EpilepsyGAN: Synthetic Epileptic Brain Activities With Privacy Preservation

Pascual, Damián; Amirshahi, Alireza; Aminifar, Amir; Atienza, David; Ryvlin, Philippe; Wattenhofer, Roger

doi:10.1109/tbme.2020.3042574

Cited by 40 publications

(24 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…They found that synthetic spectrogram sample augmentation increased the area under the receiver operating characteristic (ROC) curve (AUC) and sensitivity on two different databases respectively. Pascual et al [29] proposed a conditional GAN [30] approach, namely EpilepsyGAN, to synthesize 4s ictal samples from two selected channels for a seizure detection task. They compared the similarity between real and synthetic ictal samples using a spectral cosine metric, and obtained an average 1.3% improved sensitivity for 24 patients in the ES detection task.…”

Section: A Literature Reviewmentioning

confidence: 99%

Multichannel Synthetic Preictal EEG Signals to Enhance the Prediction of Epileptic Seizures

Xu,

Yang,

Sawan

2022

Preprint

View full text Add to dashboard Cite

Epilepsy is a chronic neurological disorder affecting 1% of people worldwide, deep learning (DL) algorithms-based electroencephalograph (EEG) analysis provides the possibility for accurate epileptic seizure (ES) prediction, thereby benefiting patients suffering from epilepsy. To identify the preictal region that precedes the onset of seizure, a large number of annotated EEG signals are required to train DL algorithms. However, the scarcity of seizure onsets leads to significant insufficiency of data for training the DL algorithms. To overcome this data insufficiency, in this paper, we propose a preictal artificial signal synthesis algorithm based on a generative adversarial network to generate synthetic multichannel EEG preictal samples. A high-quality single-channel architecture, determined by visual and statistical evaluations, is used to train the generators of multichannel samples. The effectiveness of the synthetic samples is evaluated by comparing the ES prediction performances without and with synthetic preictal sample augmentation. The leave-one-seizure-out cross validation ES prediction accuracy and corresponding area under the receiver operating characteristic curve evaluation improve from 73.0% and 0.676 to 78.0% and 0.704 by 10× synthetic sample augmentation, respectively. The obtained results indicate that synthetic preictal samples are effective for enhancing ES prediction performance.

show abstract

Section: A Literature Reviewmentioning

confidence: 99%

Multichannel Synthetic Preictal EEG Signals to Enhance the Prediction of Epileptic Seizures

Xu,

Yang,

Sawan

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…In the particular case of few existing relevant data in datasheets, transfer learning [53][54][55][56], data augmentation [57][58][59][60], or synthetic data [61][62][63][64] techniques are usually applied to generate improved machine learning training and models. Transfer learning is based on the knowledge acquired from another existing learned task to improve the performance of a new machine learning model; thus, reducing the amount of required training data.…”

Section: Regression Treesmentioning

confidence: 99%

A Survey on EEG Signal Processing Techniques and Machine Learning: Applications to the Neurofeedback of Autobiographical Memory Deficits in Schizophrenia

et al. 2021

View full text Add to dashboard Cite

In this paper, a general overview regarding neural recording, classical signal processing techniques and machine learning classification algorithms applied to monitor brain activity is presented. Currently, several approaches classified as electrical, magnetic, neuroimaging recordings and brain stimulations are available to obtain neural activity of the human brain. Among them, non-invasive methods like electroencephalography (EEG) are commonly employed, as they can provide a high degree of temporal resolution (on the order of milliseconds) and acceptable space resolution. In addition, it is simple, quick, and does not create any physical harm or stress to patients. Concerning signal processing, once the neural signals are acquired, different procedures can be applied for feature extraction. In particular, brain signals are normally processed in time, frequency, and/or space domains. The features extracted are then used for signal classification depending on its characteristics such us the mean, variance or band power. The role of machine learning in this regard has become of key importance during the last years due to its high capacity to analyze complex amounts of data. The algorithms employed are generally classified in supervised, unsupervised and reinforcement techniques. A deep review of the most used machine learning algorithms and the advantages/drawbacks of most used methods is presented. Finally, a study of these procedures utilized in a very specific and novel research field of electroencephalography, i.e., autobiographical memory deficits in schizophrenia, is outlined.

show abstract

“…Samples of pre-ictal and ictal periods were created using 4s sample size (as in [16]). Periods shorter than the sample size were discarded.…”

Section: B Pre-processing and Feature Extractionmentioning

confidence: 99%

Features importance in seizure classification using scalp EEG reduced to single timeseries

Naze

Tang

Kozloski³

et al. 2021

Preprint

View full text Add to dashboard Cite

Seizure detection and seizure-type classification are best performed using intra-cranial or full-scalp electroencephalogram (EEG). In embedded wearable systems however, recordings from only a few electrodes are available, reducing the spatial resolution of the signals to a handful of timeseries at most. Taking this constraint into account, we tested the performance of multiple classifiers using a subset of the EEG recordings by selecting a single trace from the montage or performing a dimensionality reduction over each hemispherical space. Our results support that Random Forest (RF) classifiers lead most efficient and stable classification performances over Support Vector Machines (SVM). Interestingly, tracking the feature importances using permutation tests reveals that classical EEG spectrum power bands display different rankings across the classifiers: low frequencies (delta, theta) are most important for SVMs while higher frequencies (alpha, gamma) are more relevant for RF and Decision Trees. We reach up to 94.3% +/- 5.3% accuracy in classifying absence from tonic-clonic seizures using state-of-art sampling methods for unbalanced datasets and leave-patients-out 3-fold cross-validation policy.

show abstract

EpilepsyGAN: Synthetic Epileptic Brain Activities With Privacy Preservation

Cited by 40 publications

References 39 publications

Multichannel Synthetic Preictal EEG Signals to Enhance the Prediction of Epileptic Seizures

Multichannel Synthetic Preictal EEG Signals to Enhance the Prediction of Epileptic Seizures

A Survey on EEG Signal Processing Techniques and Machine Learning: Applications to the Neurofeedback of Autobiographical Memory Deficits in Schizophrenia

Features importance in seizure classification using scalp EEG reduced to single timeseries

Contact Info

Product

Resources

About