Premium
Entropy of Electroencephalogram (EEG) Signals Varies with Sleep State
Author(s) -
Mathew Blesy,
Hudgel David,
Bruce Margaret,
Bruce Eugene
Publication year - 2006
Publication title -
the faseb journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.709
H-Index - 277
eISSN - 1530-6860
pISSN - 0892-6638
DOI - 10.1096/fasebj.20.4.a412-b
Subject(s) - sample entropy , non rapid eye movement sleep , electroencephalography , audiology , approximate entropy , entropy (arrow of time) , mathematics , statistics , standard deviation , sleep stages , psychology , eye movement , polysomnography , pattern recognition (psychology) , medicine , physics , neuroscience , time series , thermodynamics , cognitive psychology
To test our hypothesis that entropy of EEG signals changes systematically with sleep state, we analyzed the C3A2 EEG and an electrooculogram (EOG) recorded from 11 normal adult subjects awake (W) and in non rapid eye movement (NREM) and rapid eye movement (REM) sleep. Signals were sampled at 50Hz and low pass filtered at 25Hz. The EEG was filtered to remove the EOG contamination. Sample Entropy [SampEn (m, r, N)] was calculated for m=2, r = 0.2 x (standard deviation of the signal), and N=1500, (corresponding to 30s of data). Entropy was largest in W; therefore, for each subject entropy values were normalized to the average W value for that subject. Stage3/4 had the smallest entropy (0.61±0.14), followed by stage 2 (0.75±0.12), REM (0.81±0.6) and stage 1 (0.89±0.05). This pattern of entropy change with sleep state was highly consistent in all 11 subjects. Analysis of variance with p<0.05 and a Tukey post hoc test demonstrated that entropy of Stage 3/4 was significantly less than that of all other stages. In addition, Stage 2 and REM each differed from W. To determine if these findings were specific to the C3A2 leads, SampEn was also calculated using O1A2 leads from 4 of these subjects. The two leads showed no differences in entropy (2‐sample t‐test, p>0.4). Hence sample entropy of C3A2 leads (and likely of O1A2) changes systematically with sleep state. We propose that Sample Entropy may be a useful index as part of a montage for assessing sleep state. Supported by the Kentucky Science and Engineering Foundation.