Adaptive Deep Brain Stimulation for Parkinson's Disease Application

Parkinson's Disease (PD) is a neurodegenerative disease that affects between seven and ten million people worldwide. It is the second most age-related disorder after Alzheimer's. Motor symptoms that are most common in PD patients are tremor, rigidity, bradykinesia, and instability. PD patients can take medication to help control these symptoms, but medication often leads to negative long-term, systemic side effects which can be debilitating and outweigh benefits from the medication. Deep Brain Stimulation (DBS) devices were innovated to help PD patients. This implantable device delivers electrical stimulation to targeted areas in the brain which help control motor symptoms such as tremors. Currently, DBS devices work by providing constant electrical stimulation to patients which can lead to side effects due to other areas of the brain getting affected. Adaptive DBS (aDBS) is the next step to helping PD patients by controlling the stimulation settings of each patient. The way aDBS improves stimulation is by adding a feedback loop into the device that will automatically adjust the stimulation parameters such as duration. In this report, we will determine whether acceleration from an intrinsic or extrinsic accelerometer can be practically utilized as a feedback control signal for aDBS. Exploratory data analysis, including spectral analysis, will be performed on human subject data from PD patients with implanted DBS devices. This work will contribute to development of a machine learning algorithm to make the DBS device deliver stimulation in a more effective and reliable manner.