Subdural cortical surface electrodes can be placed with low morbidity and can precisely identify seizure foci and eloquent areas of the brain in children. While obvious, placement depends on the areas to be studied, which may not be easily determined prior to placement. As a result, in children, the electrode coverage may require a combination of SSE and SGE over multiple areas with different interventions. If evaluating medial temporal regions or evaluating multiple areas to lateralize, SSE can be inserted through burrholes. We have used this method on multiple occasions to define the side of abnormality prior to going to cortical grids a few weeks later. SGE are recommended primarily for those patients whose baseline studies indicate a lateral cortical, or wide areas of brain, thereby requiring a craniotomy for placement. If cortical grids are inserted, inferior and medial temporal activity can be assessed with simultaneous SSE rather than with DE. Interhemispheric SGE are useful to simultaneously monitor both sides of the spread to determine the origin and the spread of the focus. The use of SGE and SSE has been shown to be effective for mapping of the cortical abnormalities seen frequently in pediatric epilepsy, (e.g.) cortical dysplasia and architectural abnormalities.

Placement of surface electrodes is straightforward and does not require specialized technology, as does DE. For grid insertion, a large field is planned in order to permit adequate exposure of the cortical surface for the electrodes, as well as for the subsequent resection. The authors utilize an osteoplastic craniotomy when grids are being inserted so as to decrease the risk of an infected bone flap and need for subsequent removal of the bone. This is particularly true in children when larger hemicranial flaps often are needed. SGE and SSE are placed to cover the surfaces believed to be involved and/or areas of eloquent cortex that may need to be tested. Once adequate coverage has been obtained, each electrode wire is individually passed percutaneously a distance from the incision to decrease the risk of cerebrospinal fluid leak. Each wire is meticulously secured so as not to become dislodged, particularly in children. The SGE and SSE, in some instances, can be secured directly to the dura to avoid movement during the monitoring period. While contemporary SGE are increasingly thin and pliable, they may still distort and compress the brain, eliciting variable degrees of vascular compression, venous obstruction, subdural hematoma formation, pial inflammation, and resultant mass effect. Strategies to avoid the problems of cerebral compression and intracranial hypertension due to brain distortion in children include: duraplasty with autograft or allograft so as to facilitate the closure over the extensive hardware in place; careful inspection, positioning, and trimming of the array to avoid venous or cerebral compression; loose approximation of the bone flap again to avoid cerebral compression and possible injury; and lastly, perioperative dexamethasone, intraoperative mannitol, and the placement of releasing incisions in the silastic arrays. Although not commonly used, some neurosurgeons monitor the patients with an intracranial pressure monitor when a large subdural array has been placed.

Surface recordings are made from each contact in the operating room to ensure adequate electrode recording before closure. Because of the difficulty in gaining cooperation in children in regard to the care and non-manipulation of the electrodes, a watertight dural closure is attempted around each of the electrode wires, securing each to the skin individually, and applying collodion to the exit sites as a sealant. The authors utilize antibiotics continuously for the duration of the monitoring and discontinue them 48 hours after the electrodes have been removed. SGE removal and definitive resective surgery under a general anesthetic are routinely performed up to 2 weeks following grid placement though the electrodes can be left in longer if necessary to adequately define a focus and its relationship to eloquent cortex. In instances of placement of only SSE through burrholes, the electrodes are removed without resective surgery at the same setting, usually waiting 6 weeks prior to either placement of cortical SGE array or the definitive surgery to decrease the risk of infection.

The primary risk of complication with subdural electrodes is infectious. The rate is approximately 1 to 3% for long-term, indwelling electrodes. While the risk of infection tends to increase with the duration of invasive monitoring and may be reduced by the use of prophylactic antibiotics, studies to date have not been definitive. Complications are more common for SGE than for SSE, and primarily include osteomyelitis of the bone flap and increased intracranial pressure. Compared with the use of DE, however, adults monitored with surface electrodes have a lower risk of long term morbidity due to the decreased incidence of hemorrhage or intraparenchymal hematomas. Studies in children have corroborated these findings.