نوع مقاله : مقاله کامل پژوهشی
نویسندگان
1 دانشجوی دکترا، گروه مهندسی پزشکی، دانشکده برق، دانشگاه تهران
2 استادیار، گروه مهندسی پزشکی، دانشکده فنی، دانشگاه شاهد
چکیده
سیگنالهای زیستی مختلف شامل EEG، EOGو EMGبه منظور تشخیص اختلالات خواب در آزمایشگاههای خواب ثبت میشوند. تحلیل اطلاعات ثبت شده در زمان خواب بهوسیله متخصص خواب، به صورت شهودی انجام میشود. طبقهبندی شهودی مراحل خواب به دلیل طولانی بودن ثبتها، کار زمانبر و خسته کنندهای است. تحلیل خودکار خواب میتواند این امر را تسهیل کند. مهمترین گام برای طبقهبندی خودکار مراحل خواب، استخراج ویژگیهای مناسب است. در این تحقیق دو دسته ویژگی از سیگنال EEGاستخراج شدند: دسته اول ویژگیهایی هستند که از روی ضرایب تبدیل بستههای موجک (WPT) محاسبه شدهاند و دسته دوم شامل تعدادی از ویژگیهای فرکانسی و یک ویژگی زمانی یعنی دامنه سیگنال EEGهستند. در ادامه این دو مجموعه از ویژگیها به طور مجزا بهوسیله شبکههای عصبی SOMبه فضای دوبعدی نگاشته شدند. نگاشت بهدست آمده نشان داد که این ویژگیها در جدا کردن خودکار مراحل خواب بسیار مفیدند. اطلاعات استخراج شده از EEGبیداری و خواب عمیق به دو ناحیه کاملاً مجزا نگاشته شدند. این نگاشت همچنین نشان داد که سیگنالEEGبهتنهایی برای جدا کردن کامل مراحل خواب کافی نیست زیرا وقتی اطلاعات مستخرج از سیگنال EEGدر خواب REMو مرحله 1 از خواب NREMبه ناحیه یکسان نگاشت شدند، اطلاعات استخراج شده از سیگنال EEGدر مرحله 2 خواب با سایر مراحل همپوشانی دارد که این نتایج منطبق با تعاریف فیزیولوژی مراحل خواب است.
کلیدواژهها
موضوعات
عنوان مقاله [English]
Assessment of Time, Frequency, and Wavelet Packet Transform Features Extracted from EEG for Sleep Staging using Self Organizing Maps (SOM)
نویسندگان [English]
- Faride Ebrahimi 1
- Mohammad Mikaili 2
1 PhD Candidate, Bioelectric Group, School of Electrical Engineering, University of Tehran
2 Associate Professor, Biomedical Engineering Group, School of Engineering, Shahed University
چکیده [English]
Different biological signals including EEG, EOG, and EMG are recorded in sleep labs to diagnose sleep disorders. Data recorded during sleep is usually analyzed by sleep specialists visually. Since the sleep data is usually recorded for a long time period- namely a whole night- its visual inspection and classification is a very demanding and time consuming task so automatic analysis can definitely facilitate that. The key to automatic sleep staging is to extract suitable features. In the current study two classes of features are extracted from EEG signal. The first group is the features calculated from the coefficients of wavelet packet transformation (WPT) and the second group consists of a number of frequency features and a time feature, the amplitude of EEG signal itself. These two sets of features were separately mapped on a two dimensional space by SOM neural networks. The mappings indicated that these features are highly discriminative in separating sleep stages automatically. The data extracted from awake and deep sleep EEGs were mapped on two totally different regions. The mapping also indicated that EEG signal is not enough to separate stages thoroughly, as extracted data from EEG during REM and the first stage of NREM are mapped on the same region. Data extracted from EEG signals in the second stage overlapped with other stages which are in agreement with physiological definition of sleep stages.
کلیدواژهها [English]
- Sleep stages
- Feature Extraction
- Wavelet packet transformation (WPT)
- Power spectrum
- SOM neural network
[1] Rechtschaffen K.A. (Eds.), A Manual of standardized Terminology,Techniquesand Scoring System for Sleep Stages of Human Subjects; PublicHealthService,S. Government Printing Office, Washington, DC, 1968.
[2] Niedermeyer E., Lopes Da Silva F., Electro_ encephalography: basic principles, clinical applications and related fields; 3rd Ed; William & Wilkins; Part 2:Chapter 9-11, 1993.
[3] Shepard J.W. (Ed.), Atlas of Sleep Medicine. Mount Kisco, NY: Futura Publishing Co., 1991: 51-80.
[4] Thakor N., Tong S., Advances Quantitative Electroencephalogram Analysis Methods; Biomedical Engineering Department, 2004; 6: 453-495.
[5] Durka P., Malinowska U., Szelenberger W., Wakarow A., Blinowska K; High resolution parametric description of slow wave sleep; Neuroscience Methods, 2005; 147: 15-21.
[6] Acir N., Guzelis. C., Automatic recognition of sleep spindles in EEG by using artificial neural networks; Expert Systems with Applications, 2004; 27: 451-458.
[7] Ventouras E., Monoyiou E., Ktonas P., Paparrigopoulos T., Dikeos D., Uzunoglu N., Soldatos, C., Sleep spindle detection using artificial neural networkstrained with filtered time-domain EEG: A feasibility study; Computer Methods and Programs in Biomedicine, 2005; 78: 191-207.
[8] Acharya R., Faust O., Kannathal N., Chua T., Laxminarayan S; Non-Linear analysis of EEG signal at various sleep stages; Computer Methods and Programs in Biomedicine, 2005; 80: 37-45.
[9] Li J., Du Y,, Zhao L., Sleep Study with Wavelet time-frequency Analysis; IEEE, EMBS, 2005; 872-875.
[10] Oropesa E., Cycon H., Jobert M., Sleep Stage Classification using Wavelet Transform and Neural Network; International Computer Science Institute; 1999.
[11] Kiymik M., Akin M., Subasi A., Automatic recognition of alertness level by using wavelet transform and artificial neural network; Neuroscience Methods; 2004; 139: 231-240.
[12] Subasi A., Automatic recognition of alertness level from EEG by using neural network and wavelet coefficients; Expert Systems with Applications; 2005; 28: 701–711.
[13] Estrada E., Nazeran H., Nava P., Behbehani K., Burk J., Lucas E., EEG Feature extraction for classification of Sleep Stages; in Proc. IEEE, EMBS, 2004: 196-199.
[14] Van Hese P., Philips W., Koninck J., Van de Walle R., LemahieuI., Automatic DetectionofSleep Stages using the EEG; in Proc. IEEE, EMBS, 2001: 1944-1947.
[15] Barragan J., Estrada E., Nava P., Nazeran H., EEG-based Classification of Sleep Stages Using Artificial Neural Networks; Proceedings of the 27 international workshop on biomedical signal interpretation; 2005.
[16] Mogosso E., Provini. F., Montagna P., Ursino M., A wavelet based method for automatic detection of slow eye movements, Medical Engineering and physics, 2006: 860-875.
[17] Burrus C., Gopinath R., Guo H., Introduction to wavelets andwavelet transforms; Prentice Hall Pub, 1998: 10-40.
[18] Durka P., Malinowska. U., Szelenberger U., Wakarow A., Blinowska K., High resolution parametric description of slow wave sleep, J. Neuroscience Methods, 2005; 147: 15-21.
[19] Haykin S., Neural Networks, A Comprehensive Foundation; Prentice Hall International Editions, 1999: 352-472.
[20] Olbrich E., Achermann P., Meier P., Dynamics of human sleep EEG; Neuro computing; 2003; 52-54: 857-862.