In this section
@inproceedings{Nakamura:2017:10.1109/IJCNN.2017.7966411,
author = {Nakamura, T and adjei, T and alqurashi, Y and looney, D and Morrell, M and Mandic, D},
doi = {10.1109/IJCNN.2017.7966411},
pages = {4387--4394},
publisher = {IEEE},
title = {Complexity science for sleep stage classification from EEG},
url = {http://dx.doi.org/10.1109/IJCNN.2017.7966411},
year = {2017}
}
TY - CPAPER
AB - Automatic sleep stage classification is an importantparadigm in computational intelligence and promises consider-able advantages to the health care. Most current automatedmethods require the multiple electroencephalogram (EEG) chan-nels and typically cannot distinguish the S1 sleep stage fromEEG. The aim of this study is to revisit automatic sleep stageclassification from EEGs using complexity science methods. Theproposed method applies fuzzy entropy and permutation entropyas kernels of multi-scale entropy analysis. To account for sleeptransition, the preceding and following 30 seconds of epoch datawere used for analysis as well as the current epoch. Combiningthe entropy and spectral edge frequency features extracted fromone EEG channel, a multi-class support vector machine (SVM)was able to classify 93.8% of 5 sleep stages for the SleepEDFdatabase [expanded], with the sensitivity of S1 stage was 49.1%.Also, the Kappa’s coefficient yielded 0.90, which indicates almostperfect agreement.
AU - Nakamura,T
AU - adjei,T
AU - alqurashi,Y
AU - looney,D
AU - Morrell,M
AU - Mandic,D
DO - 10.1109/IJCNN.2017.7966411
EP - 4394
PB - IEEE
PY - 2017///
SN - 2161-4407
SP - 4387
TI - Complexity science for sleep stage classification from EEG
UR - http://dx.doi.org/10.1109/IJCNN.2017.7966411
UR - http://hdl.handle.net/10044/1/45327
ER -