Biomedical Engineering Reference
In-Depth Information
different electrophysiological indices of cortical activation and cortical network
dynamics.
iEEG recordings not only have excellent temporal resolution, like all
electrophysiological techniques, but they also can be considered to have a
“mesoscopic” spatial resolution that is intermediate between the microscopic scale
of single and multiunit recordings of neuronal activity, which are almost exclusively
the purvey of animal researchers, and the macroscopic scale of EEG/MEG record-
ings. This unique ability of iEEG to see “both the forest and the trees” has made it
particularly useful for studying how populations of cortical neurons organize their
activity during perceptual and cognitive tasks.
14.2
Intracranial EEG Recording Methods
iEEG recordings can be made with a variety of electrode sizes and configurations.
These include multicontact depth electrodes that may be implanted stereotactically
into deep structures such as amygdala, hippocampus, orbital-frontal cortex, or cor-
tex in the depths of the interhemispheric fissure. Information from CT, MRI, or
cerebral angiograms may be used to avoid hemorrhage from blood vessels along the
implant trajectory. An important advantage of this approach is that the holes that
must be drilled in the skull for each electrode penetration are relatively small, reduc-
ing the likelihood of infection. Furthermore, a variety of contact sizes may be used,
including microwires from which micro-EEG or local field potentials (LFPs) can be
recorded. Many epilepsy surgery centers prefer this approach, even for studying cor-
tical tissue on the lateral surface of the brain. However, comprehensive coverage of
broad cortical regions requires an increasing number of penetrations, and when a
large cortical region must be sampled, to localize a patient's seizure focus and map
eloquent cortex in and around the potential zone of resection, for example, it may be
more practical to use subdural ECoG.
Subdural ECoG (also called
epipial recording
) is performed with an array of
electrodes that usually have a larger surface area (for example, 2.3-mm-diameter
exposed surface) than depth electrodes have and are embedded in a soft silastic sheet
in a variety of configurations. These may include strips that contain a single row of
electrodes, usually up to 8 cm long with 1-cm center-to-center spacing (eight elec-
trodes). Multiple electrode strips may be implanted through a single burr hole in
order to sparsely sample a large cortical region. More comprehensive coverage of a
particular cortical region usually requires two-dimensional arrays of electrodes, or
grids. These grids vary in dimensions from 2
8cm
2
, but they can be cus-
tomized to cover the cortical territory of interest. Because they cannot fit through
burr holes, implantation of these grids requires a craniotomy (Figure 14.1), which is
a more invasive procedure with a greater likelihood of complications.
Electrode placement is dictated solely by clinical concerns and is usually based
on scalp EEG recordings of the patient's seizures and interictal epileptiform dis-
charges, as well as the proximity of the estimated seizure focus to eloquent cortical
areas, that is, motor, sensory, or language cortices. However, because the exact
locations of the seizure focus and eloquent cortex are not known a priori, electrode
coverage is not necessarily limited to cortical regions with abnormal structure or
×
4to8
×
Search WWH ::
Custom Search