Pursuit Regularity Index for Parkinson's Disease Detection via Videonystagmography

Abstract

We introduce a novel indicator for the automatic detection of Parkinson's disease, a neurodegenerative disorder characterized by motor impairments, including abnormal eye movements. This study, conducted in the context of videonystagmography (VNG), investigates the Pursuit Regularity Index (PRI), which quantifies the spectral sparsity of pupil motion during target tracking. We assess the potential of PRI to improve classification accuracy when combined with other VNG-derived features. Using a proprietary dataset collected at Razi University Hospital in Tunisia, our results demonstrate that integrating PRI into machine learning models significantly enhances the accuracy and reliability of PD detection.