Bio-Signal Complexity Analysis in Epileptic Seizure Monitoring: A Topic Review

An Epileptic EEG Detection Method Based on Data Augmentation and Lightweight Neural Network

OBJECTIVE

Epilepsy, an enduring neurological disorder, afflicts approximately 65 million individuals globally, significantly impacting their physical and mental wellbeing. Traditional epilepsy detection methods are labor-intensive, leading to inefficiencies. Although deep learning techniques for brain signal detection have gained traction in recent years, their clinical application advancement is hindered by the significant requirement for high-quality data and computational resources during training.

METHODS & RESULTS

The neural network training initially involved merging two datasets of different data quality, namely Bonn University datasets and CHB-MIT datasets, to bolster its generalization capabilities. To tackle the issues of dataset size and class imbalance, we employed small window segmentation and Synthetic Minority Over-sampling Technique (SMOTE). algorithms to augment and equalize the data. A streamlined neural network architecture was then proposed, drastically reducing the model's training parameters. Notably, a model trained with a mere 9,371 parameters yielded impressive results. The three-classification task on the combined dataset delivered an accuracy of 98.52%, sensitivity of 97.99%, specificity of 99.35%, and precision of 98.44%.

CONCLUSION

The experimental findings of this study underscore the superiority of the proposed method over existing approaches in both model size reduction and accuracy enhancement. As a result, it is more apt for deployment in low-cost, low computational hardware devices, including wearable technology, and various clinical applications. Clinical and Translational Impact Statement- This study is a Pre-Clinical Research. The lightweight neural network is easily deployed on hardware device for real-time epileptic EEG detection.

Epileptic detection in single and multi-lead EEG signals using persistent homology based on bi-directional weighted visibility graphs

Epilepsy is a widespread neurological disorder, and its recurrence and suddenness are making automatic detection of seizure an urgent necessity. For this purpose, this paper performs topological data analysis (TDA) of electroencephalographic (EEG) signals by the medium of graphs to explore the potential brain activity information they contain. Through our innovative method, we first map the time series of epileptic EEGs into bi-directional weighted visibility graphs (BWVGs), which give more comprehensive reflections of the signals compared to previous existing structures. Traditional graph-theoretic measurements are generally partial and mainly consider differences or correlations in vertices or edges, whereas persistent homology (PH), the essential part of TDA, provides an alternative way of thinking by quantifying the topology structure of the graphs and analyzing the evolution of these topological properties with scale changes. Therefore, we analyze the PH for BWVGs and then obtain the two indicators of persistence and birth-death for homology groups to reflect the topology of the mapping graphs of EEG signals and reveal the discrepancies in brain dynamics. Furthermore, we adopt neural networks (NNs) for the automatic detection of epileptic signals and successfully achieve a classification accuracy of 99.67% when distinguishing among three different sets of EEG signals from seizure, seizure-free, and healthy subjects. In addition, to accommodate multi-leads, we propose a classifier that incorporates graph structure to distinguish seizure and seizure-free EEG signals. The classification accuracies of the two subjects used in the classifier are as high as 99.23% and 94.76%, respectively, indicating that our proposed model is useful for the analysis of EEG signals.

Epileptic Seizure Detection with Hybrid Time-Frequency EEG Input: A Deep Learning Approach

The precise detection of epileptic seizure helps to prevent the serious consequences of seizures. As the electroencephalogram (EEG) reflects the brain activity of patients effectively, it has been widely used in epileptic seizure detection in the past decades. Recently, deep learning-based detection methods which automatically learn features from the EEG signals have attracted much attention. However, with deep learning-based detection methods, different input formats of EEG signals will lead to different detection performances. In this paper, we propose a deep learning-based epileptic seizure detection method with hybrid input formats of EEG signals, i.e., original EEG, Fourier transform of EEG, short-time Fourier transform of EEG, and wavelet transform of EEG. Convolutional neural networks (CNNs) are designed for extracting latent features from these inputs. A feature fusion mechanism is applied to integrate the learned features to generate a more stable syncretic feature for seizure detection. The experimental results show that our proposed hybrid method is effective to improve the seizure detection performance in few-shot scenarios.

Brain injuries can set up an epileptogenic neuronal network

Development of epilepsy or epileptogenesis promotes recurrent seizures. As of today, there are no effective prophylactic therapies to prevent the onset of epilepsy. Contributing to this deficiency of preventive therapy is the lack of clarity in fundamental neurobiological mechanisms underlying epileptogenesis and lack of reliable biomarkers to identify patients at risk for developing epilepsy. This limits the development of prophylactic therapies in epilepsy. Here, neural network dysfunctions reflected by oscillopathies and microepileptiform activities, including neuronal hyperexcitability and hypersynchrony, drawn from both clinical and experimental epilepsy models, have been reviewed. This review suggests that epileptogenesis reflects a progressive and dynamic dysfunction of specific neuronal networks which recruit further interconnected groups of neurons, with this resultant pathological network mediating seizure occurrence, recurrence, and progression. In the future, combining spatial and temporal resolution of neuronal non-invasive recordings from patients at risk of developing epilepsy, together with analytics and computational tools, may contribute to determining whether the brain is undergoing epileptogenesis in asymptomatic patients following brain injury.

Artificial Intelligence in Epilepsy

Background

The study of seizure patterns in electroencephalography (EEG) requires several years of intensive training. In addition, inadequate training and human error may lead to misinterpretation and incorrect diagnosis. Artificial intelligence (AI)-based automated seizure detection systems hold an exciting potential to create paradigms for proper diagnosis and interpretation. AI holds the promise to transform healthcare into a system where machines and humans can work together to provide an accurate, timely diagnosis, and treatment to the patients.

Objective

This article presents a brief overview of research on the use of AI systems for pattern recognition in EEG for clinical diagnosis.

Material and Methods

The article begins with the need for understanding nonstationary signals such as EEG and simplifying their complexity for accurate pattern recognition in medical diagnosis. It also explains the core concepts of AI, machine learning (ML), and deep learning (DL) methods.

Results and Conclusions

In this present context of epilepsy diagnosis, AI may work in two ways; first by creating visual representations (e.g., color-coded paradigms), which allow persons with limited training to make a diagnosis. The second is by directly explaining a complete automated analysis, which of course requires more complex paradigms than the previous one. We also clarify that AI is not about replacing doctors and strongly emphasize the need for domain knowledge in building robust AI models that can work in real-time scenarios rendering good detection accuracy in a minimum amount of time.

Graph theory in paediatric epilepsy: A systematic review

Graph theoretical studies have been designed to investigate network topologies during life. Network science and graph theory methods may contribute to a better understanding of brain function, both normal and abnormal, throughout developmental stages. The degree to which childhood epilepsies exert a significant effect on brain network organisation and cognition remains unclear. The hypothesis suggests that the formation of abnormal networks associated with epileptogenesis early in life causes a disruption in normal brain network development and cognition, reflecting abnormalities in later life. Neurological diseases with onset during critical stages of brain maturation, including childhood epilepsy, may threaten this orderly neurodevelopmental process. According to the hypothesis that the formation of abnormal networks associated with epileptogenesis in early life causes a disruption in normal brain network development, it is then mandatory to perform a proper examination of children with new-onset epilepsy early in the disease course and a deep study of their brain network organisation over time. In regards, graph theoretical analysis could add more information. In order to facilitate further development of graph theory in childhood, we performed a systematic review to describe its application in functional dynamic connectivity using electroencephalographic (EEG) analysis, focussing on paediatric epilepsy.

Variability analysis of epileptic EEG using the maximal overlap discrete wavelet transform

Purpose

To determine if there is a difference in the wavelet variances of seizure and non-seizure channels in the EEG of an epileptic subject.

Methods

A six-level decomposition was applied using the Maximal Overlap Discrete Wavelet Transform (MODWT). The wavelet variance and 95% CIs were calculated for each level of the decomposition. The number of changes in variance for each level were found using a change-point detection method of Whitcher. The Kruskal-Wallis test was used to determine if there were differences in the median number of change points within channels and across frequency bands (levels).

Results

No distinctive pattern was found for the wavelet variances to differentiate the seizure and non-seizure channels. The seizure channels tended to have lower variances for each level and overall, but this pattern only held for one of the three seizure channels (RAST4). The median number of change points did not differ between the seizure and non-seizure channels either within each channel or across the frequency bands.

Conclusion

The use of the MODWT in examining the variances and changes in variance did not show specific patterns which differentiate between seizure and non-seizure channels.

Redundancy Removed Dual-Tree Discrete Wavelet Transform to Construct Compact Representations for Automated Seizure Detection

With the development of pervasive sensing and machine learning technologies, automated epileptic seizure detection based on electroencephalogram (EEG) signals has provided tremendous support for the lives of epileptic patients. Discrete wavelet transform (DWT) is an effective method for time-frequency analysis of EEG and has been used for seizure detection in daily healthcare monitoring systems. However, the shift variance, the lack of directionality and the substantial aliasing, limit the effects of DWT in some applications. Dual-tree discrete wavelet transform (DTDWT) can overcome those drawbacks but may increase information redundancy. For classification tasks with small dataset sizes, such redundancy can greatly reduce learning efficiency and model performance. In this work, we proposed a novel redundancy removed DTDWT (RR-DTDWT) framework for automated seizure detection. Energy and modified multi-scale entropy (MMSE) features in a dual tree wavelet domain were extracted to construct a complete picture of mental states. To the best of our knowledge, this is the first study to employ MMSE as an indicator of epileptic seizures. Moreover, a compact EEG representation can be obtained after removing useless information redundancy (redundancy between wavelet trees, adjacent channels and entropy scales) by a general auto-weighted feature selection framework via global redundancy minimization (AGRM). Through validation on Bonn and CHB-MIT databases, the proposed RR-DTDWT method can achieve better performance than previous studies.

Low-Intensity Pulsed Ultrasound Stimulation Modulates the Nonlinear Dynamics of Local Field Potentials in Temporal Lobe Epilepsy

Low-intensity pulsed ultrasound stimulation (LIPUS) can inhibit seizures associated with temporal lobe epilepsy (TLE), which is the most common epileptic syndrome in adults and accounts for more than half of the cases of intractable epilepsy. Electroencephalography (EEG) signal analysis is an important method for studying epilepsy. The nonlinear dynamics of epileptic EEG signals can be used as biomarkers for the prediction and diagnosis of epilepsy. However, how ultrasound modulates the nonlinear dynamic characteristics of EEG signals in TLE is still unclear. Here, we used low-intensity pulsed ultrasound to stimulate the CA3 region of kainite (KA)-induced TLE mice, simultaneously recorded local field potentials (LFP) in the stimulation regions before, during, and after LIPUS. The nonlinear characteristics, including complexity, approximate entropy of different frequency bands, and Lyapunov exponent of the LFP, were calculated. Compared with the control group, the experimental group showed that LIPUS inhibited TLE seizure and the complexity, approximate entropy of the delta (0.5–4 Hz) and theta (4–8 Hz) frequency bands, and Lyapunov exponent of the LFP significantly increased in response to ultrasound stimulation. The values before ultrasound stimulation were higher ∼1.87 (complexity), ∼1.39 (approximate entropy of delta frequency bands), ∼1.13 (approximate entropy of theta frequency bands) and ∼1.46 times (Lyapunov exponent) than that after ultrasound stimulation (p < 0.05). The above results demonstrated that LIPUS can alter nonlinear dynamic characteristics and provide a basis for the application of ultrasound stimulation in the treatment of epilepsy.