Medical Engineering & Physics 24 (2002) 349–360
www.elsevier.com/locate/medengphy
Automatic recognition of alertness and drowsiness from EEG by
an artificial neural network
Aleksandra Vuckovic
a,
, Vlada Radivojevic
b
, Andrew C.N. Chen
a
, Dejan Popovic
a
a
Center for Sensory-Motor Interaction, Aalborg University, Aalborg, Denmark
b
Institute for the Mental Health, Belgrade, Yugoslavia
Received 19 April 2001; received in revised form 27 February 2002; accepted 14 March 2002
Abstract
We present a novel method for classifying alert vs drowsy states from 1 s long sequences of full spectrum EEG recordings in
an arbitrary subject. This novel method uses time series of interhemispheric and intrahemispheric cross spectral densities of full
spectrum EEG as the input to an artificial neural network (ANN) with two discrete outputs: drowsy and alert. The experimental
data were collected from 17 subjects. Two experts in EEG interpretation visually inspected the data and provided the necessary
expertise for the training of an ANN. We selected the following three ANNs as potential candidates: (1) the linear network with
Widrow-Hoff (WH) algorithm; (2) the non-linear ANN with the Levenberg–Marquardt (LM) rule; and (3) the Learning Vector
Quantization (LVQ) neural network. We showed that the LVQ neural network gives the best classification compared with the linear
network that uses WH algorithm (the worst), and the non-linear network trained with the LM rule. Classification properties of LVQ
were validated using the data recorded in 12 healthy volunteer subjects, yet whose EEG recordings have not been used for the
training of the ANN. The statistics were used as a measure of potential applicability of the LVQ: the t-distribution showed that
matching between the human assessment and the network output was 94.37 ± 1.95%. This result suggests that the automatic recog-
nition algorithm is applicable for distinguishing between alert and drowsy state in recordings that have not been used for the
training. © 2002 IPEM. Published by Elsevier Science Ltd. All rights reserved.
Keywords: Alert; Drowsy; EEG; Time series; Cross-spectral density; Neural networks
1. Introduction
The goals of this study were as follows: (1) estab-
lishing a method of processing input data from a full
spectrum of EEG recordings and (2) selecting an arti-
ficial neural network (ANN) that can distinguish
between alert and drowsy states in an arbitrary subject
by the use of processed EEG signals.
Spontaneous electrical brain activities, i.e. EEG sig-
nals, are dynamic, stochastic, non-linear and non-station-
ary [1–4]. The EEG recordings depend on the location
of the electrodes, their impedance and the state of alert-
ness. In addition, the EEG recordings vary substantially
between healthy subjects. Extensive expertise is required
to visually interpret the EEG recordings in order to iso-
Corresponding author. Tel.: +45-96-35-87-58; fax: +45-98-15-
40-08.
E-mail address: av@smi.auc.dk (A. Vuckovic).
1350-4533/02/$22.00 © 2002 IPEM. Published by Elsevier Science Ltd. All rights reserved.
PII:S1350-4533(02)00030-9
late and identify characteristic information from a large
amount of data. A computerized analysis of the EEG
recordings aims to facilitate the time-consuming and dif-
ficult visual inspection [5] and automatically extract
characteristic features of brain activity.
A computer-assisted EEG classification of drowsiness
has been analysed in several studies [6–15]. The classi-
fication was based on a spectral analysis of EEG rec-
ordings [6,8,9,10] and showed that a limited number of
electrodes and spectral analysis of characteristic bands
could be used as a classifier. More recently, some studies
[9,16] concentrated on detecting the information on
drowsiness available from a full EEG spectrum. Principe
et al. [12] designed a finite automaton that was capable
of categorizing the sleep into seven different stages.
McKeown et al. [17] used statistical methods for the
analysis of EEG signals and detection of vigilance
changes.
Pradhan [18] presented preliminary results for a
classification of seizure activities by applying an ANN