Biomedical Engineering Reference
In-Depth Information
per-channel sampling rate of 1,000 Hz. However, studies of low-amplitude,
high-frequency components within both phase-locked and nonphase-locked
responses would likely benefit from even greater specifications. In the near future,
optimal system configurations for iEEG will likely be capable of 24-bit recordings of
up to 256 or 512 channels at 3,000 Hz per channel. Until such systems are commer-
cially available and considered necessary for clinical purposes, cortical function
mapping with iEEG will likely require some degree of customization.
Another important consideration for cortical function mapping with iEEG is the
need to record markers for events during tasks eliciting functional activation. Analy-
sis of phase-locked, as well as nonphase-locked, iEEG responses requires a temporal
reference point for averaging signal energy in the time and frequency domains. It is
therefore necessary to record event markers with a high degree of temporal preci-
sion. Ideally, these markers should be recorded directly into the iEEG data acquisi-
tion stream, either in the iEEG channels themselves or in digital channels recorded
on the same computer bus. Systems in which task events are recorded in parallel by a
separate computer are fraught with synchronization problems and are inherently
unreliable. Most commercially available video EEG systems do not have
explicit capabilities for recording event markers other than those for seizures, but
this capability can often be added to existing configurations with relatively simple
modifications.
14.3
Localizing Cortical Function
Like noninvasive scalp EEG signals, iEEG signals can be analyzed from a variety of
different perspectives in order to correlate signal changes with functional brain acti-
vation. These analyses can be divided into those that focus on signal changes in indi-
vidual cortical regions and those that focus on interactions between cortical regions
(see Section 14.4). In the former, signal analyses are motivated by a desire to localize
functional activation to a particular cortical region and to measure the strength and
timing of activation in this region with respect to controlled parameters of the sen-
sory, motor, or cognitive task. In the latter, signal analyses are designed to identify
and establish functional relationships between different cortical regions, often in
hopes of inferring a network of cortical regions that are jointly responsible for carry-
ing out functional tasks. In this section we discuss different methods for analyzing
regional brain responses and present examples of the different kinds of responses
obtained with these methods, as well as some considerations regarding their
interpretation and application to cortical function mapping.
14.3.1 Analysis of Phase-Locked iEEG Responses
To localize functional activation at individual recording sites, iEEG signal analyses,
like those of scalp EEG, have largely focused on the measurement of signal compo-
nents that either are or are not phase-locked to a sensory task, motor output, or cog-
nitive operation. This distinction between phase-locked and nonphase-locked
components is based primarily on whether EEG or iEEG signals that are recorded
Search WWH ::
Custom Search