Distinguishing Anesthetized from Awake State in Patients: A New Approach Using One Second Segments of Raw EEG

Bjørn E Juel, Luis Romundstad, Frode Kolstad, Johan F Storm, Pål G Larsson, Bjørn E Juel, Luis Romundstad, Frode Kolstad, Johan F Storm, Pål G Larsson

Abstract

Objective: The objective of this study was to test whether properties of 1-s segments of spontaneous scalp EEG activity can be used to automatically distinguish the awake state from the anesthetized state in patients undergoing general propofol anesthesia. Methods: Twenty five channel EEG was recorded from 10 patients undergoing general intravenous propofol anesthesia with remifentanil during anterior cervical discectomy and fusion. From this, we extracted properties of the EEG by applying the Directed Transfer Function (DTF) directly to every 1-s segment of the raw EEG signal. The extracted properties were used to develop a data-driven classification algorithm to categorize patients as "anesthetized" or "awake" for every 1-s segment of raw EEG. Results: The properties of the EEG signal were significantly different in the awake and anesthetized states for at least 8 of the 25 channels (p < 0.05, Bonferroni corrected Wilcoxon rank-sum tests). Using these differences, our algorithms achieved classification accuracies of 95.9%. Conclusion: Properties of the DTF calculated from 1-s segments of raw EEG can be used to reliably classify whether the patients undergoing general anesthesia with propofol and remifentanil were awake or anesthetized. Significance: This method may be useful for developing automatic real-time monitors of anesthesia.

Keywords: consciousness; directed transfer function (DTF); electroencephalography (EEG); general anesthesia; monitoring general anesthesia.

Figures

Figure 1
Figure 1
Examples of raw EEG and time-frequency plots for a channel (Cz of patient 6) and the LDTF timecourse across the transition from wake to anesthesia. (A,B) Show zoomed segments of raw EEG samples and time-frequency plot taken from the regions marked by blue vertical bars in (C). The wavelet transform used 2 s windows, with 68 wavelets (covering 2–70 Hz) each with 12 cycles within the window. (C) Shows a raw EEG (top), time-frequency plot of wavelet transformed EEG data (middle), and LDTF source strength timecourse (bottom) accross the wakefulness to anesthesia transition (the red vertical line marks the time at which the anesthesiologist indicated the patient was anesthetized).
Figure 2
Figure 2
Explanations of the plots used to visualize the DTF analysis. Each panel shows different properties of the connectivity as quantified by the DTF in our analysis. (A) Visualization of the full connectivity matrix within a given frequency range (here 8–12 Hz) for a specific 1-s segment from a given patient (#6). Each element in this matrix represents the median LDTF from a given source channel (x-axis) to a given sink channel (y-axis), and the strength is given by the color (color bar, right). Two examples are indicated by black arrows and circles in the matrix, showing the relatively strong (red) information flow from Cz to T7, and the relatively weak (blue) information flow in the opposite direction, from T7 to Cz. (B) Topographical representation of the source channels' strengths of information outflow from the same 1-s segment presented in (A). The positons of channel labels indicate their relative locations on the head as seen from above (nose pointing up). In this 1-s segment, the medial channels are strong, whilst more lateral channels are weaker, sources of information outflow. (C) Shows how each channel's strength of information outflow changes with time (x-axis). Red arrows mark important points. Here, it marks the time of the 1-s segment visualized in the figures in (A,B). All figures use the same color scale, shown by the color bar.
Figure 3
Figure 3
Qualitative summary of the DTF analysis. (A,B) Summarize the structure of median connectivity in the “awake” and “anesthetized” states respectively, given by LDTF, calculated from the data pooled from all patients. (A2,B2) Show the full directed connectivity matrices. Each element in the matrix quantifies the median information flow, mLDTFij, between the j'th source channel (x-axis) and the i'th sink channel (y-axis). The topographical plots show the distribution of information outflow and inflow regions (sources and sinks) on the scalp. (A3,B3) Depict the median information outflow from each point, while (A1,B1) depict the information inflow to each point, on the scalp. (C) Shows the time courses of information outflow from all channels for a single patient (#7), illustrating the dynamics of the measure over time. The time-course is smoothed using a moving median calculation, taking the previous five 1-s segments into consideration. The two red arrows indicate the time of loss of verbal contact and response (left) and the time point when the patient again responded verbally (right). Note that only a few of the channels in (A2,B2,C) are labeled because of space restrictions. See Figure 2 for the full set of channel labels. All figures show the median logarithm of DTF within the alpha frequency range. The color scale on the right indicates the LDTF values for all panels: red colors indicate strong, while blue colors indicate weak, information flow.
Figure 4
Figure 4
Comparing the distributions of information outflow from different EEG channels. Each plot in (A) visualizes the distributions of information outflow values from a single channel, from the “awake” (blue) and “anesthetized” (red) states. Again, the data is pooled from all patients in the given state. The vertical positions of the stars indicate the proportion of outliers on the indicated sides of the distributions. The axes are equal in all panels [shown only for F9, upper left in (A)], with the y-axis showing the probability density of observations (range: 0 to 1), and the x-axis indicating the LDTF value (range: −10 to 0). The panels are ordered topographically as if the head is seen from above with the nose pointing up. The panels with a bright green frame mark the channels with significantly different distributions between the “awake” and “anesthetized” states. In (B), the same information is given as boxplots with the same color coding. The mean of each distribution is given by a filled circle in each box, and the whiskers indicate the standard deviation of the distributions. Outliers have been omitted. The y-axis shows the information outflow, and corresponds to the x-axes of the plots in (A). (C) Shows a topographical plot of the results from the statistical test comparing the distributions of LDTF information outflow values for each channel between awake and anesthetized states—controlling for autocorrelation in the information outflow, resampling to make the data sizes similar, and correcting for multiple comparisons. Green colors indicate regions with significantly different median information outflow distributions in awake and anesthetized states.
Figure 5
Figure 5
Visualization of results from classification algorithm. Here, data from all channels with significantly different distributions of median information outflow between states were used for continuous classification the patient state. (B–I, zoomed example in A). Show the time course of the LDTF information outflow channels for each patient, the patient's state as reported by clinical staff, and the classification made by the algorithm. The middle region of each panel, containing the LDTF values for every channel, follows the color scheme defined in the color bar. The top and bottom bars show the corresponding conscious state of the patient as suggested by the algorithm and reported by the clinical staff respectively, and have their own color scheme: blue means “awake” while red means “anesthetized.” (A) Shows the same information as (D), just enlarged to give a better impression of the dynamics, and numerated axes. Possible misclassifications can be seen as blue points or regions in the upper bars of each panel where one would expect a red color according to the clinical report (i.e., false positives: predicting “awake” when clinicians report “anesthetized”), and red regions where the clinical report suggest a blue color (false negatives: predicting anesthetized where clinicians report “awake”).

References

    1. Alkire M. T. (2008). Loss of effective connectivity during general anesthesia. Int. Anesthesiol. Clin. 46, 55–73. 10.1097/AIA.0b013e3181755dc6
    1. Alkire M. T., Hudetz A. G., Tononi G. (2008). Consciousness and anesthesia. Science 322, 876–880. 10.1126/science.1149213
    1. Avidan M. S., Zhang L., Burnside B. A., Finkel K. J., Searleman A. C., Selvidge J. A., et al. . (2008). Anesthesia awareness and the bispectral index. N. Engl. J. Med. 358, 1097–1108. 10.1056/NEJMoa0707361
    1. Bertini M., Ferrara M., De Gennaro L., Curcio G., Moroni F., Babiloni C., et al. . (2009). Directional information flows between brain hemispheres across waking, non-REM and REM sleep states: an EEG study. Brain Res. Bull. 78, 270–275. 10.1016/j.brainresbull.2008.12.006
    1. Bischoff P., Rundshagen I. (2011). Awareness under general anesthesia. Dtsch. Ärztebl. Int. 108, 1–7. 10.3238/arztebl.2011.0001
    1. Boly M., Garrido M. I., Gosseries O., Bruno M.-A., Boveroux P., Schnakers C., et al. . (2011). Preserved feedforward but impaired top-down processes in the vegetative state. Science 332, 858–862. 10.1126/science.1202043
    1. Boly M., Massimini M., Tsychiya N., Postle B. R., Koch C., Tononi G. (2017). Are the neural correlates of consciousness in the front or in the back of the cerebral cortex? Clinical and neuroimaging evidence. bioRxiv 118273. 10.1101/118273
    1. Boly M., Moran R., Murphy M., Boveroux P., Bruno M.-A., Noirhomme Q., et al. . (2012). Connectivity changes underlying spectral EEG changes during propofol-induced loss of consciousness. J. Neurosci. 32, 7082–7090. 10.1523/JNEUROSCI.3769-11.2012
    1. Brunner C., Billinger M., Seeber M., Mullen T. R., Makeig S. (2016). Volume conduction influences scalp-based connectivity estimates. Front. Comput. Neurosci. 10:121. 10.3389/fncom.2016.00121
    1. Casali A. G., Gosseries O., Rosanova M., Boly M., Sarasso S., Casali K. R., et al. . (2013). A theoretically based index of consciousness independent of sensory processing and behavior. Sci. Transl. Med. 5:198ra105. 10.1126/scitranslmed.3006294
    1. Chernik D. A., Gillings D., Laine H., Hendler J., Silver J. M., Davidson A. B., et al. . (1990). Validity and reliability of the observer's assessment of alertness/sedation scale: study with intravenous midazolam. J. Clin. Psychopharmacol. 10, 244–251.
    1. Delorme A., Makeig S. (2004). EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J. Neurosci. Methods 134, 9–21. 10.1016/j.jneumeth.2003.10.009
    1. Ding M., Bressler S. L., Yang W., Liang H. (2000). Short-window spectral analysis of cortical event-related potentials by adaptive multivariate autoregressive modeling: data preprocessing, model validation, and variability assessment. Biol. Cybern. 83, 35–45. 10.1007/s004229900137
    1. Di Perri C., Stender J., Laureys S., Gosseries O. (2014). Functional neuroanatomy of disorders of consciousness. Epilepsy Behav. 30, 28–32. 10.1016/j.yebeh.2013.09.014
    1. Dukic S., Iyer P. M., Mohr K., Hardiman O., Lalor E. C., Nasseroleslami B. (2017). Estimation of coherence using the median is robust against EEG artefacts, in 2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) (Seogwipo: ), 3949–3952.
    1. Eer A. S., Padmanabhan U., Leslie K. (2009). Propofol dose and incidence of dreaming during sedation. Eur. J. Anaesthesiol. 26, 833–836. 10.1097/EJA.0b013e32832c500c
    1. Ferrarelli F., Massimini M., Sarasso S., Casali A., Riedner B. A., Angelini G., et al. . (2010). Breakdown in cortical effective connectivity during midazolam-induced loss of consciousness. Proc. Natl. Acad. Sci. U.S.A. 107, 2681–2686. 10.1073/pnas.0913008107
    1. Florin E., Gross J., Pfeifer J., Fink G. R., Timmermann L. (2011). Reliability of multivariate causality measures for neural data. J. Neurosci. Methods 198, 344–358. 10.1016/j.jneumeth.2011.04.005
    1. Gennaro L. D., Vecchio F., Ferrara M., Curcio G., Rossini P. M., Babiloni C. (2004). Changes in fronto-posterior functional coupling at sleep onset in humans. J. Sleep Res. 13, 209–217. 10.1111/j.1365-2869.2004.00406.x
    1. Gennaro L. D., Vecchio F., Ferrara M., Curcio G., Rossini P. M., Babiloni C. (2005). Antero-posterior functional coupling at sleep onset: changes as a function of increased sleep pressure. Brain Res. Bull. 65, 133–140. 10.1016/j.brainresbull.2004.12.004
    1. Ghoneim M. M., Block R. I., Haffarnan M., Mathews M. J. (2009). Awareness during anesthesia: risk factors, causes and sequelae: a review of reported cases in the literature. Anesth. Analg. 108, 527–535. 10.1213/ane.0b013e318193c634
    1. Goddard N., Smith D. (2013). Unintended awareness and monitoring of depth of anaesthesia. Contin. Educ. Anaesth. Crit. Care 13, 213–217. 10.1093/bjaceaccp/mkt016
    1. Greenblatt R. E., Pflieger M. E., Ossadtchi A. E. (2012). Connectivity measures applied to human brain electrophysiological data. J. Neurosci. Methods 207, 1–16. 10.1016/j.jneumeth.2012.02.025
    1. He B., Dai Y., Astolfi L., Babiloni F., Yuan H., Yang L. (2011). eConnectome: a MATLAB toolbox for mapping and imaging of brain functional connectivity. J. Neurosci. Methods 195, 261–269. 10.1016/j.jneumeth.2010.11.015
    1. Höller Y., Thomschewski A., Bergmann J., Kronbichler M., Crone J. S., Schmid E. V., et al. . (2014). Connectivity biomarkers can differentiate patients with different levels of consciousness. Clin. Neurophysiol. 125, 1545–1555. 10.1016/j.clinph.2013.12.095
    1. Hudetz A. G. (2012). General anesthesia and human brain connectivity. Brain Connect. 2, 291–302. 10.1089/brain.2012.0107
    1. Hudetz A. G., Mashour G. A. (2016). Disconnecting consciousness: is there a common anesthetic end point? Anesth. Analges. 123, 1228–1240. 10.1213/ANE.0000000000001353
    1. Juel B. E., Romundstad L. G., Kolstad F., Storm J. F., Larsson P. G. (2016). Differences in effective connectivity can be used to separate conscious from unconscious states in patients undergoing general anesthesia, in 10th FENS Forum of Neuroscience (Copenhagen: ).
    1. Kaminski M., Blinowska K. J. (2014). Directed transfer function is not influenced by volume conduction—inexpedient pre-processing should be avoided. Front. Comput. Neurosci. 8:61 10.3389/fncom.2014.00061
    1. Kaminski M., Blinowska K. J. (2017). The influence of volume conduction on DTF estimate and the problem of its mitigation. Front. Comput. Neurosci. 11:36. 10.3389/fncom.2017.00036
    1. Kaminski M., Blinowska K., Szelenberger W. (1995). Investigation of coherence structure and EEG activity propagation during sleep. Acta Neurobiol. Exp. 55, 213–219.
    1. Kaminski M., Ding M., Truccolo W. A., Bressler S. L. (2001). Evaluating causal relations in neural systems: granger causality, directed transfer function and statistical assessment of significance. Biol. Cybern. 85, 145–157. 10.1007/s004220000235
    1. Kaminski M. J., Blinowska K. J. (1991). A new method of the description of the information flow in the brain structures. Biol. Cybern. 65, 203–210. 10.1007/BF00198091
    1. Koch C., Massimini M., Boly M., Tononi G. (2016). Neural correlates of consciousness: progress and problems. Nat. Rev. Neurosci. 17, 307–321. 10.1038/nrn.2016.22
    1. Kreuzer M. (2017). EEG based monitoring of general anesthesia: taking the next steps. Front. Comput. Neurosci. 11:56. 10.3389/fncom.2017.00056
    1. Laureys S. (2005). The neural correlate of (un)awareness: lessons from the vegetative state. Trends Cogn. Sci. 9, 556–559. 10.1016/j.tics.2005.10.010
    1. Lee U., Kim S., Noh G.-J., Choi B.-M., Hwang E., Mashour G. A. (2009). The directionality and functional organization of frontoparietal connectivity during consciousness and anesthesia in humans. Conscious. Cogn. 18, 1069–1078. 10.1016/j.concog.2009.04.004
    1. Lee U., Ku S., Noh G., Baek S., Choi B., Mashour G. A. (2013). Disruption of frontal-parietal communication by ketamine, propofol, and sevoflurane. Anesthesiology 118, 1264–1275. 10.1097/ALN.0b013e31829103f5
    1. Mashour G. A., Shanks A., Tremper K. K., Kheterpal S., Turner C. R., Ramachandran S. K., et al. . (2012). Prevention of intraoperative awareness with explicit recall in an unselected surgical population: a randomized comparative effectiveness trial. Anesthesiology 117, 717–725. 10.1097/ALN.0b013e31826904a6
    1. Mhuircheartaigh R. N., Warnaby C., Rogers R., Jbabdi S., Tracey I. (2013). Slow-wave activity saturation and thalamocortical isolation during propofol anesthesia in humans. Sci. Transl. Med. 5:208ra148. 10.1126/scitranslmed.3006007
    1. Minto C. F., Schnider T. W., Shafer S. L. (1997). Pharmacokinetics and pharmacodynamics of remifentanil. II. model application. Anesthesiology 86, 24–33.
    1. Musialowicz T., Lahtinen P. (2014). Current status of EEG-based depth-of-consciousness monitoring during general anesthesia. Curr. Anesthesiol. Rep. 4, 251–260. 10.1007/s40140-014-0061-x
    1. Mychaskiw G., Horowitz M., Sachdev V., Heath B. J. (2001). Explicit intraoperative recall at a bispectral index of 47. Anesth. Analg. 92, 808–809. 10.1097/00000539-200104000-00003
    1. Myles P. S., Leslie K., McNeil J., Forbes A., Chan M. T. V. (2004). Bispectral index monitoring to prevent awareness during anaesthesia: the B-aware randomised controlled trial. Lancet 363, 1757–1763. 10.1016/S0140-6736(04)16300-9
    1. Pilge S., Zanner R., Schneider G., Blum J., Kreuzer M., Kochs E. F. (2006). Time delay of index calculation: analysis of cerebral state, bispectral, and narcotrend indices. Anesthesiology 104, 488–494. 10.1097/00000542-200603000-00016
    1. Punjasawadwong Y., Phongchiewboon A., Bunchungmongkol N. (2014). Bispectral index for improving anaesthetic delivery and postoperative recovery. Cochrane Database Syst. Rev. 6:CD003843 10.1002/14651858.CD003843.pub3
    1. Purdon P. L., Pierce E. T., Mukamel E. A., Prerau M. J., Walsh J. L., Wong K. F. K., et al. . (2013). Electroencephalogram signatures of loss and recovery of consciousness from propofol. Proc. Natl. Acad. Sci. 110, E1142–E1151. 10.1073/pnas.1221180110
    1. Schartner M., Seth A., Noirhomme Q., Boly M., Bruno M.-A., Laureys S., et al. . (2015). Complexity of multi-dimensional spontaneous EEG decreases during propofol induced general anaesthesia. PLoS ONE 10:e0133532. 10.1371/journal.pone.0133532
    1. Schnakers C., Ledoux D., Majerus S., Damas P., Damas F., Lambermont B., et al. . (2008). Diagnostic and prognostic use of bispectral index in coma, vegetative state and related disorders. Brain Injury 22, 926–931. 10.1080/02699050802530565
    1. Schneider G., Wagner K., Reeker W., Hänel F., Werner C., Kochs E. (2002). Bispectral index (BIS) may not predict awareness reaction to intubation in surgical patients. J. Neurosurg. Anesthesiol. 14, 7–11. 10.1097/00008506-200201000-00002
    1. Schnider T. W., Minto C. F., Gambus P. L., Andresen C., Goodale D. B., Shafer S. L., et al. . (1998). The influence of method of administration and covariates on the pharmacokinetics of propofol in adult volunteers. Anesthesiology 88, 1170–1182. 10.1097/00000542-199805000-00006
    1. Schumacher E. M., Stiris T. A., Larsson P. G. (2015). Effective connectivity in long-term EEG monitoring in preterm infants. Clin. Neurophysiol. 126, 2261–2268. 10.1016/j.clinph.2015.01.017
    1. Schumacher E. M., Westvik A. S., Larsson P. G., Lindemann R., Westvik J., Stiris T. A. (2011). Feasibility of long-term continuous EEG monitoring during the first days of life in preterm infants: an automated quantification of the EEG activity. Pediatr. Res. 69(5Pt 1), 413–417. 10.1203/PDR.0b013e31821267d2
    1. Sebel P. S., Bowdle T. A., Ghoneim M. M., Rampil I. J., Padilla R. E., Gan T. J., et al. . (2004). The incidence of awareness during anesthesia: a multicenter United States study. Anesth. Analg. 99, 833–839. 10.1213/01.ANE.0000130261.90896.6C
    1. Storm J. F., Boly M., Casali A. G., Massimini M., Olcese U., Pennartz C. M. A., et al. . (2017). Consciousness regained: disentangling mechanisms, brain systems, and behavioral responses. J. Neurosci. 37, 10882–10893. 10.1523/JNEUROSCI.1838-17.2017
    1. Vuyk J., Lichtenbelt B. J., Vieveen J., Dahan A., Engbers F. H. M., Burm A. G. L. (2004). Low bispectral index values in awake volunteers receiving a combination of propofol and midazolam. J. Am. Soc. Anesthesiol. 100, 179–181. 10.1097/00000542-200401000-00029

Source: PubMed

Patrocinadores y Colaboradores

Condiciones médicas

Intervenciones de drogas

3
Suscribir