Seizure Prediction Using Directed Transfer Function and Convolution Neural Network on Intracranial EEG

Wang, Gang; Wang, Dong; Du, Changsheng; Li, Kuo; Zhang, Junhao; Liu, Zhian; Tao, Yi; Wang, Maode; Cao, Zehong; Yan, Xiao‐Hai

doi:10.1109/tnsre.2020.3035836

Cited by 67 publications

(36 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then, the corrupted vectorx is mapped by the encoder to the feature representation h = f (W (1)x + b (1) ) of x, where f (z) = (1 + exp(−z)) −1 is the nonlinear activation function, and W (1) ∈ R v×u and b (1) ∈ R v are the weight matrix and bias vector of encoder, respectively. Finally, the feature representation h is used by the decoder to calculate the reconstructed representation r = f (W (2) h + b (2) ) of x, where W (2) ∈ R u×v and b (2) ∈ R u are the weight matrix and bias vector of decoder, respectively. The reconstruction error Γ(x, r) = 1 2 ||r − x|| 2 .…”

Section: B Feature Extraction By the Sdae Network In The Timefrequency Domainmentioning

confidence: 99%

“…The reconstruction error Γ(x, r) = 1 2 ||r − x|| 2 . For a training set {x (1) , x (2) , • • • , x (K) }, the loss function of a DAE is as follows:…”

Section: B Feature Extraction By the Sdae Network In The Timefrequency Domainmentioning

confidence: 99%

“…1) Train the first DAE to learn the feature representation h (1) by using the corrupted versions of normalized one-dimensional timefrequency maps of EoIs. 2) Train the second DAE to learn the feature representation h (2) by using h (1) . • • • L) Train the l-th DAE to learn the feature representation h (l) by using h (l−1) , where l ∈ {3, 4, • • • , L}.…”

Section: B Feature Extraction By the Sdae Network In The Timefrequency Domainmentioning

confidence: 99%

“…S URGICAL resection of epileptogenic zones is an effective way to treat patients with drug-resistant epilepsy [1], [2]. As a surrogate region of the epileptogenic zone, the seizure onset zone (SOZ) indicates the cortex area where clinical seizures begin to occur [3].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

HFO Detection in Epilepsy: A Stacked Denoising Autoencoder and Sample Weight Adjusting Factors-Based Method

Wu¹,

Qin

Wan

et al. 2021

IEEE Trans. Neural Syst. Rehabil. Eng.

View full text Add to dashboard Cite

High-frequency oscillations (HFOs) recorded by the intracranial electroencephalography (iEEG) are the promising biomarkers of epileptogenic zones. Accurate detection of HFOs is the key to pre-operative assessment for epilepsy. Due to the subjective bias caused by manual features and the class imbalance between HFOs and false HFOs, it is difficult to obtain satisfactory detection performance by the existing methods. To solve these problems, we put forward a novel method to accurately detect HFOs based on the stacked denoising autoencoder (SDAE) and the ensemble classifier with sample weight adjusting factors. First, the adjustable threshold of Hilbert envelopes is proposed to isolate the events of interest (EoIs) from background activities. Then, the SDAE network is utilized to automatically extract features of EoIs in the time-frequency domain. Finally, the AdaBoost-based support vector machine ensemble classifier with sample weight adjusting factors is devised to separate HFOs from EoIs by using the extracted features. These adjusting factors are used to solve the class imbalance problem by adjusting sample weights when learning the base classifiers. Our HFO detection method is evaluated by using clinical iEEG data recorded from 20 patients with medically refractory epilepsy. The experimental results show that our detection method outperforms some existing methods in terms of sensitivity and false discovery rate. In addition, the HFOs detected by our method are effective for localizing seizure onset zones.

show abstract

Section: B Feature Extraction By the Sdae Network In The Timefrequency Domainmentioning

confidence: 99%

“…The reconstruction error Γ(x, r) = 1 2 ||r − x|| 2 . For a training set {x (1) , x (2) , • • • , x (K) }, the loss function of a DAE is as follows:…”

Section: B Feature Extraction By the Sdae Network In The Timefrequency Domainmentioning

confidence: 99%

Section: B Feature Extraction By the Sdae Network In The Timefrequency Domainmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

HFO Detection in Epilepsy: A Stacked Denoising Autoencoder and Sample Weight Adjusting Factors-Based Method

Wu¹,

Qin

Wan

et al. 2021

IEEE Trans. Neural Syst. Rehabil. Eng.

View full text Add to dashboard Cite

show abstract

“…SVM has become very popular in multiple application and regression problems due to its results, according to Figure 13 The internal structure or model of the NN is composed of a set of inputs x j ; synaptic weights w ij , with j = 1, ..., n; a propagation rule h i defined from the set of inputs and weights; an activation function, which simultaneously represents the neuron output and its activation state. As in SVM, there are currently libraries in different programming languages that make it possible for people without deep knowledge of NN to easily use them in troubleshooting [99]- [101]. Two examples of very common neural networks in EEG applications are described in a general way below: 1) Adaptive Neuronal Network (ANN): It is a type of neural network applied in dynamic environments [102].…”

Section: Classificationmentioning

confidence: 99%

Impact of EEG Parameters Detecting Dementia Diseases: A Systematic Review

Sánchez-Reyes¹,

Rodríguez‐Reséndiz²,

Avecilla-Ramírez³

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Dementia diseases are increasing rapidly, according to the World Health Organization (WHO), becoming an alarming problem for the health sector. The electroencephalogram (EEG) is a non-invasive study that records brain electrical activity and has a wide field of applications in the medical area, one of which is the detection of neurodegenerative diseases. The objective of this work is to present the results of a thorough review of the use of EEG systems for the detection of dementia diseases. Around 82 published papers between 2009 and 2020 were reviewed, and compared among them obtaining data such as sampling time, the number of electrodes, the most popular processing, classification, and validation techniques, as well as an analysis of the reported results. The relationship of the selected parameters with the efficiency obtained is shown, some more common combinations in the reviewed articles that demonstrated to have reliability levels greater than 90% and details to be considered at each stage of the process. An overview of the most commonly used classification tools and processing techniques is also described.

show abstract

A Method for Detecting Epileptic Seizure in Pediatrics Patients Based on EEG Signals

Chakrabarti

Swetapadma

Pattnaik

2022

Proceedings of Trends in Electronics and Health Informatics

View full text Add to dashboard Cite

Seizure Prediction Using Directed Transfer Function and Convolution Neural Network on Intracranial EEG

Cited by 67 publications

References 46 publications

HFO Detection in Epilepsy: A Stacked Denoising Autoencoder and Sample Weight Adjusting Factors-Based Method

HFO Detection in Epilepsy: A Stacked Denoising Autoencoder and Sample Weight Adjusting Factors-Based Method

Impact of EEG Parameters Detecting Dementia Diseases: A Systematic Review

A Method for Detecting Epileptic Seizure in Pediatrics Patients Based on EEG Signals

Contact Info

Product

Resources

About