An Automated System for Grading EEG Abnormality in Term Neonates with Hypoxic-Ischaemic Encephalopathy

Abstract

Automated analysis of the neonatal EEG has the potential to assist clinical decision making for neonates with hypoxic-ischaemic encephalopathy. This paper proposes a method of automatically grading the degree of abnormality in an hour long epoch of neonatal EEG. The automated grading system (AGS) was based on a multi-class linear classifier grading of short-term epochs of EEG which were converted into a long-term grading of EEG using a majority vote operation. The features used in the AGS were summary measurements of two sub-signals extracted from a quadratic time-frequency distribution: the amplitude modulation and instantaneous frequency. These sub-signals were based on a model of EEG as a multiplication of a coloured random process with a slowly varying pseudo-periodic waveform and may be related to macroscopic neurophysiological function. The 4 grade AGS had a classification accuracy of 83% compared to human annotation of the EEG (level of agreement, κ = 0.76). Features estimated on the developed sub-signals proved more effective at grading the EEG than measures based solely on the EEG and the incorporation of additional sub-grades based on EEG states into the AGS also improved performance.

Suitability of an inter-burst detection method for grading hypoxic-ischemic encephalopathy in newborn EEG

“…If the experts disagreed, they reviewed the EEG together and a consensus was reached to determine the final grade. This same dataset has been used by Stevenson et al [10] and Ahmed et al [11].…”

“…An integral part of grading the EEG for HIE injury, is to assess the level of discontinuity in the EEG by visually quantifying inter-burst activity [4], [9]. Automated grading methods developed by Stevenson et al [10] and Ahmed et al [11] which use multiple complex time-domain, frequencydomain, and information theory features. These studies do not consider detailed classification based on inter-burst interval analysis, which is an important component of visual grading of the EEG.…”

Computer assisted interpretation of the human EEG

“…One thing that is for certain is that the presence of quantitative EEG analysis will continue to grow during reviews. Many types of quantitative EEG features have been proposed to describe specific properties in the EEG, including statistical measures such as variance and kurtosis (Scherg et al, 2012;Stevenson et al, 2013), non-linear energy operators (Mukhopadhyay and Ray, 1998), smallworld networks and functional connectivity (Stam et al, 2007;Bullmore and Sporns, 2009), synchrony (Lachaux et al, 1999;van Pu en, 2003), entropy (Stam, 2005;Kannathal et al, 2005), power ratios (Kurtz et al, 2009;Cloostermans et al, 2011), bispectral index (Sigl and Chamoun, 1994), and le -right symmetry . Apart from epilepsy, clinical applications for quantitative EEG also include ICU monitoring (Friedman and Hirsch, 2010;Cloostermans et al, 2011;Foreman and Claassen, 2012), clinical psychiatry (Coburn et al, 2006;Hammond, 2010) and the diagnosis of neurodegenerative diseases (Petit et al, 2004;Babiloni et al, 2011;More i et al, 2012).…”

“…These include statistical measures such as variance, kurtosis and skewness (Scherg et al, 2012;Stevenson et al, 2013), non-linear energy operators (Mukhopadhyay and Ray, 1998), small-world networks and functional connectivity (Stam et al, 2007;Bullmore and Sporns, 2009), synchrony (Lachaux et al, 1999;van Pu en, 2003), entropy (Stam, 2005;Kannathal et al, 2005), power ratios (Kurtz et al, 2009;Cloostermans et al, 2011), bi-spectral index (Sigl and Chamoun, 1994), and le -right symmetry . Despite the variety of complex features available, relatively simple measures can be used to describe many of the background properties of an EEG.…”

Design and optimisation of time-frequency analysis for multichannel neonatal EEG background features in term neonates with hypoxic ischaemic encephalopathy: characterisation, classification and neurodevelopmental outcome prediction

“…However, EEG is a non-stationary signal, i. show outstanding performance in detecting newborn seizures [29] and automatic grading of EEG background patterns [155]. We propose that JTF-based dynamic features may also be useful in predicting outcome.…”