Subthalamic Field Potentials in Parkinson’s Disease Encodes Motor Symptoms Severity and Asymmetry

Parkinson’s Disease (PD) is a neurodegenerative progressive disorder characterized by asymmetric lateral motor impairment, usually associated with key clinical motor symptoms, such as tremor, rigidity and bradykinesia. Although it is known that specific local field potentials (LFP) band power recorded from the subthalamic nucleus (STN) correlate with some of these symptoms, there is little evidence about how both STN hemispheres information could be used to predict the patient’s clinical condition, and, possibly, used for optimizing the current deep brain stimulation protocols, since identical bilateral stimulation can lead to worsening symptoms. Bearing these ideas in mind, this work proposes a multivariable linear regression approach for patients’ motor symptoms’ prediction based on classical LFP band power (theta (4–8 Hz), alpha (8–13 Hz), beta 1 (13–22 Hz), beta 2 (22–35 Hz), gamma 1 (35–100 Hz), gamma 2 (100–150 Hz), gamma 3 (150–200 Hz)) extracted from contralateral and ipsilateral STN recordings, which aims to identify possible interhemispheric electrophysiological correlates of clinical symptoms. Besides, another prediction scenario is presented aiming to verify if there are any LFP frequency bands that could explain asymmetric clinical manifestation. As a major result, we observed that alpha and beta 2 relate to rigidity and beta 2 to bradykinesia when using only contralateral data. When combining both hemispheres, it is seen that gamma 2 is associated with tremor and beta 1, beta 2 and gamma 1 with bradykinesia. Also, adding ipsilateral data could lead to worsening prediction performance, except for beta 1 concerning the bradykinesia symptom. At last, it was found that there are frequency bands which could help explaining symptoms’ asymmetry, such as theta for rigidity and alpha for bradykinesia.