9 PD patients and 9 NTC participants performed motor tasks including finger tapping, hand flipping, arm movement, and foot stomping on both left and right sides while fNIRS and EEG were recorded from the area covering the motor cortex of both hemispheres. Each task was performed for 5 trials, each consists of 10-seconds of activity followed by 10-seconds of rest.

Oxygenated hemoglobin (HbO2) obtained from the fNIRS signal, and power spectral density (PSD) of EEG in the frequency bands of Theta [4-7 Hz], Alpha [8-15 Hz], and Beta [16-30 Hz] were used for further analysis. Support vector machine (SVM) was used to distinguish PD from NTC group. 60% of the data was used for training and the rest for testing the classifier. Three different feature sets have been used: Only fNIRS data, Only EEG data, and Hybrid fNIRS and EEG data.

The results of this pilot study demonstrate the feasibility of objectively distinguishing PD and control based on brain activity, and could be beneficial in the diagnosis of PD and quantification of symptoms. Validation in larger groups is needed.