Automated Real-Time Epileptic Seizure Detection in Scalp EEG Recordings Using an Algorithm Based on Wavelet Packet Transform

Abstract

A novel wavelet-based algorithm for real-time detection of epileptic seizures using scalp EEG is proposed. In a moving-window analysis, the EEG from each channel is decomposed by wavelet packet transform. Using wavelet coefficients from seizure and nonseizure references, a patient-specific measure is developed to quantify the separation between seizure and nonseizure states for the frequency range of 1-30 Hz. Utilizing this measure, a frequency band representing the maximum separation between the two states is determined and employed to develop a normalized index, called combined seizure index (CSI). CSI is derived for each epoch of every EEG channel based on both rhythmicity and relative energy of that epoch as well as consistency among different channels. Increasing significantly during ictal states, CSI is inspected using one-sided cumulative sum test to generate proper channel alarms. Analyzing alarms from all channels, a seizure alarm is finally generated. The algorithm was tested on scalp EEG recordings from 14 patients, totaling approximately 75.8 h with 63 seizures. Results revealed a high sensitivity of 90.5%, a false detection rate of 0.51 h(-1) and a median detection delay of 7 s. The algorithm could also lateralize the focus side for patients with temporal lobe epilepsy.

EEG seizure detection and prediction algorithms: a survey

“…Zandi et al presented a wavelet packet real-time seizure detection algorithm working on scalp EEG signals [46]. They developed a patient-specific metric to differentiate between seizure and non-seizure states in the 1-to 30-Hz frequency range based on wavelet coefficients of seizure and non-seizure references.…”

Classification of Pre-Clinical Seizure States Using Scalp EEG Cross-Frequency Coupling Features

“…Since the late 1990's, state of the art systems focus on patient specificity: they are able to achieve a sensitivity of 100%, with false detection rates around 0.02 per hour [13]. Patient specific algorithms have achieved these kinds of results using various wavelet-based, non-linear, and spectral features in combination with support vector machines (SVM) [14], recurrent neural networks [15], and logic based algorithms [16]. However, it remains a challenge to detect seizures using features that can generalize across patient datasets, and still provide a low rate of false alarms.…”

“…Early detection approaches are often capable of detecting seizures within several seconds of onset; published examples of detection latencies include 10s [18], 9.3s [13], 8s [14], and 7s [16]. Others have detections prior to the seizure onset, for example 4-10s [19].…”

A novel motor imagery EEG recognition method based on deep learning

“…WPT method can not only decompose the low-frequency and the high-frequency portion of data, but can also select the corresponding bands according to the signal characteristic, it is a more elaborate decomposition method than the wavelet. WPT retains the effective time-frequency information by filtering and reconstructing MI-EEG to ensure valid signal will not be lost [7]. In this paper, the WPT method was adopted to decompose MI-EEG by time-frequeny, and then frequency band of Mu rhythm and Bate rhythm were selected to be …”