Detection of Interictal Discharges With Convolutional Neural Networks Using Discrete Ordered Multichannel Intracranial EEG

Αντωνιάδης, Ανδρέας; Spyrou, Loukianos; Martín-López, David; Valentin, Antonio; Alarcón, Gonzalo; Sanei, Saeid; Took, Clive Cheong

doi:10.1109/tnsre.2017.2755770

Cited by 81 publications

(57 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…First, autoencoders are modeled to address the estimation problem in Eq. (1), which was not previously considered in our works 13,23,34 . Second, the classification problem Eq.…”

Section: Problem Formulationmentioning

confidence: 78%

“…Pooling has been proven to achieve invariance to transformation and a more compact representation of images. Although pooling has been proved to be beneficial to deep learning for image processing, it has been found to impede the training of CNNs with epileptic EEG data 34 . This is due to loss of information regarding IED morphology and as a result, pooling has been omitted in this work.…”

Section: Convolutional Neural Networkmentioning

confidence: 99%

See 1 more Smart Citation

Deep Neural Architectures for Mapping Scalp to Intracranial EEG

Αντωνιάδης

Spyrou

Martín-López

et al. 2018

Int. J. Neur. Syst.

Self Cite

View full text Add to dashboard Cite

Data is often plagued by noise which encumbers machine learning of clinically useful biomarkers and electroencephalogram (EEG) data is no exemption. Intracranial EEG (iEEG) data enhances the training of deep learning models of the human brain, yet is often prohibitive due to the invasive recording process. A more convenient alternative is to record brain activity using scalp electrodes. However, the inherent noise associated with scalp EEG data often impedes the learning process of neural models, achieving substandard performance. Here, an ensemble deep learning architecture for nonlinearly mapping scalp to iEEG data is proposed. The proposed architecture exploits the information from a limited number of joint scalp-intracranial recording to establish a novel methodology for detecting the epileptic discharges from the sEEG of a general population of subjects. Statistical tests and qualitative analysis have revealed that the generated pseudo-intracranial data are highly correlated with the true intracranial data. This facilitated the detection of IEDs from the scalp recordings where such waveforms are not often visible. As a real-world clinical application, these pseudo-iEEGs are then used by a convolutional neural network for the automated classification of intracranial epileptic discharges (IEDs) and non-IED of trials in the context of epilepsy analysis. Although the aim of this work was to circumvent the unavailability of iEEG and the limitations of sEEG, we have achieved a classification accuracy of 68% an increase of 6% over the previously proposed linear regression mapping.

show abstract

“…First, autoencoders are modeled to address the estimation problem in Eq. (1), which was not previously considered in our works 13,23,34 . Second, the classification problem Eq.…”

Section: Problem Formulationmentioning

confidence: 78%

Section: Convolutional Neural Networkmentioning

confidence: 99%

Deep Neural Architectures for Mapping Scalp to Intracranial EEG

Αντωνιάδης

Spyrou

Martín-López

et al. 2018

Int. J. Neur. Syst.

Self Cite

View full text Add to dashboard Cite

show abstract

“…In [14], Antoniades et al (2017) presents from intracranial EEG data along with intracranial EEG data along with clinical insight by deliberating deep learning for epileptic patients to include automated generation of feature. Hierarchical Process is utilized for automatic learning the meaningful features in a subject independent fashion.…”

Section: Literature Reviewmentioning

confidence: 99%

“…In earlier days representation based on Fourier transform and parametric methods are used. Variations in frequency sub bands due to epileptic seizure existing in EEG are given as δ(0.4-4 Hz), θ(4-8 Hz),α (8)(9)(10)(11)(12),and β (12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30).Generally Conventional frequency based methods are suitable for decomposing EEG signals due to non stationary and multicomponent signals in EEG. The better performance is observed in time-frequency based methods compared with conventional frequency based methods.…”

Section: Introductionmentioning

confidence: 99%

Feature Extraction and Enhanced Convolutional Neural Network (ECNN) for Detection and Diagnosis of Seizure using EEG Signals

Banu¹,

Suganya²

2019

IJEAT

View full text Add to dashboard Cite

Seizure detection in non-stationary electroencephalography (EEG) is perplexing and difficult task. The human examination for detecting the seizure activities in EEG signals is liable to errors. Apart from the errors, it is a time driven task and also the detection is not precise. In order to detect epileptic seizures more precisely various automatic systems have been emerged to assist neurophysiologists by researchers in various attempts. There are various limitations such as time-consuming, technical artifact issues, result variation with respect to reader expertise level, abnormalities identification. Enhanced Convolutional Neural Network (ECNN) is a technique proposed to mitigate the above mentioned limitations and to categorize more accurate epileptic seizures results. A novel automatic method to sense epileptic seizures using feature extraction and detection is proposed in this research. Linear filter is helpful in reducing the noise along with artifacts when the EEG signals are preprocessed. The noise can be still removed by applying Least Mean Square algorithm. In this proposed research the features are extracted via analytic time frequency with Cascaded wavelet transform and fractal dimension (FD) in order to detect epileptic seizures. Lastly, to analyze the EEG signal for better classification performance of the given dataset, ECNN is adopted. During this research to classify normal, preictal, and seizure classes, a 13-layer deep ECNN algorithm is implemented. This research has special characteristics such that the model yields promising classification accuracy. The experimental result demonstrates that the proposed ECNN is superior in terms of higher sensitivity, specificity, accuracy and lower time complexity rather than the existing methods.

show abstract

“…The locations of the accelerometric and Garmin sensors used to monitor the motion and heart rate data are presented in Figure 1. The methods used for the data processing include data de-noising, statistical methods, neural networks [47], and deep learning [48][49][50][51] methods with convolutional neural networks. The main goal of the present paper is the analysis of accelerometric and heart rate signals to contribute to monitoring physical activities and to the assesment of rehabilitation exercises [11,52].…”

Section: Introductionmentioning

confidence: 99%

Motion Assessment for Accelerometric and Heart Rate Cycling Data Analysis

Charvátová

Procházka

Vyšata

2020

Sensors

View full text Add to dashboard Cite

Motion analysis is an important topic in the monitoring of physical activities and recognition of neurological disorders. The present paper is devoted to motion assessment using accelerometers inside mobile phones located at selected body positions and the records of changes in the heart rate during cycling, under different body loads. Acquired data include 1293 signal segments recorded by the mobile phone and the Garmin device for uphill and downhill cycling. The proposed method is based upon digital processing of the heart rate and the mean power in different frequency bands of accelerometric data. The classification of the resulting features was performed by the support vector machine, Bayesian methods, k-nearest neighbor method, and neural networks. The proposed criterion is then used to find the best positions for the sensors with the highest discrimination abilities. The results suggest the sensors be positioned on the spine for the classification of uphill and downhill cycling, yielding an accuracy of 96.5% and a cross-validation error of 0.04 evaluated by a two-layer neural network system for features based on the mean power in the frequency bands ⟨ 3 , 8 ⟩ and ⟨ 8 , 15 ⟩ Hz. This paper shows the possibility of increasing this accuracy to 98.3% by the use of more features and the influence of appropriate sensor positioning for motion monitoring and classification.

show abstract

Detection of Interictal Discharges With Convolutional Neural Networks Using Discrete Ordered Multichannel Intracranial EEG

Cited by 81 publications

References 47 publications

Deep Neural Architectures for Mapping Scalp to Intracranial EEG

Deep Neural Architectures for Mapping Scalp to Intracranial EEG

Feature Extraction and Enhanced Convolutional Neural Network (ECNN) for Detection and Diagnosis of Seizure using EEG Signals

Motion Assessment for Accelerometric and Heart Rate Cycling Data Analysis

Contact Info

Product

Resources

About