scholarly journals Robust Deep Network with Maximum Correntropy Criterion for Seizure Detection

2014 ◽  
Vol 2014 ◽  
pp. 1-10 ◽  
Author(s):  
Yu Qi ◽  
Yueming Wang ◽  
Jianmin Zhang ◽  
Junming Zhu ◽  
Xiaoxiang Zheng
Keyword(s):  
Seizure Detection ◽  
Simultaneous Optimization ◽  
Eeg Signals ◽  
Deep Network ◽  
Scalp Eeg ◽  
Impulsive Noises ◽  
Eeg Data ◽  
The Mean ◽  
Maximum Correntropy Criterion

Effective seizure detection from long-term EEG is highly important for seizure diagnosis. Existing methods usually design the feature and classifier individually, while little work has been done for the simultaneous optimization of the two parts. This work proposes a deep network to jointly learn a feature and a classifier so that they could help each other to make the whole system optimal. To deal with the challenge of the impulsive noises and outliers caused by EMG artifacts in EEG signals, we formulate a robust stacked autoencoder (R-SAE) as a part of the network to learn an effective feature. In R-SAE, the maximum correntropy criterion (MCC) is proposed to reduce the effect of noise/outliers. Unlike the mean square error (MSE), the output of the new kernel MCC increases more slowly than that of MSE when the input goes away from the center. Thus, the effect of those noises/outliers positioned far away from the center can be suppressed. The proposed method is evaluated on six patients of 33.6 hours of scalp EEG data. Our method achieves a sensitivity of 100% and a specificity of 99%, which is promising for clinical applications.

scholarly journals Hypoglycemia Detection Based on ear-EEG

2020 ◽  
Author(s):  
Alvaro Fuentes Cabrera ◽  
Simon Lind Kappel ◽  
Line Sofie Remvig ◽  
Hans Olaf Toft ◽  
Rasmus Elsborg Madsen ◽  
...  
Keyword(s):  
Real Life ◽  
Total Power ◽  
Support Vector ◽  
Eeg Signals ◽  
Scalp Eeg ◽  
Oxide Electrodes ◽  
Dry Contact ◽  
The Right ◽  
Classification Procedures

Abstract Background: Hypoglycemia refers to the condition in which the blood glucose is severely below normal level. Hypoglycemia can cause difficulties to talk, headache, irritability, anxiety, confusion, convulsions, seizures, unconsciousness, and even death. These symptoms are a consequence of insufficient supply of glucose to the brain. It has previously been demonstrated that hypoglycemia results in characteristic changes in the electroencephalographic (EEG) signals recorded from electrodes on the scalp. Scalp EEG is not suitable for continuous measurements, due to its obtrusive nature and limited capabilities for monitoring in real-life environments. The objective of this study was to asses the feasibility of detecting hypoglycemia-induced episodes using EEG signals recorded with dry-contact in-ear electrodes, which are discreet, have the potential for long-term EEG monitoring in real-life situations, and provide similar information to that recorded with scalp EEG. The data from 5 diabetic subjects were used for this study. Six ear-EEG channels recorded from dry-contact iridium oxide electrodes fitted the right ear, and channels C3, Pz, T7, and T8 were used for the analysis and classification procedures. A Support Vector Machine (SVM) with a linear kernel was used to detect the hypoglycemic episodes, using a normalized measure of the total power of the θ,α,β, and γ frequency bands as features.Results: The results showed that there were no statistical differences between the sensitivity and specificity of the contralaterally referenced scalp-scalp and ear-scalp EEG channels. Contralaterally referenced channels showed an average sensitivity over all 5 subjects ≥90%, SD≤10% and an average specificity over all 5 subjects ≥82%, SD≤24%. The sensitivities and specificities obtained with the data from the ipsilaterally referenced ear-ear EEG channels did not exceed chance level.Trial registration: ClinicalTrials.gov NCT03022058. Registered 13 December 2016 https://clinicaltrials.gov/ct2/show/NCT03022058?term=uneeg&draw=3&rank=12.

scholarly journals Comparison of Empirical Mode Decomposition, Wavelets, and Different Machine Learning Approaches for Patient-Specific Seizure Detection Using Signal-Derived Empirical Dictionary Approach

2021 ◽  
Vol 3 ◽  
Author(s):  
Muhammad Kaleem ◽  
Aziz Guergachi ◽  
Sridhar Krishnan
Keyword(s):  
Machine Learning ◽  
Seizure Detection ◽  
Reconstruction Error ◽  
Data Driven ◽  
Patient Specific ◽  
Learning Approaches ◽  
Mode Decomposition ◽  
Eeg Data ◽  
Classifier Training

Analysis of long-term multichannel EEG signals for automatic seizure detection is an active area of research that has seen application of methods from different domains of signal processing and machine learning. The majority of approaches developed in this context consist of extraction of hand-crafted features that are used to train a classifier for eventual seizure detection. Approaches that are data-driven, do not use hand-crafted features, and use small amounts of patients' historical EEG data for classifier training are few in number. The approach presented in this paper falls in the latter category, and is based on a signal-derived empirical dictionary approach, which utilizes empirical mode decomposition (EMD) and discrete wavelet transform (DWT) based dictionaries learned using a framework inspired by traditional methods of dictionary learning. Three features associated with traditional dictionary learning approaches, namely projection coefficients, coefficient vector and reconstruction error, are extracted from both EMD and DWT based dictionaries for automated seizure detection. This is the first time these features have been applied for automatic seizure detection using an empirical dictionary approach. Small amounts of patients' historical multi-channel EEG data are used for classifier training, and multiple classifiers are used for seizure detection using newer data. In addition, the seizure detection results are validated using 5-fold cross-validation to rule out any bias in the results. The CHB-MIT benchmark database containing long-term EEG recordings of pediatric patients is used for validation of the approach, and seizure detection performance comparable to the state-of-the-art is obtained. Seizure detection is performed using five classifiers, thereby allowing a comparison of the dictionary approaches, features extracted, and classifiers used. The best seizure detection performance is obtained using EMD based dictionary and reconstruction error feature and support vector machine classifier, with accuracy, sensitivity and specificity values of 88.2, 90.3, and 88.1%, respectively. Comparison is also made with other recent studies using the same database. The methodology presented in this paper is shown to be computationally efficient and robust for patient-specific automatic seizure detection. A data-driven methodology utilizing a small amount of patients' historical data is hence demonstrated as a practical solution for automatic seizure detection.

scholarly journals Automated Detection of Ripple Oscillations in Long-Term Scalp EEG from Patients with Infantile Spasms

2020 ◽  
Author(s):  
Colin M. McCrimmon ◽  
Aliza Riba ◽  
Cristal Garner ◽  
Amy L. Maser ◽  
Daniel W. Shrey ◽  
...  
Keyword(s):  
Spatial Distribution ◽  
Visual Analysis ◽  
Infantile Spasms ◽  
Scalp Eeg ◽  
Automated Algorithm ◽  
Automated Method ◽  
Eeg Data ◽  
Acth Therapy ◽  
The Difference

AbstractObjectiveScalp high frequency oscillations (HFOs) are a promising biomarker of epileptogenicity in infantile spasms (IS) and many other epilepsy syndromes, but prior studies have relied on visual analysis of short segments of data due to the prevalence of artifacts in EEG. Therefore, we set out to develop a fully automated method of HFO detection that can be applied to large datasets, and we sought to robustly characterize the rate and spatial distribution of HFOs in IS.MethodsWe prospectively collected long-term scalp EEG data from 13 subjects with IS and 18 healthy controls. For patients with IS, recording began prior to diagnosis and continued through initiation of treatment with adenocorticotropic hormone (ACTH). The median analyzable EEG duration was 18.2 hours for controls and 83.9 hours for IS subjects (∼1300 hours total). Ripples (80-250 Hz) were detected in all EEG data using an automated algorithm.ResultsHFO rates were substantially higher in patients with IS compared to controls. In IS patients, HFO rates were higher during sleep compared to wakefulness (median 5.5/min and 2.9/min, respectively; p =0.002); controls did not exhibit a difference in HFO rate between sleep and wakefulness (median 0.98/min and 0.82/min, respectively). Spatially, the difference between IS patients and controls was most salient in the central/posterior parasaggital region, where very few HFOs were detected in controls. In IS subjects, ACTH therapy significantly decreased the rate of HFOs.DiscussionHere we show for the first time that a fully automated algorithm can be used to detect HFOs in long-term scalp EEG, and the results are accurate enough to clearly discriminate healthy subjects from those with IS. We also provide a detailed characterization of the spatial distribution and rates of HFOs associated with infantile spasms, which may have relevance for diagnosis and assessment of treatment response.

Graph Attention Network with Focal Loss for Seizure Detection on Electroencephalography Signals

2021 ◽  
pp. 2150027
Author(s):  
Yanna Zhao ◽  
Gaobo Zhang ◽  
Changxu Dong ◽  
Qi Yuan ◽  
Fangzhou Xu ◽  
...  
Keyword(s):  
Detection Method ◽  
Seizure Detection ◽  
Vital Role ◽  
Eeg Signals ◽  
Attention Network ◽  
Data Imbalance ◽  
Imbalance Problem ◽  
Eeg Data ◽  
Epilepsy Diagnosis ◽  
Electroencephalogram Eeg

Automatic seizure detection from electroencephalogram (EEG) plays a vital role in accelerating epilepsy diagnosis. Previous researches on seizure detection mainly focused on extracting time-domain and frequency-domain features from single electrodes, while paying little attention to the positional correlations between different EEG channels of the same subject. Moreover, data imbalance is common in seizure detection scenarios where the duration of nonseizure periods is much longer than the duration of seizures. To cope with the two challenges, a novel seizure detection method based on graph attention network (GAT) is presented. The approach acts on graph-structured data and takes the raw EEG data as input. The positional relationship between different EEG signals is exploited by GAT. The loss function of the GAT model is redefined using the focal loss to tackle data imbalance problem. Experiments are conducted on the CHB-MIT dataset. The accuracy, sensitivity and specificity of the proposed method are 98.89[Formula: see text], 97.10[Formula: see text] and 99.63[Formula: see text], respectively.

scholarly journals Removal of EOG Artifacts from EEG Recordings Using Stationary Subspace Analysis

2014 ◽  
Vol 2014 ◽  
pp. 1-9 ◽  
Author(s):  
Hong Zeng ◽  
Aiguo Song
Keyword(s):  
Covariance Matrix ◽  
Blind Source Separation ◽  
Source Separation ◽  
Experimental Results ◽  
Eeg Signals ◽  
Subspace Analysis ◽  
Eeg Data ◽  
The Mean ◽  
Eeg Recordings ◽  
Ocular Artifacts

An effective approach is proposed in this paper to remove ocular artifacts from the raw EEG recording. The proposed approach first conducts the blind source separation on the raw EEG recording by the stationary subspace analysis (SSA) algorithm. Unlike the classic blind source separation algorithms, SSA is explicitly tailored to the understanding of distribution changes, where both the mean and the covariance matrix are taken into account. In addition, neither independency nor uncorrelation is required among the sources by SSA. Thereby, it can concentrate artifacts in fewer components than the representative blind source separation methods. Next, the components that are determined to be related to the ocular artifacts are projected back to be subtracted from EEG signals, producing the clean EEG data eventually. The experimental results on both the artificially contaminated EEG data and real EEG data have demonstrated the effectiveness of the proposed method, in particular for the cases where limited number of electrodes are used for the recording, as well as when the artifact contaminated signal is highly nonstationary and the underlying sources cannot be assumed to be independent or uncorrelated.

Sign in / Sign up

Export Citation Format

Share Document