Side Effect Mitigation Methods for Closed-Loop Deep Brain Stimulation

Abstract

Advances in deep brain stimulation, such as the development of closed-loop algorithms that sense and respond to patient symptoms in real time, promise to deliver therapy that is increasingly tailored to individual patients and their day-to-day variations in symptoms. However, while the potential for better side-effect management through closed-loop deep brain stimulation (CLDBS) has been recognized, the majority of CLDBS research to date has focused only on tradeoffs between power savings and symptom reduction with little consideration for side-effects. Given that side-effects, when present, can be inconvenient for some patients and have a severe impact on well-being for others, developing a better understanding of side-effects and creating CLDBS algorithms for side-effect mitigation could greatly improve outcomes for recipients of DBS therapy. This dissertation describes research investigating two novel approaches to side effect mitigation for individuals with neurological movement disorders: first, the identification and use of involuntary neural biomarkers of side-effects as triggers for changes in stimulation; and second, the use of voluntarily-controlled, brain-computer interface (BCI) methods to give patients the ability to consciously adjust stimulation parameters to balance trade-offs (such as those between symptom reduction and side effect mitigation). These approaches were designed and tested with two groups of human subjects: Parkinson’s disease patients at University of California, San Francisco and essential tremor patients at University of Washington. In addition to piloting new approaches to CLDBS therapy, this research demonstrates the importance of such systems as investigative tools to better understand the underlying neuroscience of these disorders. In addition to presenting contributions to side-effect mitigation in CLDBS, this dissertation explores certain neuroethical challenges that arose within the aforementioned neural engineering research. The concept of BCI illiteracy is critiqued as a framework that is commonly used in BCI research that is intellectually and ethically problematic. Lastly, the structure and function of an ongoing ethics collaboration conducted during this research is described and recommendations are compiled for future neural engineering research teams who are considering ways to apply ethics considerations to their research projects. This research in neuroethics provides new insight into existing neural engineering topics such as BCI illiteracy; it also has potential to enable future science and engineering researchers to better approach neuroethics in their own areas of expertise.