A semi-automated method for epileptiform transient detection in the EEG of the fetal sheep using time-frequency analysis

Latent Phase Identification of High-Frequency Micro-Scale Gamma Spike Transients in the Hypoxic Ischemic EEG of Preterm Fetal Sheep Using Spectral Analysis and Fuzzy Classifiers

Premature babies are at high risk of serious neurodevelopmental disabilities, which in many cases are related to perinatal hypoxic–ischemic encephalopathy (HIE). Studies of neuroprotection in animal models consistently suggest that treatment must be started as early as possible in the first 6 h after hypoxia–ischemia (HI), the so-called latent phase before secondary deterioration, to improve outcomes. We have shown in preterm sheep that EEG biomarkers of injury, in the form of high-frequency micro-scale spike transients, develop and evolve in this critical latent phase after severe asphyxia. Real-time automatic identification of such events is important for the early and accurate detection of HI injury, so that the right treatment can be implemented at the right time. We have previously reported successful strategies for accurate identification of EEG patterns after HI. In this study, we report an alternative high-performance approach based on the fusion of spectral Fourier analysis and Type-I fuzzy classifiers (FFT-Type-I-FLC). We assessed its performance in over 2520 min of latent phase EEG recordings from seven asphyxiated in utero preterm fetal sheep exposed to a range of different occlusion periods. The FFT-Type-I-FLC classifier demonstrated 98.9 ± 1.0% accuracy for identification of high-frequency spike transients in the gamma frequency band (namely 80–120 Hz) post-HI. The spectral-based approach (FFT-Type-I-FLC classifier) has similar accuracy to our previous reverse biorthogonal wavelets rbio2.8 basis function and type-1 fuzzy classifier (rbio-WT-Type-1-FLC), providing competitive performance (within the margin of error: 0.89%), but it is computationally simpler and would be readily adapted to identify other potentially relevant EEG waveforms.

Electroencephalogram studies of hypoxic ischemia in fetal and neonatal animal models

Alongside clinical achievements, experiments conducted on animal models (including primate or non-primate) have been effective in the understanding of various pathophysiological aspects of perinatal hypoxic/ischemic encephalopathy (HIE). Due to the reasonably fair degree of flexibility with experiments, most of the research around HIE in the literature has been largely concerned with the neurodevelopmental outcome or how the frequency and duration of HI seizures could relate to the severity of perinatal brain injury, following HI insult. This survey concentrates on how EEG experimental studies using asphyxiated animal models (in rodents, piglets, sheep and non-human primate monkeys) provide a unique opportunity to examine from the exact time of HI event to help gain insights into HIE where human studies become difficult.

Examining the effect of MgSO4 on sharp wave transient activity in the hypoxic-ischemic fetal sheep model

Hypoxic-ischemic encephalopathy (HIE) due to lack of oxygen is a debilitating disorder experienced by a significant number of preterm infants during birth. Studies show that the brain undergoes different phases of injury following hypoxic insult, but the first 6-8 hours (known as a latent phase) are the key to treatment efficacy. Cerebral hypothermia is one known treatment, and for it to be effective it must be started during the latent phase and continued for several days. In order to determine the effectiveness of treatment it is important to pinpoint the time of insult. Monitoring of sharp wave transient activity in the hypoxic-ischemic (HI) electroencephalogram (EEG) could be a predictor for time of hypoxic insult. Due to practicality, it is optimal if this monitoring is performed automatically. Further, MgSO4 is a drug given to an increasing number of women in labor, due to its neuroprotective properties. This drug may influence transient activity in the HI fetal sheep EEG, leading to further complications in predicting hypoxic insult. This paper explores the effect of MgSO4 on sharp wave transient activity in the EEG of a HI fetal sheep. Demonstrated in this paper is the usage of a Wavelet-Type-II Fuzzy classifier to detect sharp wave transients during the latent phase of a control group fetal sheep and an MgSO4-treated fetal sheep. This detection was performed with an average overall performance of 93.21%±5.49 over 660 minutes of latent phase, post occlusion. There were no significant differences in number of sharp wave transients in the early- and mid-latent phases of injury for both fetal sheep. However, in the late-latent phase the MgSO4-treated fetal sheep had significantly fewer sharp wave transients than the control fetal sheep.

Using type-2 fuzzy logic systems for spike detection in the hypoxic ischemic EEG of the preterm fetal sheep

Perinatal hypoxia is a major cause of brain injury in preterm babies. Thus, neuro-protective treatments play a pivotal role during the first 6-8 hours post hypoxic-ischemic insult. However, at present it is not possible to determine which infants are suffering from hypoxic ischemia. Recent investigations suggest that there are high frequency micro-scale transients exist in the first 6-8 hours of a hypoxic ischemic EEG which could be utilized as the useful benchmarks for the prediction of hypoxia. Type-2 Fuzzy Logic Systems (Type-2 FLS) have the capability to handle inherent uncertainties in nonlinear signals. This paper describes the application of a Type-2 FLS to detect spikes in the preterm fetal sheep electroencephalogram (EEG) after asphyxia in utero. The Type-2 FLS differentiates each detected event in terms of its spikiness and specifies the potential events based on their degree of similarity to an EEG expert definition of a standard spike. An adaptive thresholding method has been employed in order to increase the spike detection ability of the purposed system. The sensitivity and selectivity verify enhanced performance of the Type-2 FLS for spike detection in fetal sheep EEG signals with a 98.1% and 93.7% respectively which are significantly improved in comparison to our previous methods.

Identifying stereotypic evolving micro-scale seizures (SEMS) in the hypoxic-ischemic EEG of the pre-term fetal sheep with a wavelet type-II fuzzy classifier

Perinatal hypoxic-ischemic encephalopathy (HIE) around the time of birth due to lack of oxygen can lead to debilitating neurological conditions such as epilepsy and cerebral palsy. Experimental data have shown that brain injury evolves over time, but during the first 6-8 hours after HIE the brain has recovered oxidative metabolism in a latent phase, and brain injury is reversible. Treatments such as therapeutic cerebral hypothermia (brain cooling) are effective when started during the latent phase, and continued for several days. Effectiveness of hypothermia is lost if started after the latent phase. Post occlusion monitoring of particular micro-scale transients in the hypoxic-ischemic (HI) Electroencephalogram (EEG), from an asphyxiated fetal sheep model in utero, could provide precursory evidence to identify potential biomarkers of injury when brain damage is still treatable. In our studies, we have reported how it is possible to automatically detect HI EEG transients in the form of spikes and sharp waves during the latent phase of the HI EEG of the preterm fetal sheep. This paper describes how to identify stereotypic evolving micro-scale seizures (SEMS) which have a relatively abrupt onset and termination in a frequency range of 1.8-3Hz (Delta waves) superimposed on a suppressed EEG amplitude background post occlusion. This research demonstrates how a Wavelet Type-II Fuzzy Logic System (WT-Type-II-FLS) can be used to automatically identify subtle abnormal SEMS that occur during the latent phase with a preliminary average validation overall performance of 78.71%±6.63 over the 390 minutes of the latent phase, post insult, using in utero pre-term hypoxic fetal sheep models.

Superiority of high frequency hypoxic ischemic EEG signals of fetal sheep for sharp wave detection using Wavelet-Type 2 Fuzzy classifiers

There is approximately a 6-8 hour window that exists from when a hypoxic-ischemic insult occurs, in utero, before significant irreversible brain injury occurs in new born infants. The focus of our work is to determine through the electroencephalogram (EEG) if such a hypoxic-ischemic insult has occurred such that neuro-protective treatment can be sought within this period. At present, there are no defined biomarkers in the EEG that are currently being used to help classify if a hypoxic ischemia insult has occurred. However, micro-scale transients in the form of spikes, sharps and slow waves exists that could provide precursory information whether a hypoxic-ischemic insult has occurred or not. In our previous studies we have successfully automatically identified spikes with high sensitivity and selectivity in the conventional 64Hz sampled EEG. This paper details the significant advantage that can be obtained in using high frequency 1024Hz sampled EEG for sharp wave detection over the typically employed 64Hz sampled EEG. This advantage is amplified when a combination of wavelet Type-2 Fuzzy Logic System (WT-Type-2-FLS) classifiers are used to identify the sharp wave transients. By applying WT-Type-2-FLS to the 1024Hz EEG record and to the same down-sampled 64Hz EEG record we demonstrate, how the sharp wave transients detection increases significantly for high resolution 1024Hz EEG over 64Hz EEG. The WT-Type-2-FLS algorithm performance was assessed over 3 standardised time periods within the first 8 hours, post occlusion of a fetal sheep, in utero. 1024Hz EEG results demonstrate the algorithm detected sharps with overall performance rates of 85%, 92%, and 87% in the Early/Mid and Late-latent phases of injury, respectively as compared to 25%, 55% and 31% in the 64Hz EEG. These results demonstrate the power of Wavelet Type-2 Fuzzy Logic System at detecting sharp waves in 1024Hz EEG and suggest that there should be a movement toward recording high frequency EEG for analysis of hypoxic ischemic micro-scale transients that does not occur at present.