differentiate between patients with and without mild cognitive impairment (MCI), and predict the severity of their motor and cognitive symptoms. However, psycholinguistic evidence on PD is limited, comes from small samples and arises from tasks that are not employed in clinical settings, which undermines the scalability of the approach. In this interdisciplinary project, we will address such challenges through an unprecedented approach for PD: the automated analysis of verbal fluency (production of words that meet a phonological or semantic criterion, for one minute) and naming (production of the words that name each of 60 pictures) tasks.