Analysis of Parkinsonian Surface Electomyography Through Advanced Signal Processing and Nonlinear Methods

Parkinson’s disease (PD) is a neurodegenerative disease that affects approx. 4% of people over 80 years of age [4]. The result of depleted dopaminergic neurons in the substantia nigra, PD is characterised with symptoms such as muscle rigidity, bradykinetic gait, and severe tremor. To distinguish Parkinsonian electromyographic (EMG) signals from those of healthy controls, recent studies have employed nonlinear methods which can capture the underlying activity of the neuromuscular system. Recurrence quantification analysis (RQA) has been shown to effectively characterise the degree of repeated synchronous structure in non-linear dynamical systems including parkinsonian EMG, through parameters such as determinism (%DET) and recurrence rate (%REC) [1]. Additional parameters such as intermuscular coherence and kurtosis have also been used to observe changes in EMG signals under various conditions [2,3]. To date, limited research has examined the potential to discern EMG of individuals with PD from healthy controls using RQA and intermuscular coherence. The work presented here aims to examine differences in Parkinsonian EMG from that of healthy controls using these measures.