치의학석사 학위논문
A Study on the Changes of
Information Processing in
Sevoflurane-Induced Sedation in
Volunteers Using
Electroencephalography
뇌파를 이용한 자원자에서 sevoflurane으로
유도된 진정상태에서 정보자극 처리의 변화에
관한 연구
2018년 2월
서울대학교 치의학대학원
치의학과
김 지 웅
A Study on the Changes of
Information Processing in
Sevoflurane-Induced Sedation in
Volunteers Using
Electroencephalography
지도 교수 신 터 전
이 논문을 치의학석사 학위논문으로 제출함
2018년 2월
서울대학교 치의학대학원
치의학과
김 지 웅
김지웅의 치의학석사 학위논문을 인준함
2018년 2월
위 원 장 현 홍 근 (인)
부위원장 신 터 전 (인)
위 원 김 정 욱 (인)
i
Abstract
1. Purpose
In this study, changes in event-related potential (ERP) in the
auditory passive oddball paradigm were investigated when sedation
was induced by sevoflurane inhalation.
2. Materials and Methods
20 volunteers (11 females, nine males, mean age 26.8 years,
age range 24–32 years) without any systemic diseases or mental
disorders were enrolled in this study. Electroencephalography
(EEG) measurements were obtained for each subject using 32-
channel EEG recording devices. In the arousal state, the subject's
baseline EEG was recorded for 5 minutes. Auditory stimulation
based on the passive oddball paradigm was delivered to the subject
via an earphone. The passive oddball paradigm consisted of a
random sequence of 500 tones; 4/5 were standard stimuli (1000
Hz) and 1/5 were target stimuli (1200 Hz). EEG measurements
were obtained during auditory stimulation. Afterwards, sevoflurane
was administered by inhalation at an initial concentration of 0.8
vol%. Sevoflurane concentration was changed to maintain the
bispectral index (BIS) value around 80. After inducing stable
sedation, EEG was performed while applying auditory stimulation in
the same manner as above. After completion of EEG measurements,
sevoflurane administration was discontinued and 100% oxygen was
administered during the subject’s recovery. After ERP was
extracted from the measured EEG, the topographic distribution of
ERP, the temporal changes of ERP in each channel, and the
statistical difference in ERP was analyzed between the arousal and
sedation states.
ii
3. Results
In the arousal state, P300 was observed at 320–360 ms latency
and was concentrated in the frontal and central areas. P300
amplitude was statistically significantly decreased in the sedation
compared with the arousal state.
4. Conclusion
Sevoflurane-induced sedation caused a decrease in P300
amplitude. This result may reflect the weakening of the cognitive
function regulating the attentional process and stimulus
discrimination during sedation.
Keyword : Event-Related Potential, P300, Oddball paradigm,
Sevoflurane, Sedation, Electroencephalography
Student Number : 2014-23003
iii
Table of Contents
Chapter 1. Introduction........................................................ 1
Chapter 2. Materials and Methods........................................ 32.1. Volunteer recruitment...............................................................3
2.2. Acquisition of EEG.....................................................................3
2.3. Auditory passive oddball paradigm..........................................3
2.4. Sevoflurane administration .......................................................4
2.5. EEG preprocessing and statistical analysis............................4
Chapter 3. Results............................................................... 63.1. Topographic distributions of ERP............................................6
3.2. Temporal changes of ERP in channel views...........................7
Chapter 4. Discussion.......................................................... 9
References ........................................................................13
Abstract in Korean.............................................................16
Figure Index
[Figure 1]........................................................................... 6
[Figure 2]........................................................................... 8
1
Chapter 1. Introduction
Sedation is increasingly being used to relieve anxiety and fear
from dental treatments and facilitate surgical procedures. Sedation
is regarded as a state at the boundary between consciousness and
unconsciousness. Determining the mechanism of sedation could help
in better understanding the process of loss and recovery of
consciousness. However, to date, few studies on the change of
sedative level from the perspective of the whole brain network have
been performed, and most studies have focused on how information
processing in the brain changes in either consciousness or complete
unconsciousness. [1-3]
The response to external stimuli is decreased in the sedated
state, and cognitive function and memory processing are most likely
to deteriorate as consciousness declines. However, little is known
regarding the difference in information processing of external
stimuli under sedation compared with the arousal state.
Oddball paradigm is an experimental method commonly used in
studies utilizing event-related potentials (ERPs). [4] This method
presents the random sequences of two distinct stimuli (either visual
or auditory); one is frequently repeated (standard) and the other is
infrequent and deviant (target). When the cognitive function is
activated to detect the rare target stimulus, a specific type of ERP,
P300, is elicited after a certain time period (approximately 300–400
ms) following the introduction of a stimulus. [5] P300 is activated
when an attention-dependent cognitive task is performed in the
arousal state. [6] Therefore, the ERP should be affected in an
environment in which consciousness is low and attention is prone to
decrease.
Sevoflurane is a widely used inhalant anesthetic agent for
2
general anesthesia. Sevoflurane is preferred for general anesthesia
when various surgical procedures are performed due to its low side
effects and high anesthetic efficacy. [7] Recently, the use of
sevoflurane as a sedative agent is increasing because sedation can
be rapidly induced at low concentrations. [8] However, changes in
cognitive function with decreasing consciousness when sevoflurane
is administered have not yet been studied from the perspective of
changes in ERP.
In the present study, the changes in ERP in an auditory passive
oddball paradigm were investigated when sedation was induced by
sevoflurane inhalation.
3
Chapter 2. Materials and Methods
2.1. Volunteer recruitment
Before the experiments were conducted on volunteers, the
experimental methods were approved by the institutional review
board (IRB No. CME17001). A total of 20 volunteers (11 females,
nine males, mean age 26.8 years, age range 24–32 years) were
enrolled in this study after providing informed consent. The
inclusion criteria were healthy adults without any systemic diseases
or mental disorders.
2.2. Acquisition of EEG
Before electroencephalography (EEG) acquisition, volunteers
were instructed to sit on a chair and listen to auditory tones
delivered through earphones. Volunteers were also instructed to
keep their eyes closed during EEG. Continuous EEG data were
obtained from all patients under comfortable conditions (sampling
rate = 2048 Hz, low passed with 417-Hz cutoff frequency). The
EEG was sampled using custom-made software with 32 electrodes
placed according to the standard 10-20 International placement
(Fp1, AF3, F7, F3, FC1, FC5, T7, C3, CP1, CP5, P7, P3, Pz, PO3,
O1, Oz, O2, PO4, P4, P8, CP6, CP2, C4, T8, FC6, FC2, F4, F8, AF4,
Fp2, Fz, Cz). We also placed one electromyography (EMG) channel
on the right outer canthus to remove muscle artifacts. Data were
saved and analyzed offline. The data were downsampled at 128 Hz
with 60-Hz notch filter. To evaluate sedation depth objectively, a
bispectral index (BIS) sensor was also applied to the forehead and
measured during EEG acquisition.
4
2.3. Auditory passive oddball paradigm
The oddball paradigm used in this study consisted of two
different auditory tone sequences, a 1000 Hz standard tone and a
1200 Hz deviant (target) tone. The ratio of standard tone (p=0.8,
n=400) to deviant tone (p=0.2, n=200) was 4. Each tone was
randomly delivered through earphones. The auditory stimulation
time of each tone was 50 ms, with a rise and fall time of 10 ms. The
time interval between each tone was 800 ms. Each subject was
subjected to an auditory oddball task that consisted of a mixture of
500 tones.
2.4. Sevoflurane administration
After acquisition of EEG simultaneously with auditory stimulus
in passive oddball paradigm through the earphone in the awake state,
sevoflurane was administered to participants. The initial
sevoflurane concentration was set to 0.8 vol%. Sevoflurane
concentration was changed to maintain the BIS value around 80. As
soon as the participants were confirmed to experience strange
feelings (such as numbness in the extremities, drowsiness, or
feeling euphoric), participants received one trial of an auditory
oddball task during sevoflurane administration similar to the awake
state.
2.5. EEG preprocessing and statistical analysis
Each channel’s linear trends in the acquired EEG signals were
removed using the detrend function in Matlab 2017b (MathWorks,
Natick, MA, USA). Detrended signals were filtered using a
bandpass filter between 1 Hz and 30 Hz. The bandpassed EEG
signals were collected with epoch from -100 to 900 ms from sound
onset. Next, the continuous signal with a 1000-ms time-window,
starting 100 ms prior to stimulus onset was epoched. The average
amplitude in each epoch’s 100-ms window prior to stimulus onset
5
was subtracted for baseline correction of each epoch. Each epoched
signal was manually inspected by the researchers to exclude
artifacts such as EMG or electrocardiography (ECG). All epoched
signals were averaged to obtain ERPs for each participant. The
two-dimensional topographic map of a scalp data field was with 40
ms window to average each channel’s voltage in awake and sedation
states and each channel’s ratio which had statistical significance in
all moments of a window using the topoplot function in EEGLAB
(Swartz Center for Computational Neuroscience, UC San Diego, La
Jolla, CA, USA) [9]. To illustrate the signals of each channel, the
signal’s plot with the window that was statistically significant was
plotted with a green window in each channel. The maximum and
minimum EEG voltages between 250 and 500 ms from stimulus
onset were obtained to P300’s peak of each channel. In that peak’s
point, the time between awake and sedation states, the peak voltage
between awake and sedation states, the peak voltage and the
voltage at that time of another condition such as the sedation
voltage at awake peak time were compared using paired t-test. In
all statistical analyses, a P value less than 0.05 was considered
statistically significant.
6
Chapter 3. Results
3.1. Topographic distributions of ERP
Figure 1 shows the average topographic distributions of ERP
amplitude in the 320–360-ms latency range. When the subjects
were in the arousal state, strong positive-going ERPs were
observed in the 320–360-ms latency range in response to both
standard and target stimuli as shown in Figure 1(a). The peak
amplitude of the ERPs was observed in the frontal and central areas.
Conversely, when the subjects were in the sedation state, the ERP
amplitude in the frontal and central areas was markedly reduced as
shown in Figure 1(b). Figure 1(c) shows the p-value distributions
of the statistical differences in ERP amplitude between arousal and
sedation states. These results show ERPs were mainly elicited in
the central area with standard stimuli and in the frontal area with
target stimuli.
Figure 1. The average topographic distributions of ERP amplitude in the
7
320–360 ms latency range. (a) The average distributions of ERP in the
arousal state. The dark blue-colored areas indicate the positive-going
ERPs, which are located in the frontal and central areas. (b) The average
distributions of ERP in the sedation state. The ERP amplitude in the frontal
and central areas was markedly reduced. (c) The p-value distributions of
the statistical differences in ERP amplitude between arousal and sedation
states. ERP was concentrated in the central area with standard stimuli and
in the frontal area with target stimuli.
3.2. Temporal changes of ERP in channel views
Based on the above results, temporal changes in ERP were
examined in the frontal and central electrodes where ERP was
strongly elicited. Figure 2(a) shows the average time-domain
graphs of ERPs measured with the Fz electrode. The characteristic
positive-going ERPs were observed in the latency range of 300–
400 ms, and the ERP amplitude when using target stimuli was
greater than when using standard stimuli. These ERP
characteristics confirmed that P300 was elicited by the target
stimuli in the auditory oddball paradigm. Figure 2(b) shows the
average time-domain graphs of ERPs measured with the Cz
electrode. The P300 detected with the Cz electrode showed the
same tendency as the P300 detected with the Fz electrode; the
P300 amplitude detected with both Fz and Cz electrodes was
statistically significantly decreased by nearly 70% in the sedation
state compared with arousal state (P<0.001).
8
Figure 2. The average time-domain graphs of ERP measured with the Fz
and Cz electrodes. The red solid lines indicate the ERP in the arousal state
and the blue solid lines indicate the ERP in the sedation state. The green
boxes show the statistically significant differences (P<0.05) between
ERPs in the arousal and sedation states. The red-dotted lines highlight the
differences between P300 peak amplitudes in the arousal and sedation
states. (a) The average ERP graphs of the measurements obtained with
the Fz electrode. (b) The average ERP graphs of the measurements
obtained with the Cz electrode. The P300 amplitude detected with both Fz
and Cz electrodes was statistically significantly decreased in the sedation
state compared with the arousal state.
9
Chapter 4. Discussion
In this study, we showed that P300 ERP was induced by
auditory stimulation based on a passive oddball paradigm and its
location was mainly concentrated in frontal and central areas. In
addition, the P300 amplitude decreased statically in the sedation
state compared with the arousal state. To the best of our knowledge,
this is the first study in which the changes of auditory stimulus
processing during the sedation state were investigated.
Reportedly, P300 consists of two subcomponents, P3a and P3b.
[5] Previous studies showed that P3a, which has a shorter latency
(360–380 ms), is different from P3b, which has a longer latency
(380–430 ms). [10] P3a is elicited in the frontal lobe when deviant
novel stimuli are detected which are task-irrelevant. P3b is elicited
in the temporal and parietal lobes when infrequent stimuli are task-
relevant (e.g., pressing a button or counting numbers). [11]
In this study, the experiments were conducted based on an
auditory passive oddball paradigm; the subjects listened to the two
types of auditory stimuli, standard and target, and were instructed
not to respond in a specific manner when target stimuli were
detected. [12] Since P3b is elicited only when the stimulus is task-
relevant, the main subcomponent of the P300 observed in this study,
which did not provide a specific task, was P3a.
Several theories have been introduced for the meaning of P300
in brain information processing. The most widely known theory is
the context-updating theory. [13] According to this theory, the
brain produces a stimulus context which is a hypothesis on
incoming stimuli based on the previously detected stimuli. If a new
stimulus is detected that is different from the previous stimuli, the
brain modifies or updates this stimulus context through internal
10
information processing. The main idea of this theory is that P300 is
elicited when the stimulus context is updated.
Previous studies have explained the meaning of P300 based on
this context-updating theory. According to a model proposed by
Polich, [14] when infrequent target stimuli appear between
standard stimuli, the frontal lobe is first activated, which
concentrates attentional resources to detect and distinguish these
stimuli. In this process, P3a is elicited in the frontal lobe. The
parietal lobe is then activated to access the memory storage and
update the memory related to the stimulus context. In this process,
P3b is elicited in the parietal lobe. [15]
The above model of P300 and the results of this study can be
summarized to explain the effect of sevoflurane-induced sedation
on brain information processing. Sevoflurane-induced sedation
significantly reduced the P300 amplitude. Since this study utilized a
passive oddball paradigm, the main component of P300 observed in
this study was P3a, which was elicited when the frontal lobe
concentrates attentional resources to discriminate infrequent
deviant stimuli. Thus, the reduction in P3a caused by sevoflurane-
induced sedation may reflect the weakening of the cognitive
function responsible for attentional process and stimuli
discrimination.
The relationship between P300 and sedation induced by several
sedative agents has been described in previous studies. Jessop, et
al. [16] observed changes in P300 when subjects were given
nitrous oxide (N2O) gas inhalation. As a result, the amplitude of the
P300 waveform decreased as the N2O concentration was increased
from the sedative level to the anesthetic level. Engelhardt, et al.
[17] observed changes in P300 when midazolam, a benzodiazepine-
based sedative, was administered, resulting in an increase in the
latency of P300 and an increase in the errors of task performance.
Reinsel, et al. [18] observed changes in P300 in sedation induced
11
by propofol administration, resulting in a 70% reduction in P300
amplitude and a 50% increase in reaction time. In contrast to our
study, previous studies did not utilize the oddball paradigm. ERP
measured in previous studies only reflected auditory processing,
not auditory attention, since the same auditory stimuli was applied
to patients. Conversely, we utilized the oddball paradigm to
investigate ERP changes under sedation. ERP measured using the
oddball paradigm has gained attention as a neural correlate of
cognitive processing. Consistent with our findings, a recent study
showed that ERP is altered in patients with cognitive impairment
when applying the auditory oddball paradigm. For example, P300
prolongation and loss of P300 potential was observed in patients
with Parkinson disease with mild cognitive impairment. [19]
Alterations of P300 were also observed in obstructive sleep apnea
associated with cognitive impairment. [20] These results indicate
that P300 may provide a diagnostic marker of cognitive impairment
associated with neurological disorders. In this regard, P300 changes
during N2O administration can be considered associated with N2O-
induced cognitive impairment.
The results from this study also indicate that information
processing may be disturbed during sedation considering that it
requires a cognitive task to discriminate rare auditory stimuli under
the auditory oddball paradigm. In fact, patients who are in a sedation
state often experience a slow response to verbal stimuli compared
with the awake state.
The present study had several limitations. First, due to the use
of a passive oddball paradigm, investigation of P3b subcomponent
could not be conducted. Therefore, in future studies, the effect of
sevoflurane-induced sedation on P3b as well as P3a should be
identified using an active oddball paradigm and independent
component analysis (ICA). [21] This study also lacked control over
the factors which may cause changes in P300. By varying the
factors such as difficulty of tasks [22] and target-to-target
12
interval [23], more accurate tendencies of P300 could be obtained.
In conclusion, the present study results showed that
sevoflurane-induced sedation caused a decrease in P300 amplitude,
which may reflect weakening of the cognitive function governing the
attentional process and stimuli discrimination during sedation.
13
References
[1] Lee H, Mashour GA, Noh GJ, Kim S, Lee U. Reconfiguration of
network hub structure after propofol-induced unconsciousness.
Anesthesiology. 2013;119(6):1347-59.
[2] Shin J, Mashour GA, Ku S, Kim S, Lee U. Subgraph
“backbone” analysis of dynamic brain networks during
consciousness and anesthesia. PLoS One. 2013;8(8):e70899.
[3] Lee U, Ku S, Noh G, Baek S, Choi B, Mashour GA. Disruption of
frontal–parietal communication by ketamine, propofol, and
sevoflurane. Anesthesiology. 2013;118(6):1264-75.
[4] Patel SH, Azzam PN. Characterization of N200 and P300:
selected studies of the Event-Related Potential. Int J Med Sci.
2005;2(4):147-54.
[5] Squires NK, Squires KC, Hillyard SA. Two varieties of long-
latency positive waves evoked by unpredictable auditory stimuli in
man. Electroencephalogr Clin Neurophysiol. 1975;38(4):387-401.
[6] Luck SJ, Woodman GF, Vogel EK. Event-related potential
studies of attention. Trends Cogn Sci. 2000;4(11):432-440.
[7] Sakai EM, Connolly LA, Klauck JA. Inhalation anesthesiology
and volatile liquid anesthetics: focus on isoflurane, desflurane, and
sevoflurane. Pharmacotherapy. 2005;25(12):1773-88.
[8] Mesnil M, Capdevila X, Bringuier S, Trine PO, Falquet Y, Charbit
J, Roustan JP, Chanques G, Jaber S. Long-term sedation in
intensive care unit: a randomized comparison between inhaled
sevoflurane and intravenous propofol or midazolam. Intensive Care
Med. 2011;37(6):933-41.
[9] Delorme A, Makeig S. EEGLAB: an open source toolbox for
analysis of single-trial EEG dynamics including independent
component analysis. J Neurosci Methods. 2004;134(1):9-21.
[10] Courchesne E, Hillyard SA, Galambos R. Stimulus novelty, task
relevance and the visual evoked potential in man.
14
Electroencephalogr Clin Neurophysiol. 1975;39(2):131–43.
[11] Conroy MA, Polich J. Normative variation of P3a and P3b from
a large sample: Gender, topography, and response time. Journal of
Psychophysiology. 2007;21(1):22–32.
[12] Bennington JY, Polich J. Comparison of P300 from passive and
active tasks for auditory and visual stimuli. Int J Psychophysiol.
1999;34(2):171-7.
[13] Donchin E. Presidential Address, 1980. Surprise!...Surprise?.
Psychophysiology. 1981;18(5):493–513.
[14] Polich J. Theoretical overview of P3a and P3b. In: Polich J,
editor. Detection of change: event-related potential and fMRI
findings. Boston, MA: Kluwer; 2003. p. 83-98.
[15] Polich J. Updating P300: an integrative theory of P3a and P3b.
Clin Neurophysiol. 2007;118(10):2128-48.
[16] Jessop J, Griffiths DE, Furness P, Jones JG, Sapsford DJ,
Breckon DA. Changes in amplitude and latency of the P300
component of the auditory evoked potential with sedative and
anaesthetic concentrations of nitrous oxide. Br J Anaesth.
1991;67(5):524-31.
[17] Engelhardt W, Friess K, Hartung E, Sold M, Dierks T. EEG and
auditory evoked potential P300 compared with psychometric tests
in assessing vigilance after benzodiazepine sedation and antagonism.
Br J Anaesth. 1992;69(1):75-80.
[18] Reinsel RA, Veselis RA, Wronski M, Marino P. The P300
event-related potential during propofol sedation: a possible marker
for amnesia?. Br J Anaesth. 1995;74(6):674-80.
[19] Yilmaz FT, Özkaynak SS, Barçin E. Contribution of auditory
P300 test to the diagnosis of mild cognitive impairment in
Parkinson's disease. Neurol Sci. 2017;38(12):2103-2109.
[20] Akcali A, Sahin E, Ergenoglu T, Neyal M. Latency of auditory
P300 response is related with cognitive deficits in Obstructive Sleep
Apnea Syndrome. Sleep and Biological Rhythms. 2015;13(1):49–56.
[21] Debener S, Makeig S, Delorme A, Engel AK. What is novel in
the novelty oddball paradigm? Functional significance of the novelty
P3 event-related potential as revealed by independent component
15
analysis. Brain Res Cogn Brain Res. 2005;22(3):309-21.
[22] Isreal JB, Chesney GL, Wickens CD, Donchin E. P300 and
tracking difficulty: evidence for multiple resources in dual-task
performance. Psychophysiology. 1980;17(3):259–73.
[23] Gonsalvez CL, Polich J. P300 amplitude is determined by
target-to-target interval. Psychophysiology. 2002;39(3):388–96.
16
초 록
1. 연구 목적
본 연구의 목적은 sevoflurane 흡입으로 유도된 진정상태에서
auditory passive oddball paradigm을 사용함으로써 event-related
potential (ERP)의 변화에 대해 조사하는 것이었다.
2. 연구 방법
전신질환이나 정신장애가 없는 20명의 자원자 (여성 11명, 남성 9명,
평균 연령 26.8세, 연령대 24-32세)가 본 연구에 참여하였다. 각
피험자의 뇌파는 32채널로 구성된 Electroencephalography (EEG)
기록 장비로 측정되었다. 각성상태에서 피험자의 평상시의 뇌파가 5분
동안 측정되었다. 이후 passive oddball paradigm에 따른 청각자극이
이어폰을 통해 환자에게 전달되었다. passive oddball paradigm은
500개의 톤으로 구성되었으며, 이 중 4/5는 표준자극 (1000 Hz),
1/5는 대상자극 (1200 Hz)이었다. 청각자극이 진행되는 동안 뇌파가
측정되었다. 이후 0.8 vol%의 초기농도로 sevoflurane이 흡입
투여되었다. Bispectral index (BIS) 수치가 80 근처에서 유지되도록
sevoflurane 농도가 조절되었다. 안정적인 진정상태가 유도된 후, 위와
동일한 방식으로 청각자극이 가해지면서 뇌파가 측정되었다. 뇌파
측정이 완료된 후 sevoflurane 투여는 종료되었고, 피험자가 회복하는
동안 100% 산소가 투여되었다. 측정된 뇌파에서 ERP가 추출된 후
ERP의 지형적 분포, 각 채널에서 ERP의 시간적 변화, 각성상태와
진정상태에서 ERP의 통계적인 차이가 분석되었다.
3. 결과
각성상태일 때, P300이 320-360 ms latency에서 관찰되었으며,
P300은 frontal 및 central area에 집중되었다. 각성상태에 비해
진정상태에서 P300 amplitude가 통계적으로 유의미하게 감소했다.
17
4. 결론
sevoflurane을 이용한 진정법은 P300 amplitude의 감소를 야기하며,
이는 진정상태에서 집중과정과 자극구별을 담당하는 인지기능이
약화됨을 반영한다고 할 수 있다.
주요어 : Event-Related Potential, P300, Oddball paradigm,
Sevoflurane, Sedation, Electroencephalography
학 번 : 2014-23003