Intracranial Monitoring with Subdural Grids and Strips

Limitations of Scalp Electroencephalography

There are a number of circumstances that complicate the interpretation of scalp EEG recordings. The scalp electrodes record primarily the positive or the negative pole of a radially oriented dipole. Scalp EEG waveforms are mostly influenced by dipoles at the apex of a gyrus, because the generators lie perpendicular to the scalp and create vertical dipoles visible to the scalp electrodes. Sulcal dipoles can also contribute to the scalp EEG but to a lesser extent than vertically oriented dipoles. Only sources that extend over multiple cortical layers comprising several square centimeters will have sufficient energy to generate detectable discharges from scalp electrodes. ^, Approximately 70% of the cortical surface lies within the depth of sulci, and also many brain areas, including medial surfaces and other deep structures such as the insula, are not optimally recorded by scalp EEG. Hence a focal epileptogenic discharge arising from a deep source or within a sulcus may not produce a sufficient field to be detected by the scalp electrodes, despite the synchronous activation of a large population of neurons. Epileptiform activity from these deeper sources may appear so small in amplitude at the scalp electrode that it may be obscured by background activity. In addition, conduction of the brain electrical activity to the scalp is impeded by the poor conductivity of the skull. Passage through the multiple layers (cerebrospinal fluid [CSF], dura, skull, and scalp), causes a smearing effect on the surface potential distribution, causing the potential from localized foci to appear in a much wider scalp area. As a result of many of these limitations of the scalp EEG, iEEG recording is considered in selected patients so as to more accurately locate the source of epileptiform activity prior to recommending a cortical resection.

There are several noninvasive tools other than scalp EEG to evaluate the epileptic focus. These include magnetoencephalography, brain magnetic resonance imaging (MRI), voxel-based morphometry, ¹⁸ F-fluorodeoxyglucose–positron emission tomography, single-photon emission computed tomography (SPECT), or subtraction ictal SPECT coregistered to MRI (SISCOM). Each has advantages and limitations, but discussion of these modalities is beyond the scope of this chapter.

Advantages of Subdural Electrodes

Compared with the scalp EEG, subdural electrodes are much closer to the source of cerebral activity. They are separated from the cortex by relatively high-conductivity media (CSF and brain parenchyma), thus the amplitudes of signals are considerably higher. Most of the artifacts seen in a scalp EEG are not present in subdural recordings or are present at a much lower relative amplitude (e.g., alternating current interference), yielding a much better signal-to-noise ratio. Subdural electrodes rapidly achieve a low impedance, which usually remains stable for the duration of the chronic recording.

Subdural recording is also less sensitive to the orientation of the dipole generators than is scalp EEG. For instance, a sheet of dipoles folded in on each other results in a net cancellation of electrical signals and thereby produce a “closed” field such as that often seen in the hippocampus. As a result, activity from these closed fields is not visible at the level of the scalp EEG. Spikes or sharp waves are recorded by scalp EEG only when a relatively large region of cortex has been involved in the epileptic activity. It had been initially reported that a minimum of 6 cm ² of cortex discharging simultaneously is needed to reflect a visible potential on the scalp surface. Experience in recordings with subdural and other invasive electrodes has demonstrated that much of the epileptiform activity cannot be visualized with scalp, nasopharyngeal, or sphenoidal electrodes. In studies comparing simultaneous scalp EEGs and iEEGs in patients with temporal lobe epilepsy, 90% of interictal spikes with a source area of >10 cm ² had associated scalp-recorded spikes, whereas only 10% of cortical spikes that covered an estimated source area of <10 cm ² had an accompanying scalp-recorded spike. The difficulty in recording these lower amplitude spikes may involve the much higher electrical impedance of the skull as compared to the brain and CSF, but also results from the low-pass properties of these layers, which attenuate higher frequencies even further. ^, Additionally, high-frequency responses likely become desynchronized over the cortex, resulting in the attenuation of the response seen at the scalp electrode. Therefore use of subdural electrode recordings provides the opportunity to detect focal EEG seizure onsets well before any changes are detected at the scalp electrode by decreasing the distance between the recording electrode and the cortical surface. As a result, activity that would be visualized on the scalp only as a diffuse or nonlocalizable ictal onset may be seen as a more focal ictal onset using subdural recordings.

Subdural electrodes can provide wide coverage over various surfaces of the cortex with contiguously spaced electrodes. Depth electrodes can allow for a better view of dipoles that are deep-seated or whose orientation is tangential to the plane of the subdural electrodes, but lack the contiguous coverage seen with subdural electrodes. Deep-seated sources are more easily localized using depth electrodes. These electrodes, combined with the anatomo-electro-clinical methodology of SEEG, are particularly suited for epileptogenic zone hypothesis–driven presurgical evaluation of the neural network in nonlesional or neocortical epilepsy. The details of differences between subdural electrodes and depth electrodes, including the technique of SEEG, are described under “Comparison With Depth Electrodes” later in the chapter. Advantages of iEEG recordings using subdural electrodes include better functional mapping. However, there may be a greater risk for symptomatic hemorrhage, infections, and permanent neurological sequelae when using subdural electrodes as compared to SEEG.

Types of Subdural Electrodes

There are generally two types of subdural electrodes arrays: grids and strips. Linear subdural strip electrodes and rectangular subdural grid electrodes can be used for intracranial monitoring depending on the location that needs to be sampled. All invasive electrodes, including depth electrodes, have the advantage of recording the signals from smaller populations of neurons that are not detectable by scalp EEG, as described previously. ^,

Subdural electrodes consist of several conductive disks held together by a flexible matter. A variety of materials have been used to make the subdural electrode contacts, including platinum, silver, and stainless steel. These small disk electrodes are embedded in a thin sheet of Silastic (Dow Corning Corporation), with a variable amount of their surface area exposed to the cortex. The composition of subdural recording grids and the techniques for their manufacture have undergone significant development in the past four decades in order to address several key issues: (1) biocompatibility and biostability (i.e., chemical inertness), (2) sufficient flexibility, (3) insolubility in body fluids, (4) impermeability to ionic species, (5) relatively low dielectric constant to reduce interelectrode crosstalk and capacitive leakage, and (6) freedom from pinholes and cracks. These grids were made to order at some epilepsy centers in the past but are now available commercially from several companies that produce high-quality subdural electrode configurations.

The standard dimensions of subdural electrodes previously used in our institution are a 0.127-mm-thick platinum electrode with a diameter measuring 3.97 mm and center-to-center interelectrode distance of 1 cm. The electrodes are embedded in a mold of flexible Silastic ( Fig. 89.1 ). The shape and size of the subdural electrodes can vary from a single row (i.e., a strip electrode) to various grid arrays comprising from 12 to 120 (e.g., rectangular array of 12 × 10 contacts with 7-mm center-to-center spacing) electrodes. The flexible nature of the electrode array allows it to be molded over various cortical regions. For example, the EEG activity from the frontoparietal lateral convexity can be recorded using a larger 8 × 8 electrode array, whereas the evaluation of EEG activity over mesial and basal surfaces of the cortex usually consists of smaller arrays or strip electrodes. Interhemispheric arrays may be curved to follow the convexity and can have contacts on both cerebral hemispheres.

Surgical Placement

Subdural grid or strip electrodes are implanted in the operating room under general anesthesia. The insertion of subdural electrodes can be accomplished using burr holes for subdural strip electrodes or via craniotomy for larger electrode arrays, which are usually placed on the cortical areas likely to undergo resection surgery ( Fig. 89.2B ). The incision and craniotomy should be planned to allow access to a large cortical area, to enable proper tunneling of the electrode wires to allow a longer tunnel to reduce infection, and to allow for a resection at the time of grid removal.

Most subdural grids are placed under direct visualization to avoid bleeding from bridging veins or bony malformations. Smaller grids or strips are often slid over or under the brain without direct visualization, especially at orbitofrontal, inferior-temporal, and interhemispheric locations. The grid cables are kept in a bundle and sutured to the dural edge to help prevent movement after closure. The dura is closed with interrupted sutures and the bone flap is loosely approximated to accommodate increased mass effect postoperatively. The cables are carefully tunneled under the skin to exit from a separate stab incision a few centimeters away from the original incision margin so as to minimize the chance for infection. A separate purse-string suture is usually placed around the cables as they exit to anchor them and to provide strain relief. Subdural electrodes are typically left implanted for a period of 5–10 days. The total duration of monitoring depends on how quickly seizures are recorded and in some cases the institution’s preferences. In rare instances, subdural electrodes have been kept in place longer in situations in which seizures had not been recorded.