Real-Time Adaptive Deep Brain Stimulation for Tremor Using Multi-Biomarker Feedback and Wearable Technology

Abstract

Tremor remains one of the most disabling symptoms of Parkinson’s disease (PD), often leading to significant motor impairment and decreased quality of life. Deep brain stimulation (DBS) is an effective therapy for managing motor symptoms. However, standard cDBS lacks the ability to adjust based on a patient's current symptoms. Furthermore, constant stimulation from cDBS leads to unnecessary power usage and adverse side-effects. To address these limitations, this thesis presents novel real-time adaptive DBS (aDBS) strategies that couple multi-biomarker feedback with wearable technology in a continuous closed-loop system to augment personalized aDBS treatment.This thesis introduces and evaluates two novel aDBS controllers: (1) a dual-input PID controller that takes local field potential (LFP) beta power and raw accelerometer readings from the wrist as inputs to update DBS stumulation (2) a tremor-adjusted PID controller that modifies the setpoint of the PID control architecure based on a tremor-likelihood score. These controllers were verified and released into a clinical trial on three PD participants with a Medtronic RC+S implant.Clinical results demonstrate that the novel aDBS controllers successfully reduce stimulation energy consumption while maintaining non-inferior symptom control compared to cDBS. Additionally, motion-based feedback was found to detect cases of breakthrough tremor which would have been missed from LFP-controlled aDBS systems. Moreover, the controllers were designed to operate outside of a controlled clinic setting and will eventually be released for further testing outside of the clinic.This study advances the next generation of aDBS by demonstrating the feasibility of multi-input, real-time neuromodulation using commercially available wearable technology. Future work will focus on further evaluating the aDBS controllers designed both in and out of the clinic with participants.