Nicoletta Nicolaou

Detection of epileptic electroencephalogram based on Permutation Entropy and Support Vector Machines

The electroencephalogram (EEG) has proven a valuable tool in the study and detection of epilepsy. This paper investigates for the first time the use of Permutation Entropy (PE) as a feature for automated epileptic seizure detection. A Support Vector Machine (SVM) is used to classify segments of normal and epileptic EEG based on PE values. The proposed system utilizes the fact that the EEG during epileptic seizures is characterized by lower PE than normal EEG. It is shown that average sensitivity of 94.38% and average specificity of 93.23% is obtained by using PE as a feature to characterize epileptic and seizure-free EEG, while 100% sensitivity and specificity were also obtained in single-trial classifications.

The use of permutation entropy to characterize sleep electroencephalograms.

This work proposes the use of Permutation Entropy (PE), a measure of time-series complexity, to characterize electroencephalogram (EEG) signals recorded during sleep. Such a measure could provide information concerning the different sleep stages and, thus, be utilized as an additional aid to obtain sleep staging information. PE has been estimated for artifact-free 30s segments from more than 80 hours of EEG records obtained from 16 subjects during all-night recordings, from which the mean PE for each sleep stage was obtained. It was found that different sleep stages are characterized by significantly different PE values, which track the physiological changes in the complexity of the EEG signals observed at the different sleep stages. This finding encourages the use of PE as an additional aide to either visual or automated sleep staging.

EEG-Based Automatic Classification of ‘Awake’ versus ‘Anesthetized’ State in General Anesthesia Using Granger Causality

Background General anesthesia is a reversible state of unconsciousness and depression of reflexes to afferent stimuli induced by administration of a “cocktail” of chemical agents. The multi-component nature of general anesthesia complicates the identification of the precise mechanisms by which anesthetics disrupt consciousness. Devices that monitor the depth of anesthesia are an important aide for the anesthetist. This paper investigates the use of effective connectivity measures from human electrical brain activity as a means of discriminating between ‘awake’ and ‘anesthetized’ state during induction and recovery of consciousness under general anesthesia. Methodology/Principal Findings Granger Causality (GC), a linear measure of effective connectivity, is utilized in automated classification of ‘awake’ versus ‘anesthetized’ state using Linear Discriminant Analysis and Support Vector Machines (with linear and non-linear kernel). Based on our investigations, the most characteristic change of GC observed between the two states is the sharp increase of GC from frontal to posterior regions when the subject was anesthetized, and reversal at recovery of consciousness. Features derived from the GC estimates resulted in classification of ‘awake’ and ‘anesthetized’ states in 21 patients with maximum average accuracies of 0.98 and 0.95, during loss and recovery of consciousness respectively. The differences in linear and non-linear classification are not statistically significant, implying that GC features are linearly separable, eliminating the need for a complex and computationally expensive non-linear classifier. In addition, the observed GC patterns are particularly interesting in terms of a physiological interpretation of the disruption of consciousness by anesthetics. Bidirectional interaction or strong unidirectional interaction in the presence of a common input as captured by GC are most likely related to mechanisms of information flow in cortical circuits. Conclusions/Significance GC-based features could be utilized effectively in a device for monitoring depth of anesthesia during surgery.

Automated Artifact Removal From the Electroencephalogram A Comparative Study

Contamination of the electroencephalogram (EEG) by artifacts greatly reduces the quality of the recorded signals. There is a need for automated artifact removal methods. However, such methods are rarely evaluated against one another via rigorous criteria, with results often presented based upon visual inspection alone. This work presents a comparative study of automatic methods for removing blink, electrocardiographic, and electromyographic artifacts from the EEG. Three methods are considered; wavelet, blind source separation (BSS), and multivariate singular spectrum analysis (MSSA)-based correction. These are applied to data sets containing mixtures of artifacts. Metrics are devised to measure the performance of each method. The BSS method is seen to be the best approach for artifacts of high signal to noise ratio (SNR). By contrast, MSSA performs well at low SNRs but at the expense of a large number of false positive corrections.

Computationally efficient classification of human transport mode using micro-doppler signatures

In this paper we present a micro-Doppler (mD) system and a computationally efficient classifier for the purpose of distinguishing different means of transport for human beings (pedestrians, inline skaters and cyclists) based on their mD time-frequency signatures. Accuracies as high as 97% are obtained while keeping the overall computational cost low.

Neural Network–Based Classification of Anesthesia/Awareness Using Granger Causality Features

This article investigates the signal processing part of a future system for monitoring awareness during surgery. The system uses features from the patients’ electrical brain activity (EEG) to discriminate between “anesthesia” and “awareness.” We investigate the use of a neural network classifier and Granger causality (GC) features for this purpose. GC captures anesthetic-induced changes in the causal relationships between pairs of signals from different brain areas. The differences in the pairwise causality estimated from the EEG activity are used as features for subsequent classification between “awake” and “anesthetized” states. EEG data from 31 subjects obtained during surgery and maintenance of anesthesia with propofol, sevoflurane, or desflurane, are classified using a neural network with one layer of hidden units. An average accuracy of 96% is obtained.

Entropy measures for discrimination of ‘awake’ Vs ‘anaesthetized’ state in recovery from general anesthesia

Approximate Entropy (ApEn) and Permutation Entropy (PE) have been recently introduced for assessment of anesthetic depth. Both measures have previously been shown to track changes in the electrical brain activity related to the administration of anesthetic agents. In this paper ApEn and PE are compared for the automatic classification of ‘awake’ and ‘anesthetized’ state using a Support Vector Machine to assess their robustness for potential use in a device for monitoring awareness during general anesthesia. It was found that both measures provide linearly separable features and we are able to discriminate between the two states with accuracy greater than 96% using either of the two entropy measures.

A hardware-efficient lowpass filter design for biomedical applications

A hardware-efficient lowpass filter design technique based on an exponentially weighted moving average (EWMA) filter architecture is proposed for the detection of general action potentials and nerve spikes in noisy signals. The EWMA VLSI architecture is compared with a basic moving average (MA) architecture and it is found that the EWMA technique is the most economical in terms of space of the two. In addition, a rule of thumb is given for converting a MA filter to the proposed filter. In the comparison, it was found that an EWMA filter is almost 85% more hardware-efficient than an MA filter.

Permutation Entropy: A new feature for Brain-Computer Interfaces

This paper investigates the use of Permutation Entropy (PE) as a feature for mental task classification for a Brain-Computer Interface system. PE is a recently introduced measure which quantifies signal complexity by measuring the departure of a time series from a random one. More regular signals are characterized by lower PE values. Here, PE is utilized to characterize signals from electroencephalograms of 3 subjects performing 4 motor imagery tasks, which are then classified using a Support Vector Machine. Even though it is possible to obtain 100% single-trial classification accuracy, this is very much subject-dependent.

Functional neuroimaging using UWB impulse radar: A feasibility study

Microwave imaging is a promising new modality for studying brain function. In the current paper we assess the feasibility of using a single chip implementation of an ultra-wideband impulse radar for developing a portable and low-cost functional neuroimaging device. A numerical model is used to predict the level of attenuation that will occur when detecting a volume of blood in the cerebral cortex. A phantom liquid is made, to study the radars performance at different attenuation levels. Although the radar is currently capable of detecting a point reflector in a phantom liquid with submillimeter accuracy and high temporal resolution, object detection at the desired level of attenuation remains a challenge.