| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
 
|
|
|
||
|
|
Anesth Analg 2004;98:1336-1340 © 2004 International Anesthesia Research Society doi: 10.1213/01.ANE.0000105867.17108.B6
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Bispectral index (BIS) integrates various electroencephalographic (EEG) parameters into a single variable. However, the exact algorithm used to synthesize the parameters to BIS values is not known. The relationship between BIS and EEG parameters was evaluated during nitrous oxide/isoflurane anesthesia. Twenty patients scheduled for elective ophthalmic surgery were enrolled in the study. After EEG recording with a BIS monitor (A-1050) was begun, general anesthesia was induced and maintained with 0.5%–2% isoflurane and 66% nitrous oxide. Using software we developed, we continuously recorded BIS, spectral edge frequency 95% (SEF95), and EEG parameters such as relative beta ratio (BetaRatio), relative synchrony of fast and slow wave (SynchFastSlow), and burst suppression ratio. BetaRatio was linearly correlated with BIS (r = 0.90; P < 0.01; n = 253) at BIS more than 60. At a BIS range of 30 to 80, SynchFastSlow (r = 0.60; P < 0.01; n = 3314) and SEF95 (r = 0.75; P < 0.01; n = 3339) were linearly correlated with BIS. The correlation between BIS and SEF95 was significantly better than the correlation between BIS and SynchFastSlow (P < 0.01). At BIS less than 30, the burst suppression ratio was inversely linearly correlated with BIS (r = 0.76; P < 0.01; n = 65). At BIS less than 80, burst-compensated SEF95 was linearly correlated with BIS (r = 0.78; P < 0.01; n = 3404). In the range of BIS from 60 to 100, BIS can be calculated from BetaRatio. At surgical levels of anesthesia, BIS and SynchFastSlow (a parameter derived from bispectral analysis) or burst-compensated SEF95 (derived from power spectral analysis) are well correlated. However, our results show that SynchFastSlow has no advantage over SEF95 in calculation of BIS.
IMPLICATIONS: The relationship between bispectral index (BIS) and electroencephalographic parameters was evaluated during nitrous oxide/isoflurane anesthesia. At surgical levels of anesthesia, BIS and the relative synchrony of fast and slow wave (a parameter derived from bispectral analysis) or burst-compensated spectral edge frequency 95% (a parameter derived from power spectral analysis) are well correlated.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The electroencephalographic (EEG)-derived bispectral index (BIS) is a sensitive index that reflects the hypnotic component of anesthesia. BIS is a dimensionless number scaled from 100 to 0, with 100 representing an awake EEG and 0 representing complete electrical silence (1). BIS values of 40–65 have been recommended for general anesthesia (1).
BIS integrates EEG parameters into a single variable (2). Briefly, the EEG is digitized and processed to detect and remove artifacts. The signal is then analyzed for suppression detection and also fast Fourier-transformed. The suppression is used to compute the burst suppression ratio (BSR). The fast Fourier transform is used to compute a relative beta ratio (BetaRatio) and also to compute the bispectrum, from which the relative synchrony of fast and slow wave (SynchFastSlow) is derived. All of these components are combined by using multipliers derived from discriminant analysis, with the result scaled from 0 to 100 (3). However, the BIS machine displays only BSR and spectral edge frequency 95% (SEF95), which is not included in the published calculation process of BIS. There is no way to know how BetaRatio and SynchFastSlow change during anesthesia. The exact algorithm used to synthesize the parameters to BIS values is still not known.
We have developed software that can perform power spectral analysis and bispectral analysis of EEG (4). It calculates the EEG parameters such as BetaRatio and SynchFastSlow. Investigating the relationship between BIS and these parameters might bring new insight into the BIS-calculation process. We evaluated the relationship between BIS and these parameters during nitrous oxide/isoflurane anesthesia.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Twenty patients scheduled for elective ophthalmic surgery were enrolled in the study. The study protocol was approved by the IRB, and informed consent was obtained from each patient.
The patients were premedicated with 0.5 mg of atropine and 2 mg of midazolam administered IM approximately 30 min before the start of anesthesia. Before the induction of anesthesia, the electrodes for BIS (BIS Sensor; Aspect Medical Systems, Natick, MA) were applied to the forehead regions. The EEG was monitored continuously by using an Aspect A-1050 monitor (BIS Version 3.4; Aspect Medical Systems). The EEG signal from the raw EEG port of the A-1050 monitor was introduced into a personal computer with Microsoft Windows ME. Then the EEG-analysis software Bispectral Analyzer (4) calculated BetaRatio, SynchFastSlow, and SEF95. BetaRatio was obtained by using the published formula described by Rampil (2). BetaRatio is the logarithm of the ratio of the EEG spectral power in the 30- to 47-Hz band to the EEG spectral power in the 11- to 20-Hz band:
|
|
SynchFastSlow is the logarithm of the ratio of the bispectral power in the waveband 40–47 Hz to that in the band 0.5–47 Hz (5):
|
|
BetaRatio, SynchFastSlow, and SEF95 were calculated from a recent 1-min EEG. BIS, BSR from the A-1050, and these parameters were recorded every minute on the computer from the start of anesthesia until the patient’s recovery of consciousness. If the signal quality index, which is an indicator of EEG signal quality by the A-1050, was less than 80, the pair of parameters was excluded. Burst-compensated SEF95 (BcSEF) was calculated off-line as a proportional reduction of the SEF95 in the presence of burst suppression (2):
|
|
General anesthesia was induced with thiopental 5 mg/kg and vecuronium 1.5 mg/kg and was maintained with 0.5%–2% isoflurane. The trachea was intubated, and the lungs were ventilated with 66% nitrous oxide in oxygen with a fresh gas flow of 6 L/min. Mechanical ventilation was adjusted to maintain an end-tidal carbon dioxide partial pressure of 35–40 mm Hg. Noninvasive mean arterial blood pressure, heart rate, and pulse oximetry were monitored continuously and maintained within a normal range (mean arterial blood pressure >60 mm Hg, heart rate between 50 and 100 bpm, and pulse oximetry >95%). The bladder temperature was monitored continuously and was maintained at normothermia (36.0°C–36.5°C) by using a water blanket (Medi-Therm II; Gaymer, NY).
At least 15 min after the start of surgery, the relationship between isoflurane concentration and EEG parameters was evaluated if the BIS value was less than 60 at an end-tidal isoflurane concentration of 0.4%. The end-tidal isoflurane concentration was increased stepwise by 0.4% up to 1.6%. Each concentration was maintained for 15 min, and EEG parameters were recorded.
After visual confirmation of a linear relationship between the BIS and EEG parameters, linear regression with subsequent calculation of the Pearson correlation coefficient was used to quantify the statistical significance. One-way analysis of variance and Scheffé’s multiple comparison test were used to evaluate the relationship between isoflurane concentration and EEG parameters. P < 0.05 was considered significant. Data are expressed as mean ± SD.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Demographic data are shown in Table 1. In 20 patients, 3509 pairs of EEG parameters were recorded and analyzed.
|
Figure 1A shows the relationship between BIS and BetaRatio. At BIS more than 60, BetaRatio was linearly correlated with BIS (BIS = 20 x BetaRatio + 95; r = 0.90; P < 0.01; n = 253).
|
Figure 1B shows the relationship between BIS and SynchFastSlow. At a BIS range of 30 to 80, SynchFastSlow was linearly correlated with BIS (r = 0.60; P < 0.01; n = 3314).
Figure 1C shows the relationship between BIS and BSR. At BIS less than 30, the BSR was inversely linearly correlated with the BIS (r = 0.76; P < 0.01; n = 65).
Figure 2A shows the relationship between BIS and SEF95. At a BIS range of 30 to 80, SEF95 was linearly correlated with BIS (r = 0.75; P < 0.01; n = 3339). The correlation between BIS and SEF95 was significantly better than the correlation between BIS and SynchFastSlow (P < 0.01) (6).
|
Figure 2B shows the relationship between BIS and BcSEF. At a BIS less than 80, BcSEF was linearly correlated with BIS (BIS = 2 x BcSEF + 16; r = 0.78; P < 0.01; n = 3404).
The relationship between isoflurane concentration and EEG parameters was obtained in seven cases (Fig. 3). BIS decreased as isoflurane concentration increased up to 1.2%. However, BIS did not change between 1.2% and 1.6% isoflurane. BcSEF decreased as isoflurane concentration increased up to 1.6% (Fig. 3A). SynchFastSlow decreased as isoflurane concentration increased to 0.8%. However, SynchFastSlow did not change between 0.8% and 1.6% isoflurane. BetaRatio did not change significantly in the study (Fig. 3B). BSR increased only when the isoflurane concentration was at 1.6% (Fig. 3C).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
BIS is a sensitive index of the hypnotic component of anesthesia. It is a dimensionless number scaled from 100 to 0, with 100 representing an awake EEG and 0 representing complete electrical silence. The BIS integrates EEG parameters such as BetaRatio, SynchFastSlow, and BSR into a single variable. However, the exact algorithm used to synthesize the parameters to a BIS value is still not known.
In this study, we attempted to show how BIS is calculated by EEG analysis, because little is known about it. This information may prove useful for introducing a new instrument into clinical care.
BetaRatio has been reported to be a good indicator for tracking the patient’s level of consciousness during the induction of anesthesia. One study has highlighted the importance of the so-called 
-band EEG oscillations (40–60 Hz) as a marker of the conscious state. BetaRatio can track the state of consciousness by using the higher EEG frequencies (up to 47 Hz) (5). The ED50 BIS value for unconsciousness in volunteers anesthetized with thiopental or propofol was 67 (7). It is reasonable that BIS and BetaRatio correlated well in the range of BIS from 60 to 100. In this range, BIS can be calculated from BetaRatio.
At surgical levels of anesthesia, SynchFastSlow may be an important component of BIS (2). The core technology for the calculation of BIS is bispectral analysis (2,8). Bispectral analysis is an advanced signal-processing technique that quantifies phase coupling among the components of a signal. However, in this study, both SEF95 (power spectral analysis derived) and SynchFastSlow (bispectral analysis derived) had a better correlation with BIS. The results indicate that SynchFastSlow may reflect the frequency changes of EEG as SEF rather than the degree of phase coupling.
SynchFastSlow is a bispectrum-derived variable that is the logarithm of a ratio of two sums of bispectral activities in different regions (2). Bispectrum values are influenced by the amplitude of signals and the degree of phase coupling, whereas bicoherence, the normalized parameter of bispectrum, directly indicates the degree of phase coupling (4). This might be why SynchFastSlow shows no advantage in calculation of BIS rather than SEF95.
BSR is a time-domain EEG parameter that was developed to quantitate burst suppression. To calculate BSR, suppression is considered as those periods longer than 0.50 s during which the EEG voltage does not exceed approximately ±5.0 µV (2). The percentage of the suppression period against the time of analysis is defined as the BSR. Similar to our results, Bruhn et al. (9) reported that BSR >40% was linearly correlated with BIS values in the range of 30 to 0. Some points have a BSR of 0 and a low BIS value in Figure 1C. However, BcSEF smoothly tracks the low BIS range in Figure 2B. This result indicates that not only BSR >40%, but also low SEF, relates to BIS less than 30. The combination of SEF with the BSR to form the BcSEF creates a parameter that appears to smoothly track the changes in EEG (2). In this study, BIS and BcSEF were linearly correlated with BIS 80, suggesting that changes of BIS were explained in greater part by the spectral change of the EEG.
Over the years, many methods have been proposed to quantify the depth of anesthesia from an EEG signal. Most of these methods quantify, by power spectral analysis, the increase in the low-frequency components of the EEG signal that occurs as the anesthetic concentration increases. Calculation of SEF95, the frequency below which 95% of the power in the spectrum resides, is one of these methods. It has been shown that SEF95 may predict the depth of anesthesia fairly well. However, the SEF response to changing depth of anesthesia is biphasic. An initial increase of SEF from the awake state was followed by a decrease in SEF (10). BIS compensates for this phenomenon by using BetaRatio during awake to light sedation.
In addition to the BSR, BIS has another burst suppression detection parameter, QUAZI, for which the exact algorithm for calculation is not in the public domain. Some mismatches between BIS and BcSEF seen at a BIS range of 30 to 50, as shown in Figure 2, might be the effect of QUAZI.
BcSEF decreased as the isoflurane concentration increased up to 1.6%. BIS did not change significantly between 0.8% and 1.6% isoflurane. In two cases, BIS decreased to less than 30 when the isoflurane concentration was increased from 1.2% to 1.6%. In other cases, BIS did not change or, rather, increased. The same paradoxical increase in BIS was reported at preburst EEG patterns at the same end-tidal isoflurane concentration (11). A change of SynchFastSlow was not obvious at this isoflurane concentration. Although it is not possible to know how QUAZI affects the BIS calculation, BIS might be a problem to evaluate at this isoflurane concentration. BcSEF was superior at tracking the changes of isoflurane concentration during nitrous oxide/isoflurane anesthesia.
Two BIS-calculation formulas were obtained in this study. We defined BIS calculated from BetaRatio as BIS-BetaRatio (BIS-BetaRatio = BetaRatio x 19.3 + 93.3). BIS calculated from BcSEF was defined as BIS-BcSEF (BIS-BcSEF = BcSEF x 2.3 + 12.0). Figure 4A shows the changes in BIS, BIS-BetaRatio, and BIS-BcSEF in a patient. During the early period of anesthetic induction and emergence from anesthesia, the changes in BIS were similar to those in BIS-BetaRatio. During surgical levels of anesthesia, changes in BIS were similar to those in BIS-BcSEF. Using BIS-BetaRatio and BIS-BcSEF, we could calculate BIS in that case.
|
Figure 4B shows the relationship between BIS and BIS-BcSEF less than BIS 80 by a Bland-Altman analysis (12). The mean difference between BIS and BIS-BcSEF was 1.6, and the SD was 6.8 (n = 3337). Ninety-five percent of data were within mean ± 2 SD (1.6 ± 13.6). The small difference suggests that BIS-BcSEF is a comparable parameter to BIS during anesthesia.
One limitation of our study is the calculation of SynchFastSlow. In the preliminary study, we could not get an equivalent value when we used the formula shown in the review by Rampil (2). We reversed the denominator and numerator in the formula, as suggested by Sleigh et al. (5). Although we were able to obtain a reasonable value for SynchFastSlow, there is no way to confirm that the formula is correct.
In conclusion, BIS and BetaRatio correlated well in the range of BIS 60 to 100. In this range, BIS can be calculated from BetaRatio. BIS and SynchFastSlow, a parameter derived from bispectral analysis, or BcSEF, a parameter derived from power spectral analysis, are well correlated at surgical levels of anesthesia. However, our results show that SynchFastSlow has no advantage over BcSEF in the calculation of BIS.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
- Johansen JW, Sebel PS. Development and clinical application of electroencephalographic bispectrum monitoring. Anesthesiology 2000; 93: 1336–44.[Web of Science][Medline]
- Rampil IJ. A primer for EEG signal processing in anesthesia. Anesthesiology 1998; 89: 980–1002.[Web of Science][Medline]
- Bloom MJ. Electroencephalography and monitoring of anesthetic depth. In: Lake CL, Hines RL, Blitt CD, eds. Clinical monitoring. Philadelphia: WB Saunders, 2001: 92–102.
- Hagihira S, Takashina M, Mori T, et al. Practical issue in bispectral analysis of electroencephalographic signals. Anesth Analg 2001; 93: 966–70.
[Abstract/Free Full Text] - Sleigh JW, Steyn-Ross DA, Steyn-Ross ML, et al. Comparison of changes in electroencephalographic measures during induction of general anaesthesia: influence of the gamma frequency band and electromyogram signal. Br J Anaesth 2001; 86: 50–8.
[Abstract/Free Full Text] - Zar JH. Simple linear correlation. In: Zar JH, ed. Biostatistical analysis. 4th ed. Upper Saddle River, NJ: Prentice-Hall, 1999: 377–412.
- Glass PS, Bloom M, Kearse L, et al. Bispectral analysis measures sedation and memory effects of propofol, midazolam, isoflurane, and alfentanil in healthy volunteers. Anesthesiology 1997; 86: 836–47.[Web of Science][Medline]
- Sigl JC, Chamoun NG. An introduction to bispectral analysis for the electroencephalogram. J Clin Monit 1994; 10: 392–404.[Web of Science][Medline]
- Bruhn J, Bouillon TW, Shafer SL. Bispectral index (BIS) and burst suppression: revealing a part of the BIS algorithm. J Clin Monit Comput 2000; 16: 593–6.[Medline]
- Katoh T, Suzuki A, Ikeda K. Electroencephalographic derivatives as a tool for predicting the depth of sedation and anesthesia induced by sevoflurane. Anesthesiology 1998; 88: 642–50.[Web of Science][Medline]
- Detsch O, Schneider G, Kochs E, et al. Increasing isoflurane concentration may cause paradoxical increases in the EEG bispectral index in surgical patients. Br J Anaesth 2000; 84: 33–7.
[Abstract/Free Full Text] - Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307–10.[Web of Science][Medline]
This article has been cited by other articles:
|
|
 |
|
  M. S. Avidan, L. Zhang, B. A. Burnside, K. J. Finkel, A. C. Searleman, J. A. Selvidge, L. Saager, M. S. Turner, S. Rao, M. Bottros, et al. Anesthesia Awareness and the Bispectral Index N. Engl. J. Med., March 13, 2008; 358(11): 1097 - 1108. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Morimoto, S. Hagihira, S. Yamashita, Y. Iida, M. Matsumoto, S. Tsuruta, and T. Sakabe Changes in electroencephalographic bicoherence during sevoflurane anesthesia combined with intravenous fentanyl. Anesth. Analg., September 1, 2006; 103(3): 641 - 645. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Fodale, C. Pratico, M. Tescione, S. Tanania, T. Lucanto, and L. B. Santamaria Tramadol does not modify the Bispectral Index during anaesthesia with sevoflurane and remifentanil Br. J. Anaesth., August 1, 2005; 95(2): 212 - 215. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Heller, R. Hatami, P. Mullin, R. R. Sciacca, A. G. Khandji, M. Hamberger, R. Emerson, and E. J. Heyer Bilateral Bispectral Index Monitoring During Suppression of Unilateral Hemispheric Function Anesth. Analg., July 1, 2005; 101(1): 235 - 341. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Morimoto, A. Matsumoto, Y. Koizumi, T. Gohara, T. Sakabe, and S. Hagihira Changes in the Bispectral Index During Intraabdominal Irrigation in Patients Anesthetized with Nitrous Oxide and Sevoflurane Anesth. Analg., May 1, 2005; 100(5): 1370 - 1374. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Morimoto, Y. Monden, K. Ohtake, T. Sakabe, and S. Hagihira The Detection of Cerebral Hypoperfusion with Bispectral Index Monitoring During General Anesthesia Anesth. Analg., January 1, 2005; 100(1): 158 - 161. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
