Neural Closed-loop Deep Brain Stimulation for Tremor Mitigation

Abstract

Parkinson’s disease and essential tremor are two common neurodegenerative disorders affecting millions of individuals. The hallmarks of these diseases are their interference with normal movement, which can greatly affect their sufferers’ quality of life. Deep brain stimulation has emerged as therapy that can have a profound therapeutic effect on the symptoms of Parkinson’s disease and essential tremor. Being a relatively new therapy, there is still much that can be done to optimize deep brain stimulation for clinical effect and device performance. Closed-loop stimulation, wherein stimulation is only given when the patient is experiencing symptoms, is a potential method for improving therapy. This closed-loop stimulation can be achieved using neural signals from the brains of Parkinson’s and essential tremor patients to determine when they are experiencing symptoms. This work describes the development and testing of a closed-loop DBS system for treating the symptoms of neurodegenerative movement disorders. First, the optimal neural signal source for a closed-loop system is investigated. After determining the optimal neural biofeedback signal, a closed-loop system is constructed that uses machine learning to build patient-specific models relating neural activity and tremor-inducing movements. This novel system is tested under various conditions, and is able to reliably deliver stimulation at the correct time and provide a highly therapeutic effect. This work serves as a proof-of-concept for the efficacy of machine-learning based, neural closed-loop deep brain stimulation systems for the treatment of neurodegenerative movement disorders.