The primary study objective is to measure the electrical fields evoked by tDCS using subjects who have implanted intracranial electrodes as part of their evaluation for epilepsy surgery. The measurements obtained in these subjects and their brain MRI will be employed to validate existing mathematical models.

In the future, these refined models can be used to target tDCS to predetermined brain regions in healthy and subjects and patient populations. As described above in the safety section, the intensities of stimulation applied in this project are not expected to produce changes in brain function, are below intensities commonly applied in clinical trials, and fall well below safety limits suggested by animal studies.

Noninvasive brain stimulation (NBS) represents a promising set of tools for neurotherapeutics and rehabilitation. In a literature search, NBS has been tested for over seventy neurologic and psychiatric conditions. NBS may complement existing medical treatments, especially for neurologic indications without suitable pharmacotherapies (e.g. tinnitus, dyskinesias) or for patients with pharmaco-resistant illness (e.g. intractable epilepsy, severe depression).

In particular, transcranial direct current stimulation (tDCS) modulates brain activity by delivering low intensity unidirectional current through the scalp. Rather than induce action potentials, tDCS modulates resting neuronal transmembrane potential to influence brain plasticity. Moreover, from a pragmatic perspective, tDCS' benefits include its low cost, portability, and ease of use. Furthermore, tDCS can easily be combined with other interventions such as mental imagery, computerized cognitive interventions, or robot-assisted motor activity.

Current physiological understanding of how TDCS affects brain plasticity at a synaptic, cellular, and a network level is limited. Experimentally, spontaneous neuronal firing activity under the anode generally increases, while firing activity under the cathode decreases, although the precise effects probably depend on the orientation of the axons to the electric field (Nitsche and Paulus, 2000, Bindman et al., 1964, Creutzfeldt et al., 1962, Purpura and McMurtry, 1965). The neuromodulatory effects of tDCS have also been broadly attributed to LTP- and LTD-like mechanisms of synaptic plasticity, involving modulation of NMDA-receptor activity, and sodium and calcium channel activity (Hattori et al., 1990, Islam et al., 1995, Liebetanz et al., 2002). Furthermore, functional neuroimaging studies have revealed both local and distant network effects induced by tDCS, probably mediated by interneuronal circuits (Lefaucheur, 2008).

Advancing the investigators mechanistic understanding of how tDCS affects cortical excitability on a local and distributed level is necessary to (1) customize stimulation parameters (e.g. electrode size, positioning, current intensity and duration) to precisely target brain regions and maximize therapeutic outcomes, (2) confirm safety outcomes for vulnerable patient populations (e.g. children, patients with skull defects and implanted hardware). Previously, patients with a scalp or skull defect have been excluded from stimulation (Bikson, 2012) protocols because of a theoretical risk of current shunting through highly conductive CSF collections. However patients with penetrating brain injury, stroke, or previous brain surgery are precisely those who may most benefit from these technologies.

Computational models using finite element methods (FEM) aim to determine the pattern and intensity of current flow through the brain by incorporating both (1) stimulation parameters and (2) patient characteristics such as underlying anatomy and tissue properties (e.g. size and position of skull defect relative to electrode configuration) (Bikson 2012). For example, one computational model incorporating electrode configuration and skull defect size and properties (Datta et al., 2010) predicts that the majority of electrode configurations surrounding the skull defect (with the exception of stimulating directly on top of a small skull defect) will not significantly increase the peak cortical electrical field intensity. Rather, current is directed to the edges of the bony defect, which may be counterproductive to therapeutic goals. Another computational case study on a stroke patient demonstrated that a relatively conductive stroke lesion concentrated current in the perilesional areas, and that placement of the reference electrode (e.g. right should, right mastoid, right orbitofrontal, and contralateral hemisphere) significantly altered the path of greatest current flow (Datta et al., 2011).

Yet, these modeling predictions are limited in their clinical application, as experimental validation is necessary. Quantitative determination of the electrical field at the neural tissue level is required to establish efficacy and safety for a given individual (Bikson 2012). To the investigators knowledge, there are no published studies that have empirically confirmed the predicted patterns and current flow intensities predicted by these models. This proposed experimental study represents the first-in-kind to quantify voltage intensities, as measured at the brain surface, in response to various stimulation parameters, and will represent a significant advance in the field of noninvasive neurostimulation

Inclusion Criteria:

Subjects for this study will be recruited from patients undergoing routine intracranial electrode placement as part of the evaluation for surgical treatment of medically refractory epilepsy. Inclusion criteria include:

1. Age ≥ 18 years
2. Placement of intracranial electrode arrays (grid, strip and or depth electrodes) for seizure focus localization and/or mapping of eloquent cortex
3. Able to provide written informed consent
4. English-speaking patients only

Exclusion Criteria:

1. Cognitive impairment (Intelligence Quotient <70)
2. Facial or forehead skin breakdown that would interfere with surface electrode placement
3. Contraindication to MRI
4. Known adhesive allergy
5. Space occupying intracranial pathology including brain tumor, ateriovenous malformation, cavernous malformation, prior surgical resection or significant encephalomalcia that would create unknown tissue inhomogeneity that cannot be accurately modeled.
6. Subjects who have had an electrographic or clinical seizure within one hour prior to the stimulation procedure