Mostefa Mesbah

Time-frequency feature extraction of newborn EEG seizure using SVD-based techniques

The nonstationary and multicomponent nature of newborn EEG seizures tends to increase the complexity of the seizure detection problem. In dealing with this type of problems, time-frequency-based techniques were shown to outperform classical techniques. This paper presents a new time-frequency-based EEG seizure detection technique. The technique uses an estimate of the distribution function of the singular vectors associated with the time-frequency distribution of an EEG epoch to characterise the patterns embedded in the signal. The estimated distribution functions related to seizure and nonseizure epochs were used to train a neural network to discriminate between seizure and nonseizure patterns.

Signal enhancement by time-frequency peak filtering

Time-frequency peak filtering (TFPF) allows the reconstruction of signals from observations corrupted by additive noise by encoding the noisy signal as the instantaneous frequency (IF) of a frequency modulated (FM) analytic signal. IF estimation is then performed on the analytic signal using the peak of a time-frequency distribution (TFD) to recover the filtered signal. This method is biased when the peak of the Wigner-Ville distribution (WVD) is used to estimate the encoded signals instantaneous frequency. We characterize a class of signals for which the method implemented using the pseudo WVD is approximately unbiased. This class contains deterministic bandlimited nonstationary multicomponent signals in additive white Gaussian noise (WGN). We then derive the pseudo WVD window length that gives a reduced bias when TFPF is used for signals from this class. Testing of the method on both synthetic and real life newborn electroencephalogram (EEG) signals shows clean recovery of the signals in noise level down to a signal-to-noise ratio (SNR) of -9 dB.

A Nonstationary Model of Newborn EEG

The detection of seizure in the newborn is a critical aspect of neurological research. Current automatic detection techniques are difficult to assess due to the problems associated with acquiring and labelling newborn electroencephalogram (EEG) data. A realistic model for newborn EEG would allow confident development, assessment and comparison of these detection techniques. This paper presents a model for newborn EEG that accounts for its self-similar and nonstationary nature. The model consists of background and seizure submodels. The newborn EEG background model is based on the short-time power spectrum with a time-varying power law. The relationship between the fractal dimension and the power law of a power spectrum is utilized for accurate estimation of the short-time power law exponent. The newborn EEG seizure model is based on a well-known time-frequency signal model. This model addresses all significant time-frequency characteristics of newborn EEG seizure which include; multiple components or harmonics, piecewise linear instantaneous frequency laws and harmonic amplitude modulation. Estimates of the parameters of both models are shown to be random and are modelled using the data from a total of 500 background epochs and 204 seizure epochs. The newborn EEG background and seizure models are validated against real newborn EEG data using the correlation coefficient. The results show that the output of the proposed models have a higher correlation with real newborn EEG than currently accepted models (a 10% and 38% improvement for background and seizure models, respectively)

A time-frequency approach for newborn seizure detection

Techniques previously designed for seizure detection in newborns using the electroencephalogram (EEG) have been relatively inefficient due to their assumption of local stationarity of the EEG. To overcome the problem raised by the nonstationarity of the EEG signal, current methods are extended to a time-frequency approach. This allows the analysis and characterization of the different newborn EEG patterns that are intended to be the first step toward an automatic time-frequency seizure detection and classification. An in-depth analysis of both the autocorrelation and spectrum seizure detection techniques identified the detection criteria that can be extended to the time-frequency domain. The selected method uses a high-resolution reduced interference time-frequency distribution referred to as the B-distribution (BD). Here, the authors present the various patterns of observed time-frequency seizure signals and relate them to current knowledge of seizures. In particular, initial results indicate that a quasilinear instantaneous frequency (IF) can be used as a critical feature of the EEG seizure characteristics.

Time-frequency based newborn EEG seizure detection using low and high frequency signatures

The nonstationary and multicomponent nature of newborn EEG seizures tend to increase the complexity of the seizure detection problem. In dealing with this type of problem, time-frequency based techniques were shown to outperform classical techniques. Neonatal EEG seizures have signatures in both low frequency (lower than 10 Hz) and high frequency (higher than 70 Hz) areas. Seizure detection techniques have been proposed that concentrate on either low frequency or high frequency signatures of seizures. They, however, tend to miss seizures that reveal themselves only in one of the frequency areas. To overcome this problem, we propose a detection method that uses time-frequency seizure features extracted from both low and high frequency areas. Results of applying the proposed method on five newborn EEGs are very encouraging.

Newborn Seizure Detection Based on Heart Rate Variability

In this paper, we investigate the use of heart rate variability (HRV) for automatic newborn seizure detection. The proposed method consists of a sequence of processing steps, namely, obtaining HRV from the ECG, extracting a discriminating HRV feature set, selecting an optimal subset from the full feature set, and, finally, classifying the HRV into seizure/nonseizure using a supervised statistical classifier. Due to the fact that HRV signals are nonstationary, a set of time-frequency features from the newborn HRV is proposed and extracted. In order to achieve efficient HRV-based automatic newborn seizure detection, a two-phase wrapper-based feature selection technique is used to select the feature subset with minimum redundancy and maximum class discriminability. Tested on ECG recordings obtained from eight newborns with identified EEG seizure, the proposed HRV-based neonatal seizure detection algorithm achieved 85.7% sensitivity and 84.6% specificity. These results suggest that the HRV is sensitive to changes in the cardioregulatory system induced by the seizure, and therefore, can be used as a basis for an automatic seizure detection.

A sampling limit for the empirical mode decomposition

The aim of this paper is to investigate the effect of sampling on the empirical mode decomposition (EMD). To this end, an experiment utilising linear frequency modulated (LFM) signals was used to simulate different sampling rates. This experiment showed that as the frequency content of the signal (fc) approached the sampling frequency (fs) the EMD performed poorly due to poor amplitude resolution. This led to a definition of a sampling limit that was 5 times the Nyqvist rate (fs/10) to improve the performance of the EMD. Comparative simulation with this sampling limit was conducted on a simulated and a real world signal. The results exhibited significant improvement in intrinsic mode function (IMF) orthogonality, the distribution of IMF energy and IMF coherence.

A matching pursuit-based signal complexity measure for the analysis of newborn EEG

This paper presents a new relative measure of signal complexity, referred to here as relative structural complexity (RSC), which is based on the matching pursuit (MP) decomposition. By relative, we refer to the fact that this new measure is highly dependent on the decomposition dictionary used by MP. The structural part of the definition points to the fact that this new measure is related to the structure, or composition, of the signal under analysis. After a formal definition, the proposed RSC measure is used in the analysis of newborn electroencephalogram (EEG). To do this, firstly, a time–frequency decomposition dictionary is specifically designed to compactly represent the newborn EEG seizure state using MP. We then show, through the analysis of synthetic and real newborn EEG data, that the relative structural complexity measure can indicate changes in EEG structure as it transitions between the two EEG states; namely seizure and background (non-seizure).

Neonatal EEG seizure detection using low and high frequency signatures

The nonstationary and multicomponent nature of newborn EEG seizures tend to increase the complexity of the seizure detection problem. In dealing with this type of problem, time-frequency based techniques were shown to outperform classical techniques. Neonatal EEG seizures have signatures in both low frequency (lower than 10 Hz) and high frequency (higher than 70 Hz) areas. Seizure detection techniques have been proposed that concentrate on either low frequency or high frequency signatures of seizures. They, however, tend to miss seizures that reveal themselves only in one of the frequency areas. To overcome this problem, we propose a detection method that uses time-frequency seizure features extracted from both low and high frequency areas. Results of applying the proposed method on five newborn EEGs are very encouraging.

Detection of seizures in newborns using time-frequency analysis of EEG signals

This paper presents a time-frequency approach for electroencephalographic (EEG) seizure detection. The proposed method uses the high-resolution reduced interference B time-frequency distribution. An in-depth analysis of the seizure detection techniques of Gotman (frequency domain) (see Gotman et al., Electroenc. and Clin. Neurophysy., 103, pp.363-9, 1997) and Liu (time domain) (see Liu et al., Electroenc. and Clin. Neurophysy., 82, pp.30-7, 1992) has been performed in order to compare with the detection criteria used in the time-frequency domain. Both synthetic and real neonatal EEG signals have been used for testing.