Assessing Levodopa Effectiveness in Parkison's Disease During Gait Using Electroencephalogram and Machine Learning

Rafael Gonçalves Pires, Clayton Reginaldo Pereira, Tiago Penedo, João Paulo Papa, Mateus Roder, Fabio Augusto Barbieri

Abstract: This study aims to investigate the temporal effects of levodopa, a fundamental medication for treating Parkinson’s disease, a neurodegenerative condition that causes diverse motor alterations as it progresses. Our goal is to identify, at some specific intervals after the first levodopa intake, whether a machine learning approach can determine the effectiveness of that drug during gait exercises using electroencephalogram signals. The analysis will focus on identifying specific temporal patterns associated with levodopa administration and its impact on brain electrical activities during gait. We observed the proposed approach can accurately identify drug administration 90 minutes after the first levodopa intake, showing machine learning techniques are promising to cope with such a task.