Professional Documents
Culture Documents
Patrick L. Purdon, Ph.D., Aaron Sampson, B.S., Kara J. Pavone, B.S., Emery N. Brown, M.D., Ph.D.
The Electroencephalogram and Brain patients over time as a three-dimensional plot (power by fre-
Monitoring under General Anesthesia quency vs. time).11 Fleming and Smith12 devised the density-
modulated or density spectral array, the two-dimensional plot
Almost 80 yr ago, Gibbs et al. demonstrated that system-
of the spectrogram, for the same purpose.13 Levy14 later sug-
atic changes occur in the electroencephalogram and patient
gested using multiple electroencephalogram features to track
arousal level with increasing doses of ether or pentobarbital.
anesthetic effects. Despite further documentation of system-
They stated that “a practical application of these observations
atic relations among anesthetic doses, electroencephalogram
might be the use of electroencephalogram as a measure of the patterns, and patient arousal levels,4,15–20 use of the unpro-
depth of anesthesia.”1 Several subsequent studies reported on cessed electroencephalogram and the spectrogram to monitor
the relation between electroencephalogram activity and the the states of the brain under general anesthesia and sedation
behavioral states of general anesthesia.2–6 Faulconer7 showed never became a standard practice in anesthesiology.
in 1949 that a regular progression of the electroencephalo- Instead, since the 1990s, depth of anesthesia has been
gram patterns correlated with the concentration of ether in tracked using indices computed from the electroencephalo-
arterial blood. Bart et al. and Findeiss et al. used the spec- gram and displayed on brain monitoring devices.21–25 The
trum—the decomposition of the electroencephalogram signal indices have been developed by recording simultaneously
into the power in its frequency components—to show that the the electroencephalogram and the behavioral responses to
electroencephalogram was organized into distinct oscillations various anesthetic agents in patient cohorts.26 Some of the
at particular frequencies under general anesthesia.8,9 Bickford indices have been derived by using regression methods to
et al.10 introduced the compressed spectral array or spectrogram relate selected electroencephalogram features to the behav-
to display the electroencephalogram activity of anesthetized ioral responses.26–29 One index has been constructed by
This article is featured in “This Month in Anesthesiology,” page 1A. Drs. Purdon and Brown have summarized some of this work on the
Web site www.anesthesiaEEG.com and have given several seminars on this topic in multiple venues during the last 2 yr.
Submitted for publication April 8, 2013. Accepted for publication May 18, 2015. From the Department of Anesthesia, Critical Care, and
Pain Medicine, Massachusetts General Hospital, Boston, Massachusetts, and Department of Anesthesia, Harvard Medical School, Boston, Mas-
sachusetts (P.L.P.); Department of Anesthesia, Critical Care, and Pain Medicine, Massachusetts General Hospital, Boston, Massachusetts (A.S.,
K.J.P.); and Department of Anesthesia, Critical Care, and Pain Medicine, Massachusetts General Hospital, Boston, Massachusetts; Department
of Anesthesia, Harvard Medical School, Boston, Massachusetts; Institute for Medical Engineering and Science and Harvard-Massachusetts
Institute of Technology, Health Sciences and Technology Program; and Department of Brain and Cognitive Sciences, Massachusetts Institute
of Technology, Cambridge, Massachusetts (E.N.B.).
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. All Rights Reserved. Anesthesiology 2015; 123:937–60
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Electroencephalography for Anesthesiologists
using classifier methods to derive a continuum of arousal lev- signatures observable in the unprocessed electroencephalo-
els ranging from awake to profound unconsciousness from gram and the spectrogram. The new concept of defining the
visually categorized electroencephalogram recordings.30,31 anesthetic state by using drug-specific electroencephalogram
Another index has been constructed by relating the entropy signatures that relate to molecular and neural circuit mecha-
of the electroencephalogram signal—its degree of disorder— nisms of anesthetic action would allow anesthesiologists to
to the behavioral responses of the patients.32,33 The indices make more detailed and accurate assessments than those
are computed from the electroencephalogram in near real based on electroencephalogram-based indices. The potential
time and displayed on the depth-of-anesthesia monitor benefits of using the unprocessed electroencephalogram to
as values scaled from 0 to 100, with low values indicating monitor anesthetic states have been recently stated.48,49 Cur-
greater depth of anesthesia. The algorithms used in many rent brain monitors display the unprocessed electroencepha-
of the current depth-of-anesthesia monitors to compute the logram and the spectrogram.22,31,50
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
EDUCATION
the cortex can be measured at the scalp as the electroenceph- the scalp because the electric field decreases in strength as the
alogram (fig. 1B). square of the distance from its source.56 However, because cor-
The organization of the pyramidal neurons in the cortex tical and subcortical structures are richly interconnected, scalp
favors the production of large local field potentials because the electroencephalogram patterns reflect the states of both cortical
dendrites of the pyramidal neurons run parallel with each other and subcortical structures.57 Thus, the electroencephalogram
and perpendicular to the cortical surface (fig. 1C). This geom- provides a window into the brain’s oscillatory states.
etry creates a biophysical transmitting antenna that generates A growing body of evidence suggests that anesthet-
large extracellular currents whose potentials can be measured ics induce oscillations that alter or disrupt the oscillations
through the skull and scalp as the electroencephalogram.53,55,56 produced by the brain during normal information process-
Subcortical regions, such as the thalamus (fig. 1D), produce ing.19,20,57–63 These anesthesia-induced oscillations are read-
much smaller potentials that are more difficult to detect at ily visible in the electroencephalogram.
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Electroencephalography for Anesthesiologists
Data Analysis spectral analysis by computing the spectrum (fig. 3C) and
To show the unprocessed electroencephalogram recordings the spectrogram (figs. 3D and 3E).64,65,69
and their spectrograms computed with the same methods For a given segment of electroencephalogram data, the
on comparable displays, we have taken examples from cases spectrum (fig. 3C) provides a decomposition of the segment
of patients receiving general anesthesia and sedation at our into its frequency components usually computed by Fourier
institution. These data were recorded following a protocol methods.64,65 The advantage of the spectrum is that it shows
approved by the Massachusetts General Hospital Human the frequency decomposition of the electroencephalogram
Research Committee to construct a database of electroen- segment for all of the frequencies in a given range (fig. 3C)
cephalogram of recordings of patients receiving general anes- by plotting frequency on the x-axis and power on the y-axis.
thesia and sedation. Electroencephalograms were recorded Power is commonly represented in decibels, defined as 10
using the Sedline monitor (Masimo Corporation, USA). times the log base 10 of the squared amplitude of a given
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
EDUCATION
Fig. 2. Unprocessed electroencephalogram signatures of propofol-induced sedation and unconsciousness. (A) Awake eyes
open electroencephalogram pattern. (B) Paradoxical excitation. (C) Alpha and beta oscillations commonly observed during
propofol-induced sedation (fig. 5). (D) Slow-delta and alpha oscillations commonly seen during unconsciousness. (E) Slow os-
cillations commonly observed during unconsciousness at induction with propofol (fig. 6) and sedation with dexmedetomidine
(fig. 11) and with nitrous oxide (fig. 13). (F) Burst suppression, a state of profound anesthetic-induced brain inactivation com-
monly occurring in elderly patients,68 anesthetic-induced coma, and profound hypothermia (fig. 6, B and D). (G) Isoelectric
electroencephalogram pattern commonly observed in anesthetic-induced coma and profound hypothermia. With the exception
of the isoelectric state, the amplitudes of the electroencephalogram signatures of the anesthetized states are larger than the
amplitudes of the electroencephalogram in the awake state by a factor of 5 to 20. All electroencephalogram recordings are from
the same subject. Reproduced, with permission, from Brown et al. Chapter 50 in Miller’s Anesthesia, 8th edition, 2014.
three-dimensional structure (fig. 3D). However, it is plotted in is termed the density spectral array (fig. 3E),12,13 whereas the
two dimensions by placing time on the x-axis, frequency on the three-dimensional plot of the spectrogram is termed the com-
y-axis, and power through color coding on the z-axis (fig. 3E). As pressed spectral array (fig. 3D).10,22 We display the spectrogram
discussed earlier, this two-dimensional plot of the spectrogram as a density spectral array and refer to it as the spectrogram.
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Electroencephalography for Anesthesiologists
Fig. 3. Construction of the spectrogram. (A) A 10-s electroencephalogram (EEG) trace recorded under propofol-induced
unconsciousness. (B) The electroencephalogram trace in A filtered into its two principal oscillations: the blue curve, an alpha
(8 to 12 Hz) oscillation, and the green curve, a slow (0.1 to 1 Hz) oscillation. (C) The spectrum provides a decomposition of
the electroencephalogram in A into power by frequency for all of the frequencies in a specified range. The range here is 0.1 to
30 Hz. Power at a given frequency is defined in decibels as the 10 times the log base 10 of the squared amplitude. The green
horizontal line underscores the slow-delta frequency band and the blue horizontal line underscores the alpha frequency band
used to compute the filtered signals in B. The median frequency, 3.4 Hz (dashed vertical line), is the frequency that divides the
power in the spectrum in half. The spectral edge frequency, 15.9 Hz (solid vertical line), is the frequency such that 95% of the
power in the spectrum lies below this value. (D) The three-dimensional (3D) spectrogram (compressed spectral array) displays
the successive spectra computed on a 32-min electroencephalogram recording from a patient anesthetized with propofol.
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
EDUCATION
Spectral analyses make it easier to visualize frequency con- Neurophysiology and Clinical
tent, especially oscillations, and to detect subtle changes in fre- Electrophysiology of Selected Intravenous
quency structure. Nevertheless, it is important to know both Anesthetics
the time domain and spectral representations of a given behav-
We review the neuropharmacology and clinical electro-
ioral or neurophysiological state induced by an anesthetic. We
physiology of propofol, dexmedetomidine, and ketamine.
present both in our discussions in the following sections for
For each anesthetic, we discuss the putative mechanism
commonly used intravenous and inhaled anesthetics.
through which its actions at specific molecular targets in
specific neural circuits produce the electroencephalogram
Fig. 3. (Continued ). Each spectrum is computed on a 3-s inter- signatures and the behavioral changes associated with its
val and adjacent spectra have 0.5 s of overlap. The black curve anesthetic state.
at minute 24 is the spectrum in C. (E) The spectrogram in D plot-
ted in two dimensions (density spectral array). The black vertical Table 1. Spectral Frequency Bands
curve is the spectrum in D. The lower white curve is the time
course of the median frequency and the upper white curve is the Frequency Range
time course of the spectral edge frequency. A–E were adapted, Name (Hertz, Cycles per Second)
with permission, from Purdon and Brown, Clinical Electroen-
Slow <1
cephalography for the Anesthesiologist (2014), from the Partners Delta 1–4
Healthcare Office of Continuing Professional Development.69 Theta 5–8
Adaptations are themselves works protected by copyright. In or- Alpha 9–12
der to publish this adaptation, authorization has been obtained Beta 13–25
both from the owner of the copyright of the original work and Gamma 26–80
from the owner of copyright of the translation or adaptation.
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Electroencephalography for Anesthesiologists
Molecular and Neural Circuit Mechanisms of Propofol oscillations are larger than those of the gamma oscillations seen
Propofol, the most widely administered anesthetic agent, is in the awake electroencephalogram (fig. 2A). When general
used as an induction agent for sedation and maintenance of anesthesia has been maintained by a propofol infusion, a simi-
general anesthesia. Low-dose propofol is frequently adminis- lar beta oscillation pattern is visible in patients after extuba-
tered either as small boluses or by infusion for surgeries and tion as they lie quietly before transfer to the postanesthesia care
diagnostic procedures that require only sedation. unit. A beta oscillation is also seen during paradoxical excita-
The molecular mechanism of propofol has been well char- tion (fig. 2B), the state of euphoria or dysphoria with move-
acterized. Propofol binds postsynaptically to γ-aminobutyric ments, that can occur when patients are sedated. The state is
acid type A (GABAA) receptors where it induces an inward termed paradoxical because a dose of propofol intended to
chloride current that hyperpolarizes the postsynaptic neurons sedate results in excitation. Two mechanisms have been pro-
thus leading to inhibition.72,73 Because the drug is lipid soluble posed to explain propofol-induced paradoxical excitation. One
and GABAergic inhibitory interneurons are widely distrib- involves GABAA1-mediated inhibition of inhibitory inputs
uted throughout the cortex, thalamus, brainstem, and spinal from the globus pallidus to the thalamus leading to increased
cord, propofol induces changes in arousal through its actions excitatory inputs from the thalamus to the cortex.46 This mech-
at multiple sites (fig. 4). In the cortex, propofol induces inhi- anism is also the one through which the sedative zolpidem is
bition by enhancing GABA-mediated inhibition of pyramidal postulated to induce arousal in minimally conscious patients.76
neurons.73 Propofol decreases excitatory inputs from the thala- The second mechanism, established in simulation studies, pos-
mus to the cortex by enhancing GABAergic inhibition at the tulates that low-dose propofol induces transient blockade of
thalamic reticular nucleus, a network that provides important slow potassium currents in cortical neurons.76
inhibitory control of thalamic output to the cortex. Because
the thalamus and cortex are highly interconnected, the inhibi- The Electroencephalogram Signatures of Propofol Are
tory effects of propofol lead not to inactivation of these circuits Slow-delta Oscillations on Induction
but rather to oscillations in the beta (figs. 2C and 5) and alpha The electroencephalogram patterns observed during propo-
(figs. 2D, 6 and 7) ranges. Propofol also enhances inhibition in fol general anesthesia depend critically on several factors, the
the brainstem at the GABAergic projections from the preoptic most important of which is the rate of drug administration.
area of the hypothalamus to the cholinergic, monoaminergic, When propofol is administered as a bolus for induction of
and orexinergic arousal centers (fig. 4). Decreasing excitatory general anesthesia, the electroencephalogram changes within
inputs from the thalamus and the brainstem to the cortex 10 to 30 s from an awake pattern with high-frequency, low-
enhances hyperpolarization of cortical pyramidal neurons, an amplitude gamma and beta oscillations (fig. 2A) to patterns
effect that favors the appearance of slow and delta oscillations of high-amplitude slow and delta oscillations (fig. 2E). The
on the electroencephalogram (figs. 5–7).20,60,75 slow-delta and delta oscillations appear in the spectrogram
as increased power between 0.1 to 5 Hz (fig. 6, A and B,
The Electroencephalogram Signatures of Propofol between minutes 0 and 5) and in the time domain as high-
Sedation and Paradoxical Excitation Are Beta-gamma amplitude oscillations (fig. 6C, minute 5.5, and fig. 6D,
Oscillations minute 7.1). The electroencephalogram change is dramatic
The electroencephalogram patterns seen during sedation are as the slow-delta oscillation amplitudes can be 5 to 20 times
organized, regular beta-gamma oscillations (figs. 2C and 5) larger than the amplitudes of the gamma and beta oscilla-
and slow-delta oscillations (fig. 5). The amplitudes of these tions seen in the electroencephalogram of an awake patient.43
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
EDUCATION
The appearance of slow and delta oscillations (figs. 2 oscillations on the electroencephalogram with loss of con-
and 6, C and D) within seconds of administering propo- sciousness (LOC).20,60,75 Loss of the oculocephalic reflex is
fol for induction of general anesthesia coincides with loss of consistent with the anesthetic acting at cranial nerve nuclei
responsiveness, loss of the oculocephalic reflex, apnea, and III, IV, and VI in the midbrain and pons.46,74 Apnea is most
atonia.46,74 These electroencephalogram signatures and the likely due to the drug’s inhibition of the ventral and dorsal
clinical signs are consistent with a rapid action of the anes- respiratory centers in the medulla and pons,76 whereas the
thetic in the brainstem. After bolus administration, propofol brainstem component of atonia is most likely due to inhibi-
quickly reaches the GABAergic inhibitory synapses emanat- tion of the pontine and medullary reticular nuclei.46
ing from the preoptic area of the hypothalamus onto the Administration of an additional propofol bolus, either
major arousal centers in the brainstem and hypothalamus before or after intubation, can result in either enhance-
(fig. 4). Action of the anesthetic at these synapses inhibits ment of the slow oscillation or the conversion of the slow
the excitatory arousal inputs from the brainstem, favoring oscillation into burst suppression (fig. 6B, minutes 7 to
hyperpolarization of the cortex, the appearance of slow-delta 14, and fig. 6D, minute 11.5). Burst suppression is a state
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Electroencephalography for Anesthesiologists
of unconsciousness and profound brain inactivation in as vertical lines in the spectrogram (fig. 6B, minutes 7 to
which the electroencephalogram shows periods of electrical 14). When propofol is administered as an induction bolus,
activity alternating with periods of isoelectricity or electri- patients, particularly elderly patients, can enter burst sup-
cal silence. In the spectrogram, burst suppression appears pression within seconds.68
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
EDUCATION
As the effects of the bolus doses of propofol recede, the incoherent slow oscillations represent an impediment
the electroencephalogram evolves into slow-delta oscilla- to normal intracortical communications.20,57,60,80 Together,
tion and alpha oscillation patterns (fig. 6A, minute 7 and these two mechanisms likely contribute to the patient being
fig. 6B, minute 15). The time of transition from the slow unconscious when the slow and alpha oscillations are vis-
oscillations or burst suppression into a combined slow ible in the spectrogram of patients receiving propofol. These
oscillation and alpha oscillation patterns depends on how highly coherent alpha oscillations, incoherent slow oscilla-
profound the effect of the bolus dose was. Even if no addi- tions, and anteriorization most certainly contribute to the
tional propofol is administered after the first bolus, it can loss of frontal-parietal effective connectivity80–82 that is also
take several minutes for the transition to occur. The patient associated with propofol-induced LOC.
shown in figure 6A received a second bolus dose of 50 mg
at minute 5. An infusion of propofol was started at a rate Electroencephalogram Signatures of Emergence from
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Electroencephalography for Anesthesiologists
the electroencephalogram is isoelectric (fig. 2G). Within the measure of the instantaneous probability of the brain being
bursts, the electroencephalogram can, in some cases, main- in a state of suppression that can be reliably computed using
tain the brain dynamics that were present before the start state-space methods and used to track burst suppression
of burst suppression.85,91,92 For example, if the patient is in in real time and to implement control systems for medical
a state of slow and alpha oscillations before the burst sup- coma.99,100 A burst suppression probability of 0.5 means a
pression, then the same pattern is present within the bursts 0.5 probability of being suppressed, whereas a burst suppres-
(figs. 6B and 8A). In addition to deep general anesthesia, sion probability of 0.75 means a 0.75 probability of being
burst suppression is also observed in other conditions of pro- suppressed.
found brain inactivation including coma93 and in children
with significantly compromised brain development.94 The Ketamine
observation that multiple different mechanisms induce burst
suppression are consistent with a recently proposed neuro- Neural Circuit Mechanisms of Ketamine
nal metabolic mechanism of burst suppression.85 Transient Ketamine, an anesthetic adjunct and an analgesic, acts
increases and decreases in extracellular calcium, leading to primarily by binding to N-methyl-d-aspartate (NMDA)
synaptic disfacilitation, could also play a role in determining receptors in the brain and spinal cord.45 Ketamine is a
suppression duration.84 channel blocker so that to be effective, the channels have
The burst suppression ratio or suppression ratio is a time to be open.101 Because in general, the channels on inhibi-
domain measure used to track quantitatively the level of tory interneurons are more active than those on pyramidal
burst suppression. The burst suppression ratio is a number neurons, ketamine at low-to-moderate doses has its primary
between 0 and 1, which measures the fraction of time in effect on inhibitory interneurons (fig. 9A).102,103 By block-
a given time interval that the electroencephalogram is sup- ing inputs to inhibitory interneurons, ketamine allows
pressed.95,96 The burst suppression ratio is displayed on some downstream excitatory neurons to become disinhibited or
brain function monitors and is one of the measures used more active.45,46 This is why cerebral metabolism increases
in electroencephalogram-based indices to assess depth of with low doses of ketamine. Hallucinations, dissociative
anesthesia.97,98 A refinement of the burst suppression ratio, states, euphoria, and dysphoria are common with low-dose
termed the burst suppression probability, has been recently ketamine because brain regions, such as the cortex, hippo-
developed (fig. 8C).100 The burst suppression probability is a campus, and the amygdala, continue to communicate but
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
EDUCATION
with less modulation and control by the inhibitory interneu- in the prefrontal cortex due in part to increased glutamate
rons. Information is processed without proper coordination activity at non-NMDA glutamate receptors.104 Analgesia is
in space and time.45,46 The hallucinatory effects are likely due in part to the action of ketamine on glutamate NMDA
enhanced by disruption of dopaminergic neurotransmission receptors at the dorsal root ganglia, the first synapse of the
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Electroencephalography for Anesthesiologists
pain pathway in the spinal cord where glutamate is the pri- of these inhibitory pathways from the preoptic area to the
mary neurotransmitter (fig. 9A).101 As the dose of ketamine arousal centers. Activation of inhibitory inputs from the pre-
is increased, the NMDA receptors on the excitatory gluta- optic area is postulated to be an essential component of how
matergic neurons are also blocked and consciousness is lost. nonrapid eye movement sleep is initiated.113,114 Sedation
by dexmedetomidine is further enhanced due to blockage
The Electroencephalogram Signatures of Ketamine of presynaptic release of norepinephrine, leading to loss of
Sedation Are Beta and Gamma Oscillations excitatory inputs from the locus ceruleus to the basal fore-
Given the preference of ketamine for NMDA receptors on brain, intralaminar nucleus of the thalamus, and cortex115
inhibitory interneurons, whose inhibition results in increased and decreased thalamocortical connectivity.116
cerebral metabolic rate, cerebral blood flow, and hallucina-
tions,105–107 it is no surprise that ketamine is associated with The Electroencephalogram Signatures of
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
EDUCATION
for the lower dose of dexmedetomine (fig. 11A). The spectro- unresponsive to verbal queries from the anesthesiologist and
gram is that of a 48-yr-old, 65-kg female patient who received but moved in response to changes in the level of nociceptive
dexmedetomidine as a loading infusion of 1 μg/kg over 10 min stimulation from the procedure.
and a maintenance infusion of 0.85 μg kg−1 h−1 for placement Propofol frontal alpha oscillations (fig. 7A), dexmedeto-
of a left forearm arteriovenous fistula. The intense slow-delta midine spindles45 (fig. 11, A and B), and sleep spindles117
oscillation persisted for the duration of the procedure. Dur- are all thought to be generated by thalamocortical loop
ing the surgery, the patient was sedated, meaning that she was mechanisms (figs. 7B and 10B). Although the propofol
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Electroencephalography for Anesthesiologists
frontal alpha oscillations are highly coherent and continuous brief interruptions in neuronal activity.118 Similarly, both
in time,19,20,57 the sleep and dexmedetomidine spindles are slow oscillations and spindles under dexmedetomidine are
brief and episodic.43 Sleep spindles are a marker for nonrapid smaller than their counterparts under propofol, which may
eye movement stage II sleep, which is a state of unconscious- reflect a lower level of disruption in neuronal activity under
ness that is less profound than nonrapid eye movement stage dexmedetomidine compared with propofol.43 Consequently,
III or slow-wave sleep.66 Dexmedetomidine spindles are con- cortical and thalamocortical activities are likely to be more
sistent with a light state of sedation.43 profoundly inhibited under propofol-induced unconscious-
Propofol-induced slow oscillations likely result from ness compared with either slow-wave sleep or dexmedeto-
decreased excitatory inputs to the cortex due to propofol’s midine-induced sedation. This difference in cortical and
GABAA-mediated inhibition arousal centers in the midbrain, thalamocortical activity suggests why patients can be aroused
pons, and hypothalamus (fig. 4). Dexmedetomine-induced from sleep and dexmedetomidine-induced sedation but not
slow oscillations likely result from decreased excitatory from propofol-induced unconsciousness.43,60
inputs to the cortex due to dexmedetomidine’s disinhibi-
tion of the inhibitory circuits emanating from the preoptic Clinical Electrophysiology of the Inhaled
area of the hypothalamus to the arousal centers along with
Anesthetics
decreased adrenergically-mediated excitatory inputs to the
basal forebrain, the intralaminar nucleus of the thalamus, Neurophysiological Mechanisms of Inhaled of Anesthetic
and to the cortex directly (fig. 10A). Propofol-induced slow Action
oscillations gate brief periods of neuronal activity (fig. 7B).60 The principal inhaled anesthetics are the ether derivatives
In contrast, sleep slow-wave oscillations are associated with sevoflurane, isoflurane, and desflurane. The inhaled agents are
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
EDUCATION
total anesthetics in that they can maintain all of the required time, the slow and delta oscillations dissipate. The loss of the
behavioral and physiological characteristics of general anes- alpha and slow-delta oscillation power again appears in the
thesia without adjunct. In practice, the inhaled anesthetics spectrogram as a zipper opening pattern.
are rarely used alone but more commonly in conjunction Isoflurane (fig. 12, E and F) and desflurane (fig. 12, G and
with intravenous drugs, nitrous oxide, and muscle relaxants H) have similar patterns to sevoflurane (fig. 12, C and D). At
to provide balanced general anesthesia, which maximizes the sub-MAC concentrations, they also show strong alpha and
desired effects while minimizing side effects. The inhaled slow-delta oscillations. When the concentration of isoflurane
agents are known to create their behavioral and physiologi- or desflurane is increased to MAC levels and above, a theta
cal effects through binding at multiple targets in the brain oscillation fills in between the delta and alpha bands (fig.
and central nervous system including binding to GABAA 12E, minute 30, and fig. 12G, minute 28). As in the case
receptors and enhancing GABAergic inhibition; blockade of of sevoflurane, on emergence from isoflurane and oxygen or
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Electroencephalography for Anesthesiologists
13A, minute 84). Four minutes after the switch, a profound 5 Hz for approximately 4 min (fig. 13, minutes 86 to 90,
slow-delta oscillation appears in the electroencephalogram blue area in the spectrogram) before they evolve to the beta-
(fig. 13A, minute 86, and fig. 13B, minute 86). These slow- gamma oscillations that are more commonly associated with
delta oscillations dominate with near loss of all power above nitrous oxide (fig. 13A, minute 90, and fig. 13B, minute 90).
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
EDUCATION
The appearance of the beta-gamma oscillations and the loss of median pontine reticular formation to the basal forebrain
the slow-delta oscillations take place from minutes 90 to 94. and to the thalamus.124,125 The beta-gamma oscillations
The patient was extubated at minute 110 (not shown). may have a mechanism similar to that of ketamine (fig. 9B)
The appearance of slow-delta oscillations we commonly whose increased beta-gamma activity depends critically on
observe in the switch from one of the ether anesthetics, halo- NMDA-mediated inactivation of inhibitory interneurons.45
thane or propofol, to a high concentration (> 70%) nitrous
oxide has been previously reported.122–124 The slow-delta Discussion
oscillations tend to be transient,122–124 and the beta-gamma Anesthesiologists administer selected combinations of drugs
oscillations usually reported with nitrous oxide40,41 appear as to create the anesthetic state best suited for the patient and
the slow-delta oscillations dissipate. These slow-delta oscil- the given surgical or diagnostic procedure. A growing body of
lations are noticeably different from the slow waves seen evidence suggests that the behavioral effects of the anesthet-
during propofol induction (fig. 6) and during deep dexme- ics are due to neural oscillations induced by their actions at
detomidine sedation (fig. 11D) in that they are accompanied specific molecular targets in specific neural circuits. Anesthe-
by a substantial reduction in spectral power at all frequencies sia-induced oscillations are 5 to 20 times larger than normal
above 10 Hz (fig. 13A, minutes 86 to 90). Although the brain oscillations, likely disrupt normal brain communication
mechanism of this slow oscillation is unknown, one possible (fig. 1) and are readily visible in the unprocessed electroenceph-
explanation is the blockade of NMDA receptor–mediated alogram (fig. 2) and its spectrogram (fig. 3). Therefore, for three
excitatory inputs from the parabrachial nucleus and the widely used intravenous anesthetics—propofol, ketamine, and
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Electroencephalography for Anesthesiologists
dexmedetomidine—we related the electroencephalogram sig- to reduce or not use certain anesthetics, for example, not to use
natures to the molecular targets and the neural circuits in the ether anesthetics during spinal cord monitoring, to facilitate
brain at which these drugs most likely act (figs. 4–11). For the the detection of clinically significant changes in unprocessed
inhaled ether-derived anesthetics such as sevoflurane, isoflu- electroencephalogram patterns.120, 130 Hence, the neurologist’s
rane, and desflurane, we observed that, with the exception of paradigm does not require understanding the electroencepha-
the theta oscillations that appear around 1 MAC and beyond, logram signatures of the anesthetics.
their electroencephalogram patterns during maintenance and Our paradigm synthesizes research from the last 80 yr.
emergence closely resemble those seen in propofol (fig. 12). Gibbs et al.1 showed that under general anesthesia, electro-
Nitrous oxide is known to be associated with increased beta encephalogram patterns changed with level of unconscious-
and gamma oscillations and likely decreased slow-delta oscilla- ness. Several investigators showed that different anesthetics
tions. However, we demonstrated that nitrous oxide also pro- have different electroencephalogram patterns and that these
duces profound slow-delta oscillations during the transition patterns were visible in the spectra of the electroencepha-
from an inhaled ether anesthetic (fig. 13). logram.10,12–14,22 In 1959, Martin et al.6 wrote, “The big
In contrast to brain state monitoring based on the electro- question, of course, remains unanswered; namely, do the
encephalogram-derived indices, which assumes that the same electroencephalographic patterns defined in the several clas-
index value defines for any anesthetic the same level of uncon- sifications presented constitute a valid measure of depth of
sciousness, the unprocessed electroencephalogram and the anesthesia. We believe that this question unanswerable until
spectrogram define a broader range of brain states. Therefore, a precise definition of ‘depth of anesthesia’ is made.” Relating
under our paradigm, their use should enable a more nuanced the electroencephalogram patterns to the mechanisms was not
characterization of brain states guided by mechanistic insights. possible at the time because lipid solubility was the primary
Our paradigm for brain state monitoring differs also from that theory of anesthetic mechanisms. The concept that anesthet-
used by neurologists and clinical neurophysiologists. These ics bind to specific molecular targets was not promulgated
clinicians commonly use unprocessed electroencephalogram until 1984,131 and the detailed neural circuit descriptions
patterns to identify seizures in the intensive care unit,67,126 to of how anesthetic actions at specific molecular targets could
characterize evoked potentials,127 to detect ischemia during lead to altered states of arousal were not reported until more
neurophysiological monitoring in the operating room,128 and than 25 yr later.45,46 In recent years, more has been learned
to characterize sleep stages.129 Their paradigm relies on identify- about the biophysics of the electroencephalogram53 and about
ing patterns in the unprocessed electroencephalogram to diag- how altered states of arousal induced by anesthetics relate to
nose seizures or ischemia irrespective of the anesthetic. To do electroencephalogram activity.19,20,43,47,57,60 Hence, it is now
so, clinical neurophysiologists frequently ask anesthesiologists possible to link electroencephalogram signatures to anesthetic
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
EDUCATION
state and the actions of the drugs at specific molecular targets 3. Faulconer A, Pender JW, Bickford RG: The influence of par-
tial pressure of nitrous oxide on the depth of anesthesia and
and in specific neural circuits (fig. 14). the electro-encephalogram in man. Anesthesiology 1949;
Today, the unprocessed electroencephalogram and the spec- 10:601–9
trogram are displayed on several brain monitors.31,50,132 Accurate 4. Kiersey DK, Bickford RG, Faulconer A Jr: Electro-
spectral analyses are required to track accurately the anesthetic encephalographic patterns produced by thiopental sodium
during surgical operations; description and classification. Br
effects. For this reason, we computed our spectrograms using J Anaesth 1951; 23:141–52
multitaper spectral methods. For a given data length, multita- 5. Galla SJ, Rocco AG, Vandam LD: Evaluation of the traditional
per methods have been shown to be the optimal nonparametric signs and stages of anesthesia: An electroencephalographic
spectral techniques in the sense of giving the spectral estimates and clinical study. Anesthesiology 1958; 19:328–38
6. Martin JT, Faulconer A Jr, Bickford RG: Electroencephalography
with the highest resolution and the lowest variance.64,65,133 As in anesthesiology. Anesthesiology 1959; 20:359–76
a result, the multitaper methods make it easier to identify the 7. Faulconer A Jr: Correlation of concentrations of ether in arte-
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Electroencephalography for Anesthesiologists
23. Palanca BJ, Mashour GA, Avidan MS: Processed electroen- 42. Foster BL, Liley DT: Effects of nitrous oxide sedation on rest-
cephalogram in depth of anesthesia monitoring. Curr Opin ing electroencephalogram topography. Clin Neurophysiol
Anaesthesiol 2009; 22:553–9 2013; 124:417–23
24. Bol CJJG, Danhof M, Stanski DR, Mandema JW: Pharmacokinetic- 43. Akeju O, Pavone KJ, Westover MB, Vazquez R, Prerau MJ,
pharmacodynamic characterization of the cardiovascular, hyp- Harrell PG, Hartnack KE, Rhee J, Sampson AL, Habeeb K,
notic, EEG and ventilatory responses to dexmedetomidine in Gao L, Lei G, Pierce ET, Walsh JL, Brown EN, Purdon PL:
the rat. J Pharmacol Exp Ther 1997; 283:1051–8 A comparison of propofol- and dexmedetomidine-induced
25. Barnard JP, Bennett C, Voss LJ, Sleigh JW: Can anaesthe- electroencephalogram dynamics using spectral and coher-
tists be taught to interpret the effects of general anaesthesia ence analysis. Anesthesiology 2014; 121:978–89
on the electroencephalogram? Comparison of performance 44. Huupponen E, Maksimow A, Lapinlampi P, Särkelä M,
with the BIS and spectral entropy. Br J Anaesth 2007; Saastamoinen A, Snapir A, Scheinin H, Scheinin M, Meriläinen
99:532–7 P, Himanen SL, Jääskeläinen S: Electroencephalogram spin-
26. Kearse LA Jr, Manberg P, Chamoun N, deBros F, Zaslavsky dle activity during dexmedetomidine sedation and physi-
A: Bispectral analysis of the electroencephalogram cor- ological sleep. Acta Anaesthesiol Scand 2008; 52:289–94
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
EDUCATION
61. Li D, Voss LJ, Sleigh JW, Li X: Effects of volatile anesthetic 80. Lee U, Ku S, Noh G, Baek S, Choi B, Mashour GA: Disruption
agents on cerebral cortical synchronization in sheep. of frontal-parietal communication by ketamine, propofol,
Anesthesiology 2013; 119:81–8 and sevoflurane. Anesthesiology 2013; 118:1264–75
62. Wang K, Steyn-Ross ML, Steyn-Ross DA, Wilson MT, Sleigh 81. Imas OA, Ropella KM, Ward BD, Wood JD, Hudetz AG:
JW: EEG slow-wave coherence changes in propofol-induced Volatile anesthetics disrupt frontal-posterior recurrent infor-
general anesthesia: Experiment and theory. Front Syst mation transfer at gamma frequencies in rat. Neurosci Lett
Neurosci 2014; 8:215 2005; 387:145–50
63. Vizuete JA, Pillay S, Ropella KM, Hudetz AG: Graded defrag- 82. Jordan D, Ilg R, Riedl V, Schorer A, Grimberg S, Neufang S,
mentation of cortical neuronal firing during recovery of con- Omerovic A, Berger S, Untergehrer G, Preibisch C, Schulz E,
sciousness in rats. Neuroscience 2014; 275:340–51 Schuster T, Schröter M, Spoormaker V, Zimmer C, Hemmer
64. Mitra PP, Bokil H: Nonparametric Quadratic Estimates, B, Wohlschläger A, Kochs EF, Schneider G: Simultaneous
Observed Brain Dynamics. Edited by Mirta PP, Bokil H. New electroencephalographic and functional magnetic reso-
York, Oxford University Press, 2008, pp 194–196 nance imaging indicate impaired cortical top-down process-
ing in association with anesthetic-induced unconsciousness.
65. Babadi B, Brown EN: A review of multitaper spectral analy-
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Electroencephalography for Anesthesiologists
100. Shanechi MM, Chemali JJ, Liberman M, Solt K, Brown EN: connectivity is a neural correlate of dexmedetomidine-
A brain-machine interface for control of medically-induced induced unconsciousness. Elife 2014; 4:e04499
coma. PLoS Comput Biol 2013; 9: e1003284 117. Destexhe A, Contreras D, Steriade M: Mechanisms underly-
101. Sinner B, Graf BM: Ketamine. Handb Exp Pharmacol 2008; ing the synchronizing action of corticothalamic feedback
182:313–33 through inhibition of thalamic relay cells. J Neurophysiol
102. Olney JW, Farber NB: Glutamate receptor dysfunction and 1998; 79:999–1016
schizophrenia. Arch Gen Psychiatry 1995; 52:998–1007 118. Nir Y, Staba RJ, Andrillon T, Vyazovskiy VV, Cirelli C, Fried
103. Seamans J: Losing inhibition with ketamine. Nat Chem Biol I, Tononi G: Regional slow waves and spindles in human
2008; 4:91–3 sleep. Neuron 2011; 70:153–69
104. Moghaddam B, Adams B, Verma A, Daly D: Activation of 119. Antognini JF, Schwartz K: Exaggerated anesthetic
glutamatergic neurotransmission by ketamine: A novel step requirements in the preferentially anesthetized brain.
in the pathway from NMDA receptor blockade to dopami- Anesthesiology 1993; 79:1244–9
nergic and cognitive disruptions associated with the pre- 120. Lyon R, Feiner J, Lieberman JA: Progressive suppression
frontal cortex. J Neurosci 1997; 17:2921–7 of motor evoked potentials during general anesthesia: The
Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.