The primary objective of this exploratory study is to prospectively evaluate the feasibility of image-guided programming of pallidal deep brain stimulation (DBS) for dystonia. The dystonias are a heterogeneous group of movement disorders that share the core clinical feature of abnormal involuntary muscle contractions in common. Pallidal DBS is an established therapy for severe cases with an average improvement in dystonia severity of 50-60%. However, outcomes are variable and difficult to predict, and clinical trials report up to 25% of Nonresponders. Variability in electrode placement and inappropriate stimulation settings may account for much of this variability in outcome. In addition, improvement in dystonia is delayed, often days to weeks after a change in DBS therapy, complicating programming. Our group recently developed a computer model to predict optimal individualized stimulation settings in patients based on the outcome of a large cohort of of chronically treated patients. In-silico testing showed a 16.3% better mean group improvement with computer-assisted programming compared with physician-assisted programming and a dramatic reduction in non-responders (from 25% to 5%). In this prospective study, the computer model will be compared in a randomized, controlled, and double blinded setting against best clinical DBS programming. The primary outcome will be a responder analysis in which dystonia severity will be compared between conventional clinical and model-based programming will be compared.