Analysis and Discrimination of Electroencephalography Features for Parkinson’s Disease During the Induction of Negative Mood

Purpose: Parkinson’s disease (PD) is the second-most-common neurodegenerative disorder. Early intervention/treatment relies on the early diagnosis of PD. It is known that depressive disturbances are common in PD patients and that they influence many other clinical aspects of the disease. This study investigated the electroencephalography characteristics induced by emotional video clips between PD patients and healthy control (HC) subjects, with the aim of distinguishing PD patients from HC subjects according to the electroencephalography (EEG) information using a support vector machine (SVM) classifier. Methods: Nineteen PD patients and 20 HC subjects participated in experiments that involved watching and scoring sad audiovisual clips. Five characteristics of the brain activity were calculated in PD patients and compared the results with those in HC subjects. Results: The rating scores of feelings toward emotional video clips did not differ significantly between the two groups. 228 features calculated from the mean frequency, relative power, interhemisphere power asymmetry, coherence, and bispectrum of EEG signals were found to differ significantly between the two groups. Conclusion: Using this information to distinguish PD patients from HC subjects with the SVM classifier resulted in a mean accuracy of 92.50%. In the future, we plan to rank EEG characteristics with the aim of reducing the number of features, and to examine at different stages of the disease, which may enable the preclinical detection of PD.