ABSTRACT
INTRODUCTIONDeep brain stimulation (DBS) is a surgical treatment for movement and psychiatric disorders that achieves therapeutic benefit through the delivery of high-frequency electrical stimulation (typically 100-130 Hz) to specific brain regions associated with pathological activity [1-3]. Sustaining these therapeutic effects over time while minimizing stimulation-induced adverse effects requires periodic manual adjustment of stimulation parameters [4]. As such, clinical DBS programming is an iterative, time consuming, and expensive process [5, 6]. During a programming session, active electrode contacts, as well as stimulation frequency, amplitude, and pulse duration are empirically adjusted through a process that depends on the subjective experience of the patient, acute clinical observations, and the clinical experience of the programmer. Additionally, only a limited subset of
the stimulation parameter space can be explored in a given programming session. As a consequence, many DBS patients require several months of regular parameter adjustments following implantation of the DBS system before an optimal balance between therapeutic benefit and adverse effects can be achieved [7-12]. These parameter adjustments are further complicated by the dynamic and comorbid nature of most neurologic and psychiatric disorders.The limited stimulation parameter space that can be explored clinically, coupled with the subjective nature of the parameter adjustments underscore the need for better alternatives to clinical DBS programming. One alternative is the use of closed-loop systems that continuously monitor the dynamic environment within the brain and respond by automatically adjusting stimulation parameters to achieve and sustain optimal therapeutic efficacy [13-15].
