A novel approach for detection of consciousness level in comatose patients from EEG signals with 1-D convolutional neural network

• Autorzy: Altıntop Çiğdem Gülüzar , Latifoğlu Fatma , Akın Aynur Karayol , Çetin Bilge

Identyfikatory

DOI

10.1016/j.bbe.2021.11.003

• Języki publikacji: EN

Abstrakty

EN

Coma is an unresponsive state of unconsciousness from which a person cannot be awakened. Glasgow Coma Score (GCS) is a clinical scale for determining the depth and length of a coma. GCS plays an important role in effective and accurate patient evaluation and is critical in planning the right treatment modalities and patient care because it shows patient outcomes and is a measurement performed several times a day. The GCS is universally accepted as a gold standard and validated scale for assessing a patient’s level of consciousness. However, the scale’s success has been questioned due to variations in interobserver reliability performance. In this study, the data set generated from Electroencephalography (EEG) signals obtained from 39 comatose patients was used in the training of deep neural networks for the classification of consciousness level. The EEG signals were recorded during nurse and family interaction with comatose patients. The level of consciousness was classified with the proposed 1D-CNN model. Consequently, the two classes that we label as low and high consciousness are classified with 83.3% accuracy. To our best knowledge, no prior studies are using 1D-CNN for the classification of EEG-based level of consciousness using the proposed recording process. Our study is unique from other studies in terms of recording procedure and methods.

Słowa kluczowe

EN

Glasgow coma score; electroencephalography; convolutional neural network; level of consciousness

PL

skala Glasgow; elektroencefalografia; sieć neuronowa konwolucyjna; poziom świadomości

• Wydawca: Nałęcz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences ; Elsevier
• Czasopismo: Biocybernetics and Biomedical Engineering
• Rocznik: 2022
• Tom: Vol. 42, no. 1
• Strony: 16--26
• Opis fizyczny: Bibliogr. 76 poz., rys., tab., wykr.

Twórcy

autor

Altıntop Çiğdem Gülüzar

• Department of Biomedical Engineering, Erciyes University, Kayseri, Turkey

autor

Latifoğlu Fatma

• Erciyes University, Engineering Faculty, Biomedical Eng. Dept., 39039 Kayseri, Turkey

autor

Akın Aynur Karayol

• Department of Anesthesiology and Reanimation, Erciyes University, Kayseri, Turkey

autor

Çetin Bilge

• Department of Anesthesiology and Reanimation, Memorial Hospital, Kayseri, Turkey

Bibliografia

• [1] Posner J, Plum F, Saper C. Plum and Posner’s Diagnosis of Stupor and Coma. New York, NY: Oxford University Press; 2007.
• [2] Young, Coma GB. Ann N Y Acad Sci 2009;1157(1):32–47.
• [3] Teasdale G. Forty years on: updating the Glasgow Coma Scale. Nurs Times 2014;110(42):12–6.
• [4] Teasdale G, Murray G, Parker L, Jennett B. Adding up the Glasgow Coma Score. Acta Neurochir Suppl (Wien) 1979;28 (1):13–6.
• [5] Mehta R, Chinthapalli K. Glasgow coma scale explained. BMJ 2019;365 l1296.
• [6] Bledsoe BE, Casey MJ, Feldman J, Jhonson L, Diel S, et al. Glasgow Coma Scale scoring is often inaccurate. Prehosp Disaster Med. 2015;30(1):46–53.
• [7] Gill MR, Reiley DG, Green SM. Interrater reliability of Glasgow Coma Scale scores in the emergency department. Ann Emerg Med 2004;43(2):215–23.
• [8] Kebapçı A, Dikeç G, Topçu S. Interobserver reliability of glasgow coma scale scores for intensive care unit patients. Crit Care Nurse 2020;40(4):e18–26.
• [9] Moe HK, Follestad T, Andelic N, Håberg AK, Flusund AMH, et al. Traumatic axonal injury on clinical MRI: association with the Glasgow Coma Scale score at scene of injury or at admission and prolonged posttraumatic amnesia. J Neurosurg 2020;1(aop):1–12.
• [10] Jiang M, Su Y, Liu G, Chen W, Gao D. Predicting the nonsurvival outcome of large hemispheric infarction patients via quantitative electroencephalography: Superiority to visual electroencephalography and the Glasgow Coma Scale. Neurosci Lett 2019;706:88–92.
• [11] Baker M. Reviewing the application of the Glasgow Coma Scale: Does it have interrater reliability? Br J Neurosci Nurs 2008;4:342–7.
• [12] Altıntop C¸ G, Latifoğlu F, Akın AK, Ileri R, Yazar MA. Analysis of consciousness level using galvanic skin response during therapeutic effect. J Med Syst 2021;45(1):1–12.
• [13] Reith FC, Synnot A, Van den Brande R, Gruen RL, Maas AI. Factors influencing the reliability of the Glasgow Coma Scale: a systematic review. Neurosurgery 2017;80(6):829–39.
• [14] Kubota Y, Nakamoto H, Egawa S, Kawamata T. Continuous EEG monitoring in ICU. J Intens Care 2018;6(1):1–8.
• [15] Herman ST, Abend NS, Bleck TP, Chapman KE, Drislane FW, et al. Consensus statement on continuous EEG in critically ill adults and children, Part II: personnel, technical specifications and clinical practice. J Clin Neurophysiol 2015;32(2):96.
• [16] LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015;521 (7553):436–44.
• [17] Ravì D, Wong C, Deligianni F, Berthelot M, Andreu-Perez J, et al. Deep learning for health informatics. IEEE J Biomed Health Inf 2016;21(1):4–21.
• [18] Lecun Y, Bengio Y. Convolutional networks for images, speech, and time series. In The handbook of brain theory and neural networks. MIT Press, 1998.
• [19] Rumelhart DE, Hinton GE, Williams RJ. Learning representations by back-propagating errors. Nature 1986;323:533–6.
• [20] Huang S, Tang J, Dai J, Wang Y. Signal status recognition based on 1DCNN and its feature extraction mechanism analysis. Sensors 2019;19(9):2018.
• [21] Kiranyaz S, Avci O, Abdeljaber O, Ince T, Gabbouj M, et al. 1D convolutional neural networks and applications: A survey. Mech Syst Sig Process 2021;151 107398.
• [22] Kiranyaz S, Ince T, Gabbouj M. Personalized monitoring and advance warning system for cardiac arrhythmias. Sci Rep Nat 2017;7(1):1–8.
• [23] Luo C, Jiang H, Li Q, Rao N. Multi-label classification of abnormalities in 12-lead ecg using 1d cnn and lstm. In Machine Learning and Medical Engineering for Cardiovascular Health and Intravascular Imaging and Computer Assisted Stenting, 2019, pp. 55-63, Springer, Cham.
• [24] Nannavecchia A, Girardi F, Fina PR, Scalera M, Dimauro G. Personal heart health monitoring based on 1D convolutional neural network. J Imaging 2021;7(2):26.
• [25] Mukhopadhyay R, Panigrahy P. S, Misra G, Chattopadhyay P. Quasi 1D CNN-based fault diagnosis of induction motor drives. In: 2018 5th International Conference on Electric Power and Energy Conversion Systems (EPECS); Kitakyushu, Japan; 2018. pp. 1–5.
• [26] Eren L, Ince T, Kiranyaz S. A generic intelligent bearing fault diagnosis system using compact adaptive 1D CNN classifier. J Signal Process Syst 2019;91(2):179–89.
• [27] Liu C, Wang F, Hu J, et al. Risk prediction with electronic health records: a deep learning approach. In: ACM International Conference on Knowledge Discovery and Data Mining; Sydney, NSW, Australia; 2015. pp. 705–14.
• [28] Nurmaini S, Tondas AE, Darmawahyuni A, Rachmatullah MN, Partan RU, et al. Robust detection of atrial fibrillation from short-term electrocardiogram using convolutional neural networks. Fut Gen Comput Syst 2020;113:304–17.
• [29] Li F, Liu M, Zhao Y, Kong L, Dong L, et al. Feature extraction and classification of heart sound using 1D convolutional neural networks. EURASIP J Adv Signal Process 2019;1:1–11.
• [30] Khalili E, Asl BM. Automatic sleep stage classification using temporal convolutional neural network and new data augmentation technique from raw single-channel EEG. Comput Methods Programs Biomed 2021;204 106063.
• [31] Phang CR, Norman F, Hussain H, Ting CM, Ombao HA. Multidomain connectome convolutional neural network for identifying Schizophrenia from EEG connectivity patterns. IEEE J Biomed Health Inf 2019;24(5):1333–43.
• [32] Liu Y, Sivathamboo S, Goodin P, Bonnington P, Kwan P, et al. Epileptic seizure detection using convolutional neural network: A multi-biosignal study. In: Proceedings of the Australasian Computer Science Week Multiconference. p. 1–8.
• [33] Chaabene S, Bouaziz B, Boudaya A, Hökelmann A, Ammar A, Chaari L. Convolutional neural network for drowsiness detection using EEG signals. Sensors 2021;21(5):1734.
• [34] Zhou M, Tian C, Cao R, Wang B, Niu Y, Hu T, et al. Epileptic seizure detection based on EEG signals and CNN. Front Neuroinf 2018;12:95.
• [35] Yang B, Zhu X, Liu Y, Liu H. A single-channel EEG based automatic sleep stage classification method leveraging deep one-dimensional convolutional neural network and hidden Markov model. Biomed Signal Process Control 2021;68 102581.
• [36] Abel JH, Badgeley MA, Meschede-Krasa B, Schamberg G, Garwood IC, Lecamwasam K, et al. Machine learning of EEG spectra classifies unconsciousness during GABAergic anesthesia. PLoS One 2021;16(5) e0246165.
• [37] Oh SL, Hagiwara Y, Raghavendra U, et al. A deep learning approach for Parkinson’s disease diagnosis from EEG signals. Neural Comput Appl 2020;32:10927–33.
• [38] Khanmohammadi S, Laurido-Soto O, Eisenman LN, Kummer TT, Ching S. Intrinsic network reactivity differentiates levels of consciousness in comatose patients. Clin Neurophysiol 2018;129(11):2296–305.
• [39] Lechinger J, Bothe K, Pichler G, et al. CRS-R score in disorders of consciousness is strongly related to spectral EEG at rest. J Neurol 2013;260:2348–56.
• [40] La Foresta F, Mammone N, Morabito FC. PCA–ICA for automatic identification of critical events in continuous coma-EEG monitoring. Biomed Signal Process Control 2009;4 (3):229–35.
• [41] Wieser M, Koenig BA, Riener R, Quantitative description of the state of awareness of patients in vegetative and minimally conscious state, 32nd Annual International Conference of the IEEE EMBS Buenos Aires, Argentina, 5533– 5536, August 31–September 4, 2010.
• [42] Prichep LS, Jacquin A, Filipenko J, Dastidar SG, Zabele S, Vodenčarević A, et al. Classification of traumatic brain injury severity using informed data reduction in a series of binary classifier algorithms. IEEE Trans Neural Syst Rehabil Eng 2012;20(6):806–22.
• [43] Armanfard N, Komeili M, Reilly JP, Connolly JF. A machine learning framework for automatic and continuous MMN detection with preliminary results for coma outcome prediction. IEEE J Biomed Health Inf 2019;23(4):1794–804.
• [44] Schartner M, Seth A, Noirhomme Q, Boly M, Bruno M-A, Laureys S, et al. Complexity of multi-dimensional spontaneous EEG decreases during propofol induced general anaesthesia. PLoS ONE 2015;10:1–21.
• [45] Cui G, Yin Y, Zhao Q, Cichocki A, Cao J. Patients’ consciousness analysis using Dynamic Approximate Entropy and MEMD method, 2013 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, pp. 1–5, 2013.
• [46] Gonzalez Andino S, Grave de Peralta R, Menendez, Thut G, Spinelli L, Blanke O, et al. Measuring the complexity of time series: an application to neurophysiological signals. Hum Brain Mapp 2000;11:46–57.
• [47] Tang L, Lv H, Yang F, Yu L. Complexity testing techniques for time series data: A comprehensive literature review. Chaos Soliton Fract 2015;81:117–35.
• [48] Zhang N, Zhang J, Cao J, Miao Y, Chu L, Kong W. Analysis of deep coma and brain death based on phase synchronization with EEG. In: Proceedings of ICSP2018. p. 143–7.
• [49] Lin M, Chan H, Fang S. Linear and nonlinear EEG indexes in relation to the severity of coma. Conf Proc IEEE Eng Med Biol Soc 2005;5:4580–3.
• [50] Górska U, Rupp A, Celikel T, Englitz B. Assessing the state of consciousness for individual patients using complex, statistical stimuli. NeuroImage: Clinical 2021;29 102471.
• [51] Michal T. Fundamental of EEG measurement. Meas Sci Rev 2002;2(2):1–11.
• [52] Nunez PL. Neocortical dynamics and human EEG rhythms. New York: Oxford University Press; 1995.
• [53] Blakemore & Frith. The Learning Brain. Blackwell Publishing; 2005.
• [54] Schacter Daniel L, Gilbert Daniel T, Wegner Daniel M. Psychology. 2nd ed. New York: Worth Publishers; 2010.
• [55] Toraman S. Automatic recognition of preictal and interictal EEG signals using 1D-capsule networks. Comput Electr Eng 2021;91 107033.
• [56] Chakrabarti S, Swetapadma A, Pattnaik PK. A channel independent generalized seizure detection method for pediatric epileptic seizures. Comput Methods Programs Biomed 2021;209 106335.
• [57] Jijomon CM, Vinod AP. Person-identification using familiarname auditory evoked potentials from frontal EEG electrodes. Biomed Signal Process Control 2021;68 102739.
• [58] Liu X, Xie Q, Lv J, Huang H, Wang W. P300 event-related potential detection using one-dimensional convolutional capsule networks. Expert Syst Appl 2021;174 114701.
• [59] Satapathy SK, Loganathan D, Narayanan P, Sharathkumar S. Convolutional neural network for classification of multiple sleep stages from dual-channel EEG signals. In: 2020 IEEE 4th Conference on Information & Communication Technology (CICT). IEEE; 2020. p. 1–16.
• [60] Sun Y, Lo FPW, Lo B. EEG-based user identification system using 1D-convolutional long short-term memory neural networks. Expert Syst Appl 2019;125:259–67.
• [61] LeCun Y, Kavukcuoglu K, Farabet C. Convolutional networks and applications in vision. In: Proceedings of 2010 IEEE International Symposium on Circuits and Systems. p. 253–6.
• [62] Zhang Q, Zhang M, Chen T, Sun Z, Ma Y, et al. Recent advances in convolutional neural network acceleration. Neurocomputing 2019:37–51.
• [63] Wiatowski T, Bölcskei H. A mathematical theory of deep convolutional neural networks for feature extraction. IEEE Trans Inf Theory 2018;64(3):1845–66.
• [64] Munce SEP, Webster F, Christian J, Gonzalez-Lara LE, Owe AM, et al. Experiences of family of individuals in a locked in, minimally conscious state, or vegetative state with the health care system. Brain Inj 2021:1–7.
• [65] Van Stralen KJ, Stel VS, Reitsma JB, Dekker FW, Zoccali C, et al. Diagnostic methods I: sensitivity, specificity, and other measures of accuracy. Kidney Int 2009;75(12):1257–63.
• [66] Schnakers C, Vanhaudenhuyse A, Giacino JT, Ventura M, Boly M, et al. Diagnostic accuracy of the vegetative and minimally conscious state: clinical consensus versus standardized neurobehavioral assessment. BMC Neurol 2009;9(1):1–5.
• [67] Kulyk OV. Nonlinear analysis of quantitative EEGs in patients with syndromes of post-coma disorders of consciousness after severe traumatic brain injury. Neurophysiology 2018;50:456–65.
• [68] Lehembre R, Bruno MA, Vanhaudenhuyse A, Chatelle C, Cologan V, et al. Resting-state EEG study of comatose patients: a connectivity and frequency analysis to find differences between vegetative and minimally conscious states. Funct Neurol 2012;27(1):41–7.
• [69] Gosseries O, Schnakers C, Ledoux D, Vanhaudenhuyse A, Bruno MA, et al. Automated EEG entropy measurements in coma, vegetative state/unresponsive wakefulness syndrome and minimally conscious state. Funct Neurol 2011;26 (1):25–30.
• [70] Miao Y, Cao J. Descriptive statistical analysis based on patients’ EEG energy in coma and quasi-brain-death state. Int J Comput Technol 2018;17(1).
• [71] Shi Q, Yang J, Cao J, Tanaka T, Wang R, et al. EEG data analysis based on EMD for coma and quasi-brain-death patients. J Exp Theor Artif Intell 2011;23(1):97–110.
• [72] Xu Z, Wang J, Wang R, Zhang Z, Yang S. Deep convolutional neural network for detection of disorders of consciousness. In: 2020 39th Chinese Control Conference (CCC). p. 7084–9.
• [73] Yuan L, Cao J. Patients’ EEG data analysis via spectrogram image with a convolution neural network. In: International Conference on Intelligent Decision Technologies; 2017. p. 13–21.
• [74] Lee M, Yeom SK, Baird B, Gosseries O, Nieminen JO, et al. Spatio-temporal analysis of EEG signal during consciousness using convolutional neural network. In: 2018 6th International Conference on Brain-Computer Interface (BCI). p. 1–3.
• [75] Sun H, Kimchi E, Akeju O, Nagaraj SB, McClain LM, et al. Automated tracking of level of consciousness and delirium in critical illness using deep learning. NPJ Digital Med 2019;2 (1):1–8.
• [76] Liu Q, Cai J, Fan SZ, Abbod MF, Shieh JS, et al. Spectrum analysis of EEG signals using CNN to model patient’s consciousness level based on anesthesiologists’ experience. IEEE Access 2019;7:53731–42.