Biomedical Engineering Reference
In-Depth Information
and a reference. This type of amplifier detects two signals—the voltage between
plus and the reference, and between minus and the reference—and then subtracts
the second signal from the first.
The contribution of the reference electrode is common to both signals and thus
cancels. Attenuation of common-mode artifact signals will be complete only if each
of the plus and minus electrodes is attached to the skin with identical contact imped-
ance. If the electrodes do not have equal contact impedances, the amplitude of the
common-mode signal will differ between the plus and minus electrodes, and they
will not cancel exactly. Most commonly, the EEG is measured (indirectly) between
two points on the scalp with a reference electrode on the ear or forehead. If the refer-
ence electrode is applied far from the scalp (i.e., on the thorax or leg), there is always
a chance that large common-mode signals such as the ECG will not be ideally can-
celed, leaving some degree of a contaminating artifact.
Some artifacts, like the EMG, characteristically have most of their energy in a
frequency range different from that of the EEG. Hence, the amplifier can bandpass
filter the signal, passing the EEG and attenuating the nonoverlapping EMG. How-
ever, it is not possible to completely eliminate EMG contamination when it is active.
At present, most commercial EEG monitors quantify and report EMG activity on
the screen.
9.3.2 Signal Processing
Signal processing of an EEG is done to enhance and aid the recognition of informa-
tion in the EEG that correlates with the physiology and pharmacology of interest.
Metaphorically, the goal is to separate this “needle” from an electrical haystack.
The problem in EEG-based assessment of the anesthetic state is that the characteris-
tics of this needle are unknown, and since our fundamental knowledge of the central
nervous system (CNS) remains relatively limited, our models of these “needles”
will, for the foreseeable future, be based on empirical observation.
Assuming a useful target is identified in the raw EEG waveform, it must be mea-
sured and reduced to a qEEG parameter. The motivation for quantitation is three-
fold: to reduce the clinician's workload in analyzing intraoperative EEGs, to reduce
the level of specialized training required to take advantage of EEG, and finally to
develop a parameter that might, in the future, be used in an automated closed-loop
titration of anesthetic or sedative drugs. The following section introduces some of
the mechanics and mathematics behind signal processing.
Although it is possible to perform various types of signal enhancement on ana-
log signals, the speed, flexibility, and economy of digital circuits has produced revo-
lutionary changes in the field of signal processing. To use digital circuits, it is,
however, necessary to translate an analog signal into its digital counterpart.
Analog signals are continuous and smooth. They can be measured or displayed
with any degree of precision at any moment in time. The EEG is an analog signal:
The scalp voltage varies smoothly over time. Digital signals are fundamentally dif-
ferent in that they represent discrete points in time and their values are quantified to
a fixed resolution rather than continuous. The binary world of computers and digi-
Search WWH ::
Custom Search