EEG Oscillatory Power and Complexity for Epileptic Seizure Detection

Abstract

Monitoring patients at risk of epileptic seizure is critical for optimal treatment and ensuing the reduction of seizure risk and complications. In general, seizure detection is done manually in hospitals and involves time-consuming visual inspection and interpretation by experts of electroencephalography (EEG) recordings. The purpose of this study is to investigate the pertinence of band-limited spectral power and signal complexity in order to discriminate between seizure and seizure-free EEG brain activity. The signal complexity and spectral power are evaluated in five frequency intervals, namely, the delta, theta, alpha, beta, and gamma bands, to be used as EEG signal feature representation. Classification of seizure and seizure-free data was performed by prevalent potent classifiers. Substantial comparative performance evaluation experiments were performed on a large EEG data record of 341 patients in the Temple University Hospital EEG seizure database. Based on statistically validated criteria, results show the efficiency of band-limited spectral power and signal complexity when using random forest and gradient-boosting decision tree classifiers (95% of the area under the curve (AUC) and 91% for both F-measure and accuracy). These results support the use of these automatic classification schemes to assist the practicing neurologist interpret EEG records more accurately and without tedious visual inspection.

Supervised Machine Learning and Deep Learning Techniques for Epileptic Seizure Recognition Using EEG Signals—A Systematic Literature Review

“…Power spectral density (PSD) is a measure of the distribution of power of the discrete frequencies that compose the signal retrieved by applying Welch’s FFT method. In [ 13 , 14 , 20 , 35 , 52 ], PSD was used to analyze the frequency power in various brain states, which have discriminative ability between seizures and non-seizures. In [ 53 ], PSD and additional statistical features were applied after the signal was transformed into a time-frequency image (TFI).…”

“…Non-seizure activities had low PSD with a non-stationary pattern, but seizure activities had high PSD values at certain frequencies with spike-wave patterns, demonstrating the potential of PSD in detecting seizures. Other power spectral measures include absolute and relative spectral powers, which were used to investigate the phase shifts between interictal and preictal states of the brain [ 52 , 54 ].…”

“…If the entropy value is above a certain threshold, samples are not confidently identified by the current iteration of the proposed algorithm, and further training is required. It has also been applied to signal histogram data calculated via PSD to quantify the complexity of the signal [ 13 ] and was used as one of the features in [ 52 ]. Spectral entropy (SEN), commonly called power spectrum entropy, is calculated on the normalized PSD values using classical Shannon’s entropy to quantify the spectral complexity of the signal in the frequency domain [ 13 , 35 , 39 , 52 ].…”

“…It has also been applied to signal histogram data calculated via PSD to quantify the complexity of the signal [ 13 ] and was used as one of the features in [ 52 ]. Spectral entropy (SEN), commonly called power spectrum entropy, is calculated on the normalized PSD values using classical Shannon’s entropy to quantify the spectral complexity of the signal in the frequency domain [ 13 , 35 , 39 , 52 ]. In [ 48 ], it was observed that healthy patients had lower median spectral entropy than epileptic patients.…”

Intelligent Epileptic Seizure Detection and Classification Model Using Optimal Deep Canonical Sparse Autoencoder

“…At a molecular level, several pathways were implicated in the apoptosis of premyelinating oligodendrocytes or subplate neurons that exist in perinatal brain growth. A rising concentration of glutamate or free radical reactive species in hypoxic-ischemic encephalopathy, inflammatory cytokines such as TNF-α, IL-1b, IL-6, 12, 15, and 18 from activated microglia and astrocytes, low pH in infections, and free iron secondary to cerebral hemorrhage were broadly stated in both white- and grey-matter injuries as significant triggers for epileptic events [ 3 , 4 ]. An epileptic seizure is an ephemeral instance of complete or partial abnormal, unintended movement of the body that may be integrated with a loss of consciousness [ 5 , 6 ].…”