The EPILEPSIAE database: An extensive electroencephalography database of epilepsy patients

Addressing data limitations in seizure prediction through transfer learning

According to the literature, seizure prediction models should be developed following a patient-specific approach. However, seizures are usually very rare events, meaning the number of events that may be used to optimise seizure prediction approaches is limited. To overcome such constraint, we analysed the possibility of using data from patients from an external database to improve patient-specific seizure prediction models. We present seizure prediction models trained using a transfer learning procedure. We trained a deep convolutional autoencoder using electroencephalogram data from 41 patients collected from the EPILEPSIAE database. Then, a bidirectional long short-term memory and a classifier layers were added on the top of the encoder part and were optimised for 24 patients from the Universitätsklinikum Freiburg individually. The encoder was used as a feature extraction module. Therefore, its weights were not changed during the patient-specific training. Experimental results showed that seizure prediction models optimised using pretrained weights present about four times fewer false alarms while maintaining the same ability to predict seizures and achieved more 13% validated patients. Therefore, results evidenced that the optimisation using transfer learning was more stable and faster, saving computational resources. In summary, adopting transfer learning for seizure prediction models represents a significant advancement. It addresses the data limitation seen in the seizure prediction field and offers more efficient and stable training, conserving computational resources. Additionally, despite the compact size, transfer learning allows to easily share data knowledge due to fewer ethical restrictions and lower storage requirements. The convolutional autoencoder developed in this study will be shared with the scientific community, promoting further research.

A scalable platform for acquisition of high-fidelity human intracranial EEG with minimal clinical burden

Human neuroscience research has been significantly advanced by neuroelectrophysiological studies from people with refractory epilepsy-the only routine clinical intervention that acquires multi-day, multi-electrode human intracranial electroencephalography (iEEG). While a sampling rate below 2 kHz is sufficient for manual iEEG review by epileptologists, computational methods and research studies may benefit from higher resolution, which requires significant technical development. At adult and pediatric Stanford hospitals, research ports of commercial clinical acquisition systems were configured to collect 10 kHz iEEG of up to 256 electrodes simultaneously with the clinical data. The research digital stream was designed to be acquired post-digitization, resulting in no loss in clinical signal quality. This novel framework implements a near-invisible research platform to facilitate the secure, routine collection of high-resolution iEEG that minimizes research hardware footprint and clinical workflow interference. The addition of a pocket-sized router in the patient room enabled an encrypted tunnel to securely transmit research-quality iEEG across hospital networks to a research computer within the hospital server room, where data was coded, de-identified, and uploaded to cloud storage. Every eligible patient undergoing iEEG clinical evaluation at both hospitals since September 2017 has been recruited; participant recruitment is ongoing. Over 350+ terabytes (representing 1000+ days) of neuroelectrophysiology were recorded across 200+ participants of diverse demographics. To our knowledge, this is the first report of such a research integration within a hospital setting. It is a promising approach to promoting equitable participant enrollment and building comprehensive data repositories with consistent, high-fidelity specifications towards new discoveries in human neuroscience.

Artificial intelligence in epilepsy - applications and pathways to the clinic

Artificial intelligence (AI) is rapidly transforming health care, and its applications in epilepsy have increased exponentially over the past decade. Integration of AI into epilepsy management promises to revolutionize the diagnosis and treatment of this complex disorder. However, translation of AI into neurology clinical practice has not yet been successful, emphasizing the need to consider progress to date and assess challenges and limitations of AI. In this Review, we provide an overview of AI applications that have been developed in epilepsy using a variety of data modalities: neuroimaging, electroencephalography, electronic health records, medical devices and multimodal data integration. For each, we consider potential applications, including seizure detection and prediction, seizure lateralization, localization of the seizure-onset zone and assessment for surgical or neurostimulation interventions, and review the performance of AI tools developed to date. We also discuss methodological considerations and challenges that must be addressed to successfully integrate AI into clinical practice. Our goal is to provide an overview of the current state of the field and provide guidance for leveraging AI in future to improve management of epilepsy.

Concept-drifts adaptation for machine learning EEG epilepsy seizure prediction

Seizure prediction remains a challenge, with approximately 30% of patients unresponsive to conventional treatments. Addressing this issue is crucial for improving patients' quality of life, as timely intervention can mitigate the impact of seizures. In this research field, it is critical to identify the preictal interval, the transition from regular brain activity to a seizure. While previous studies have explored various Electroencephalogram (EEG) based methodologies for prediction, few have been clinically applicable. Recent studies have underlined the dynamic nature of EEG data, characterised by data changes with time, known as concept drifts, highlighting the need for automated methods to detect and adapt to these changes. In this study, we investigate the effectiveness of automatic concept drift adaptation methods in seizure prediction. Three patient-specific seizure prediction approaches with a 10-minute prediction horizon are compared: a seizure prediction algorithm incorporating a window adjustment method by optimising performance with Support Vector Machines (Backwards-Landmark Window), a seizure prediction algorithm incorporating a data-batch (seizures) selection method using a logistic regression (Seizure-batch Regression), and a seizure prediction algorithm with a dynamic integration of classifiers (Dynamic Weighted Ensemble). These methods incorporate a retraining process after each seizure and use a combination of univariate linear features and SVM classifiers. The Firing Power was used as a post-processing technique to generate alarms before seizures. These methodologies were compared with a control approach based on the typical machine learning pipeline, considering a group of 37 patients with Temporal Lobe Epilepsy from the EPILEPSIAE database. The best-performing approach (Backwards-Landmark Window) achieved results of 0.75 ± 0.33 for sensitivity and 1.03 ± 1.00 for false positive rate per hour. This new strategy performed above chance for 89% of patients with the surrogate predictor, whereas the control approach only validated 46%.

Combining general and personal models for epilepsy detection with hyperdimensional computing

Epilepsy is a highly prevalent chronic neurological disorder with great negative impact on patients' daily lives. Despite this there is still no adequate technological support to enable epilepsy detection and continuous outpatient monitoring in everyday life. Hyperdimensional (HD) computing is a promising method for epilepsy detection via wearable devices, characterized by a simpler learning process and lower memory requirements compared to other methods. In this work, we demonstrate additional avenues in which HD computing and the manner in which its models are built and stored can be used to better understand, compare and create more advanced machine learning models for epilepsy detection. These possibilities are not feasible with other state-of-the-art models, such as random forests or neural networks. We compare inter-subject model similarity of different classes (seizure and non-seizure), study the process of creating general models from personal ones, and finally posit a method of combining personal and general models to create hybrid models. This results in an improved epilepsy detection performance. We also tested knowledge transfer between models trained on two different datasets. The attained insights are highly interesting not only from an engineering perspective, to create better models for wearables, but also from a neurological perspective, to better understand individual epilepsy patterns.

EEG epilepsy seizure prediction: the post-processing stage as a chronology

Almost one-third of epileptic patients fail to achieve seizure control through anti-epileptic drug administration. In the scarcity of completely controlling a patient's epilepsy, seizure prediction plays a significant role in clinical management and providing new therapeutic options such as warning or intervention devices. Seizure prediction algorithms aim to identify the preictal period that Electroencephalogram (EEG) signals can capture. However, this period is associated with substantial heterogeneity, varying among patients or even between seizures from the same patient. The present work proposes a patient-specific seizure prediction algorithm using post-processing techniques to explore the existence of a set of chronological events of brain activity that precedes epileptic seizures. The study was conducted with 37 patients with Temporal Lobe Epilepsy (TLE) from the EPILEPSIAE database. The designed methodology combines univariate linear features with a classifier based on Support Vector Machines (SVM) and two post-processing techniques to handle pre-seizure temporality in an easily explainable way, employing knowledge from network theory. In the Chronological Firing Power approach, we considered the preictal as a sequence of three brain activity events separated in time. In the Cumulative Firing Power approach, we assumed the preictal period as a sequence of three overlapping events. These methodologies were compared with a control approach based on the typical machine learning pipeline. We considered a Seizure Prediction horizon (SPH) of 5 mins and analyzed several values for the Seizure Occurrence Period (SOP) duration, between 10 and 55 mins. Our results showed that the Cumulative Firing Power approach may improve the seizure prediction performance. This new strategy performed above chance for 62% of patients, whereas the control approach only validated 49% of its models.

Seizure forecasting using minimally invasive, ultra-long-term subcutaneous electroencephalography: Individualized intrapatient models

OBJECTIVE

One of the most disabling aspects of living with chronic epilepsy is the unpredictability of seizures. Cumulative research in the past decades has advanced our understanding of the dynamics of seizure risk. Technological advances have recently made it possible to record pertinent biological signals, including electroencephalogram (EEG), continuously. We aimed to assess whether patient-specific seizure forecasting is possible using remote, minimally invasive ultra-long-term subcutaneous EEG.

METHODS

We analyzed a two-center cohort of ultra-long-term subcutaneous EEG recordings, including six patients with drug-resistant focal epilepsy monitored for 46-230 days with median 18 h/day of recorded data, totaling >11 000 h of EEG. Total electrographic seizures identified by visual review ranged from 12 to 36 per patient. Three candidate subject-specific long short-term memory network deep learning classifiers were trained offline and pseudoprospectively on preictal (1 h before) and interictal (>1 day from seizures) EEG segments. Performance was assessed relative to a random predictor. Periodicity of the final forecasts was also investigated with autocorrelation.

RESULTS

Depending on each architecture, significant forecasting performance was achieved in three to five of six patients, with overall mean area under the receiver operating characteristic curve of .65-.74. Significant forecasts showed sensitivity ranging from 64% to 80% and time in warning from 10.9% to 44.4%. Overall, the output of the forecasts closely followed patient-specific circadian patterns of seizure occurrence.

SIGNIFICANCE

This study demonstrates proof-of-principle for the possibility of subject-specific seizure forecasting using a minimally invasive subcutaneous EEG device capable of ultra-long-term at-home recordings. These results are encouraging for the development of a prospective seizure forecasting trial with minimally invasive EEG.

Seizure-Cluster-Inception CNN (SciCNN): A Patient-Independent Epilepsy Tracking SoC With 0-Shot-Retraining

Epilepsy tracking System-on-Chips (SoC) usually perform patient-specific classification to deal with the patient-to-patient seizure pattern variation from a surface electroencephalogram (EEG). However, the patient-specific classifier training requires the EEG signals from the target patients a priori, which involves costly and time-consuming hospitalization for the inpatient data collection. To address this issue, this paper presents a patient-independent epilepsy tracking SoC that is trained with pre-existing databases and can be directly deployed to the target patients without collecting their data and performing cumbersome patient-specific training beforehand. The proposed SoC adopts a Seizure-Cluster-Inception Convolutional Neural Network (SciCNN) Neural Processor (SNP) to reduce SRAM access rate by 179.05× with the Kernel-Wise Pipeline (KWP). The 22-Ch. SoC achieves event-based sensitivity of 90.3%/90.4%/83.3% and specificity of 93.6%/95.7%/88.6% on unseen patients from CHB-MIT database/EU database/local hospital patient, respectively.

A comparison of neuroelectrophysiology databases

As data sharing has become more prevalent, three pillars - archives, standards, and analysis tools - have emerged as critical components in facilitating effective data sharing and collaboration. This paper compares four freely available intracranial neuroelectrophysiology data repositories: Data Archive for the BRAIN Initiative (DABI), Distributed Archives for Neurophysiology Data Integration (DANDI), OpenNeuro, and Brain-CODE. The aim of this review is to describe archives that provide researchers with tools to store, share, and reanalyze both human and non-human neurophysiology data based on criteria that are of interest to the neuroscientific community. The Brain Imaging Data Structure (BIDS) and Neurodata Without Borders (NWB) are utilized by these archives to make data more accessible to researchers by implementing a common standard. As the necessity for integrating large-scale analysis into data repository platforms continues to grow within the neuroscientific community, this article will highlight the various analytical and customizable tools developed within the chosen archives that may advance the field of neuroinformatics.

PreEpiSeizures: description and outcomes of physiological data acquisition using wearable devices during video-EEG monitoring in people with epilepsy

The PreEpiSeizures project was created to better understand epilepsy and seizures through wearable technologies. The motivation was to capture physiological information related to epileptic seizures, besides Electroencephalography (EEG) during video-EEG monitorings. If other physiological signals have reliable information of epileptic seizures, unobtrusive wearable technology could be used to monitor epilepsy in daily life. The development of wearable solutions for epilepsy is limited by the nonexistence of datasets which could validate these solutions. Three different form factors were developed and deployed, and the signal quality was assessed for all acquired biosignals. The wearable data acquisition was performed during the video-EEG of patients with epilepsy. The results achieved so far include 59 patients from 2 hospitals totaling 2,721 h of wearable data and 348 seizures. Besides the wearable data, the Electrocardiogram of the hospital is also useable, totalling 5,838 h of hospital data. The quality ECG signals collected with the proposed wearable is equated with the hospital system, and all other biosignals also achieved state-of-the-art quality. During the data acquisition, 18 challenges were identified, and are presented alongside their possible solutions. Though this is an ongoing work, there were many lessons learned which could help to predict possible problems in wearable data collections and also contribute to the epilepsy community with new physiological information. This work contributes with original wearable data and results relevant to epilepsy research, and discusses relevant challenges that impact wearable health monitoring.

General and patient-specific seizure classification using deep neural networks

Seizure prediction algorithms have been central in the field of data analysis for the improvement of epileptic patients’ lives. The most recent advancements of which include the use of deep neural networks to present an optimized, accurate seizure prediction system. This work puts forth deep learning methods to automate the process of epileptic seizure detection with electroencephalogram (EEG) signals as input; both a patient-specific and general approach are followed. EEG signals are time structure series motivating the use of sequence algorithms such as temporal convolutional neural networks (TCNNs), and long short-term memory networks. We then compare this methodology to other prior pre-implemented structures, including our previous work for seizure prediction using machine learning approaches support vector machine and random under-sampling boost. Moreover, patient-specific and general seizure prediction approaches are used to evaluate the performance of the best algorithms. Area under curve (AUC) is used to select the best performing algorithm to account for the imbalanced dataset. The presented TCNN model showed the best patient-specific results than that of the general approach with, AUC of 0.73, while ML model had the best results for general classification with AUC of 0.75.

Removing artefacts and periodically retraining improve performance of neural network-based seizure prediction models

The development of seizure prediction models is often based on long-term scalp electroencephalograms (EEGs) since they capture brain electrical activity, are non-invasive, and come at a relatively low-cost. However, they suffer from major shortcomings. First, long-term EEG is usually highly contaminated with artefacts. Second, changes in the EEG signal over long intervals, known as concept drift, are often neglected. We evaluate the influence of these problems on deep neural networks using EEG time series and on shallow neural networks using widely-used EEG features. Our patient-specific prediction models were tested in 1577 hours of continuous EEG, containing 91 seizures from 41 patients with temporal lobe epilepsy who were undergoing pre-surgical monitoring. Our results showed that cleaning EEG data, using a previously developed artefact removal method based on deep convolutional neural networks, improved prediction performance. We also found that retraining the models over time reduced false predictions. Furthermore, the results show that although deep neural networks processing EEG time series are less susceptible to false alarms, they may need more data to surpass feature-based methods. These findings highlight the importance of robust data denoising and periodic adaptation of seizure prediction models.

Data Archive for the BRAIN Initiative (DABI)

Data sharing is becoming ubiquitous and can be advantageous for most biomedical research. However, some data are inherently more amenable to sharing than others. For example, human intracranial neurophysiology recordings and associated multimodal data have unique features that warrant special considerations. The associated data are heterogeneous, difficult to compare, highly specific, and collected from small cohorts with treatment resistant conditions, posing additional complications when attempting to perform generalizable analyses across projects. We present the Data Archive for the BRAIN Initiative (DABI) and describe features of the platform that are designed to overcome these and other challenges. DABI is a data repository and portal for BRAIN Initiative projects that collect human and animal intracranial recordings, and it allows users to search, visualize, and analyze multimodal data from these projects. The data providers maintain full control of data sharing privileges and can organize and manage their data with a user-friendly and intuitive interface. We discuss data privacy and security concerns, example analyses from two DABI datasets, and future goals for DABI.

Patient-independent seizure detection based on long-term iEEG and a novel lightweight CNN

Objective.Patient-dependent seizure detection based on intracranial electroencephalography (iEEG) has made significant progress. However, due to the difference in the locations and number of iEEG electrodes used for each patient, patient-independent seizure detection based on iEEG has not been carried out. Additionally, current seizure detection algorithms based on deep learning have outperformed traditional machine learning algorithms in many performance metrics. However, they still have shortcomings of large memory footprints and slow inference speed.Approach.To solve the above problems of the current study, we propose a novel lightweight convolutional neural network model combining the Convolutional Block Attention Module (CBAM). Its performance for patient-independent seizure detection is evaluated on two long-term continuous iEEG datasets: SWEC-ETHZ and TJU-HH. Finally, we reproduce four other patient-independent methods to compare with our method and calculate the memory footprints and inference speed for all methods.Main results.Our method achieves 83.81% sensitivity (SEN) and 85.4% specificity (SPE) on the SWEC-ETHZ dataset and 86.63% SEN and 92.21% SPE on the TJU-HH dataset. In particular, it takes only 11 ms to infer 10 min iEEG (128 channels), and its memory footprint is only 22 kB. Compared to baseline methods, our method not only achieves better patient-independent seizure detection performance but also has a smaller memory footprint and faster inference speed.Significance.To our knowledge, this is the first iEEG-based patient-independent seizure detection study. This facilitates the application of seizure detection algorithms to the future clinic.

Generalizable epileptic seizures prediction based on deep transfer learning

Predicting seizures before they happen can help prevent them through medication. In this research, first, a total of 22 features were extracted from 5-s segmented EEG signals. Second, tensors were developed as inputs for different deep transfer learning models to find the best model for predicting epileptic seizures. The effect of Pre-ictal state duration was also investigated by selecting four different intervals of 10, 20, 30, and 40 min. Then, nine models were created by combining three ImageNet convolutional networks with three classifiers and were examined for predicting seizures patient-dependently. The Xception convolutional network with a Fully Connected (FC) classifier achieved an average sensitivity of 98.47% and a False Prediction Rate (FPR) of 0.031 h-1 in a 40-min Pre-ictal state for ten patients from the European database. The most promising result of this study was the patient-independent prediction of epileptic seizures; the MobileNet-V2 model with an FC classifier was trained with one patient's data and tested on six other patients, achieving a sensitivity rate of 98.39% and an FPR of 0.029 h-1 for a 40-min Pre-ictal scheme.

Unsupervised EEG preictal interval identification in patients with drug-resistant epilepsy

Typical seizure prediction models aim at discriminating interictal brain activity from pre-seizure electrographic patterns. Given the lack of a preictal clinical definition, a fixed interval is widely used to develop these models. Recent studies reporting preictal interval selection among a range of fixed intervals show inter- and intra-patient preictal interval variability, possibly reflecting the heterogeneity of the seizure generation process. Obtaining accurate labels of the preictal interval can be used to train supervised prediction models and, hence, avoid setting a fixed preictal interval for all seizures within the same patient. Unsupervised learning methods hold great promise for exploring preictal alterations on a seizure-specific scale. Multivariate and univariate linear and nonlinear features were extracted from scalp electroencephalography (EEG) signals collected from 41 patients with drug-resistant epilepsy undergoing presurgical monitoring. Nonlinear dimensionality reduction was performed for each group of features and each of the 226 seizures. We applied different clustering methods in searching for preictal clusters located until 2 h before the seizure onset. We identified preictal patterns in 90% of patients and 51% of the visually inspected seizures. The preictal clusters manifested a seizure-specific profile with varying duration (22.9 ± 21.0 min) and starting time before seizure onset (47.6 ± 27.3 min). Searching for preictal patterns on the EEG trace using unsupervised methods showed that it is possible to identify seizure-specific preictal signatures for some patients and some seizures within the same patient.

Distinct signatures of loss of consciousness in focal impaired awareness versus tonic-clonic seizures

Loss of consciousness is a hallmark of many epileptic seizures and carries risks of serious injury and sudden death. While cortical sleep-like activities accompany loss of consciousness during focal impaired awareness seizures, the mechanisms of loss of consciousness during focal to bilateral tonic-clonic seizures remain unclear. Quantifying differences in markers of cortical activation and ictal recruitment between focal impaired awareness and focal to bilateral tonic-clonic seizures may also help us to understand their different consequences for clinical outcomes and to optimize neuromodulation therapies. We quantified clinical signs of loss of consciousness and intracranial EEG activity during 129 focal impaired awareness and 50 focal to bilateral tonic-clonic from 41 patients. We characterized intracranial EEG changes both in the seizure onset zone and in areas remote from the seizure onset zone with a total of 3386 electrodes distributed across brain areas. First, we compared the dynamics of intracranial EEG sleep-like activities: slow-wave activity (1-4 Hz) and beta/delta ratio (a validated marker of cortical activation) during focal impaired awareness versus focal to bilateral tonic-clonic. Second, we quantified differences between focal to bilateral tonic-clonic and focal impaired awareness for a marker validated to detect ictal cross-frequency coupling: phase-locked high gamma (high-gamma phased-locked to low frequencies) and a marker of ictal recruitment: the epileptogenicity index. Third, we assessed changes in intracranial EEG activity preceding and accompanying behavioural generalization onset and their correlation with electromyogram channels. In addition, we analysed human cortical multi-unit activity recorded with Utah arrays during three focal to bilateral tonic-clonic seizures. Compared to focal impaired awareness, focal to bilateral tonic-clonic seizures were characterized by deeper loss of consciousness, even before generalization occurred. Unlike during focal impaired awareness, early loss of consciousness before generalization was accompanied by paradoxical decreases in slow-wave activity and by increases in high-gamma activity in parieto-occipital and temporal cortex. After generalization, when all patients displayed loss of consciousness, stronger increases in slow-wave activity were observed in parieto-occipital cortex, while more widespread increases in cortical activation (beta/delta ratio), ictal cross-frequency coupling (phase-locked high gamma) and ictal recruitment (epileptogenicity index). Behavioural generalization coincided with a whole-brain increase in high-gamma activity, which was especially synchronous in deep sources and could not be explained by EMG. Similarly, multi-unit activity analysis of focal to bilateral tonic-clonic revealed sustained increases in cortical firing rates during and after generalization onset in areas remote from the seizure onset zone. Overall, these results indicate that unlike during focal impaired awareness, the neural signatures of loss of consciousness during focal to bilateral tonic-clonic consist of paradoxical increases in cortical activation and neuronal firing found most consistently in posterior brain regions. These findings suggest differences in the mechanisms of ictal loss of consciousness between focal impaired awareness and focal to bilateral tonic-clonic and may account for the more negative prognostic consequences of focal to bilateral tonic-clonic.

The performance evaluation of the state-of-the-art EEG-based seizure prediction models

The recurrent and unpredictable nature of seizures can lead to unintentional injuries and even death. The rapid development of electroencephalogram (EEG) and Artificial Intelligence (AI) technologies has made it possible to predict seizures in real-time through brain-machine interfaces (BCI), allowing advanced intervention. To date, there is still much room for improvement in predictive seizure models constructed by EEG using machine learning (ML) and deep learning (DL). But, the most critical issue is how to improve the performance and generalization of the model, which involves some confusing conceptual and methodological issues. This review focuses on analyzing several factors affecting the performance of seizure prediction models, focusing on the aspects of post-processing, seizure occurrence period (SOP), seizure prediction horizon (SPH), and algorithms. Furthermore, this study presents some new directions and suggestions for building high-performance prediction models in the future. We aimed to clarify the concept for future research in related fields and improve the performance of prediction models to provide a theoretical basis for future applications of wearable seizure detection devices.

EPIC: Annotated epileptic EEG independent components for artifact reduction

Scalp electroencephalogram is a non-invasive multi-channel biosignal that records the brain’s electrical activity. It is highly susceptible to noise that might overshadow important data. Independent component analysis is one of the most used artifact removal methods. Independent component analysis separates data into different components, although it can not automatically reject the noisy ones. Therefore, experts are needed to decide which components must be removed before reconstructing the data. To automate this method, researchers have developed classifiers to identify noisy components. However, to build these classifiers, they need annotated data. Manually classifying independent components is a time-consuming task. Furthermore, few labelled data are publicly available. This paper presents a source of annotated electroencephalogram independent components acquired from patients with epilepsy (EPIC Dataset). This dataset contains 77,426 independent components obtained from approximately 613 hours of electroencephalogram, visually inspected by two experts, which was already successfully utilised to develop independent component classifiers.

Measurement(s)	Electroencephalogram Independent Components
Technology Type(s)	MATLAB and Python

Weak self-supervised learning for seizure forecasting: a feasibility study

This paper proposes an artificial intelligence system that continuously improves over time at event prediction using initially unlabelled data by using self-supervised learning. Time-series data are inherently autocorrelated. By using a detection model to generate weak labels on the fly, which are concurrently used as targets to train a prediction model on a time-shifted input data stream, this autocorrelation can effectively be harnessed to reduce the burden of manual labelling. This is critical in medical patient monitoring, as it enables the development of personalized forecasting models without demanding the annotation of long sequences of physiological signal recordings. We perform a feasibility study on seizure prediction, which is identified as an ideal test case, as pre-ictal brainwaves are patient-specific, and tailoring models to individual patients is known to improve forecasting performance significantly. Our self-supervised approach is used to train individualized forecasting models for 10 patients, showing an average relative improvement in sensitivity by 14.30% and a reduction in false alarms by 19.61% in early seizure forecasting. This proof-of-concept on the feasibility of using a continuous stream of time-series neurophysiological data paves the way towards a low-power neuromorphic neuromodulation system.

Interictal epileptiform discharges show distinct spatiotemporal and morphological patterns across wake and sleep

Presurgical evaluation of mesial temporal and neocortical focal pharmacoresistant epilepsy patients using intracranial EEG recordings has led to the generation of extensive data on interictal epileptiform discharges, located within or remotely from seizure onset zones. In this study, we used this data to investigate how interictal epileptiform discharges are modulated and how their spatial distribution changes during wake and sleep and analysed the relationship between these discharge events and seizure onset zones. Preoperative evaluation data from 11 adult patients with focal pharmacoresistant epilepsy were extracted from the Epilepsiae database. Interictal epileptiform discharges were automatically detected during wakefulness and over several hours of continuous seizure-free sleep (total duration of EEG recordings:106.7 h; mean per patient: 9.7 h), and analysed across four brain areas (mesial temporal, lateral neocortical, basal cortical and the temporal pole). Sleep stages were classified manually from scalp EEG. Discharge events were characterized according to their rate and morphology (amplitude, sharpness and duration). Eight patients had a seizure onset zone over mesial areas and three patients over lateral neocortical areas. Overall, discharge rates varied across brain areas during wakefulness and sleep [wake/sleep stages × brain areas interaction; Wald χ 2(df = 6) = 31.1, P < 0.0001]. N2-N3 non-rapid eye movement sleep increased interictal epileptiform discharges in mesial areas compared with wakefulness and rapid eye movement sleep (P < 0.0001), and to other areas (P < 0.0001 for all comparisons). This mesial pattern was observed both within and outside of seizure onset zones. During wakefulness, the rate of interictal epileptiform discharges was significantly higher than during N2-N3 non-rapid eye movement sleep (P = 0.04), and rapid eye movement sleep (P = 0.01) in lateral neocortical areas (referred to as lateral neocortical pattern), a finding that was more pronounced in seizures onset zones (P = 0.004). The morphological characteristics of the discharge events were modulated during wakefulness and sleep stages across brain areas. The effect of seizure onset zones on discharge morphology was conditioned by brain area and was particularly marked in temporal pole areas. Our analysis of discharge patterns in relation to cerebral localization, vigilance state and the anatomical affiliation of seizure onset zones revealed the global and local aspects of the complex relationship between interictal discharges, sleep and seizure onset zones. This novel approach may lead to a better understanding of cognitive decline and responses to therapy, as well as to adaptation of surgical interventions for epileptic patients.

On the clinical acceptance of black-box systems for EEG seizure prediction

Seizure prediction may be the solution for epileptic patients whose drugs and surgery do not control seizures. Despite 46 years of research, few devices/systems underwent clinical trials and/or are commercialized, where the most recent state-of-the-art approaches, as neural networks models, are not used to their full potential. The latter demonstrates the existence of social barriers to new methodologies due to data bias, patient safety, and legislation compliance. In the form of literature review, we performed a qualitative study to analyze the seizure prediction ecosystem to find these social barriers. With the Grounded Theory, we draw hypotheses from data, while with the Actor-Network Theory we considered that technology shapes social configurations and interests, being fundamental in healthcare. We obtained a social network that describes the ecosystem and propose research guidelines aiming at clinical acceptance. Our most relevant conclusion is the need for model explainability, but not necessarily intrinsically interpretable models, for the case of seizure prediction. Accordingly, we argue that it is possible to develop robust prediction models, including black-box systems to some extent, while avoiding data bias, ensuring patient safety, and still complying with legislation, if they can deliver human- comprehensible explanations. Due to skepticism and patient safety reasons, many authors advocate the use of transparent models which may limit their performance and potential. Our study highlights a possible path, by using model explainability, on how to overcome these barriers while allowing the use of more computationally robust models.

Spectral and network characterization of focal seizure types and phases

BACKGROUND AND OBJECTIVE

Currently, epileptic seizure characterization relies on several clinical features that allow their classification into different types. The present work aims to characterize both seizure types and phases based exclusively on electrophysiological characteristics.

METHODS

Based on the analysis of intracranial EEG recordings of 129 seizures from 22 patients obtained from the European Epilepsy Database, network and spectral measures were calculated in five-second temporal windows. Statistically significant differences between each window of the seizure phases (preictal, ictal, and postictal) and the interictal phase were used to identify/classify seizure types and their phases. A support vector machine (SVM) working on a multidimensional feature space of network and spectral measures was implemented for the classification of each seizure type; a traditional statistical approach was also conducted to highlight the underlying patterns to each seizure type or phase.

RESULTS

The percentage of correct classification of seizure types, corrected by chance, provided by the SVM exceeded 70%, considering all measures and the entire seizure (preictal + ictal + postictal). This percentage increased to more than 80% when all the measures during the ictal period for the depth electrodes or during the postictal for subdural electrodes were considered. Regarding the statistical approach, several measures presented a monotonic ascending and descending behavior with respect to seizure severity; these changes were observed during the ictal and postictal periods. Some measures were specific of each seizure type.

CONCLUSIONS

Our results provide a new framework to seizure characterization and reveal the possibility of an exclusively intracranial EEG-based classification. This could be used to build an automatic seizure classification system and provides new evidence of the network-related physiopathology of epilepsies. Thus, the novelty of this work is the possibility of differentiating seizure types based exclusively on the EEG recordings, providing evidence of the underlying patterns or characteristics to each seizure type and/or phase that would allow their optimal classification.

Engineering nonlinear epileptic biomarkers using deep learning and Benford's law

In this study, we designed two deep neural networks to encode 16 features for early seizure detection in intracranial EEG and compared them and their frequency responses to 16 widely used engineered metrics to interpret their properties: epileptogenicity index (EI), phase locked high gamma (PLHG), time and frequency domain Cho Gaines distance (TDCG, FDCG), relative band powers, and log absolute band powers (from alpha, beta, theta, delta, low gamma, and high gamma bands). The deep learning models were pretrained for seizure identification on the time and frequency domains of 1 s, single-channel clips of 127 seizures (from 25 different subjects) using "leave-one-out" (LOO) cross validation. Each neural network extracted unique feature spaces that were interpreted using spectral power modulations before being used to train a Random Forest Classifier (RFC) for seizure identification. The Gini Importance of each feature was calculated from the pretrained RFC, enabling the most significant features (MSFs) for each task to be identified. The MSFs were extracted to train another RFC for UPenn and Mayo Clinic's Seizure Detection Kaggle Challenge. They obtained an AUC score of 0.93, demonstrating a transferable method to identify and interpret biomarkers for seizure detection.

Interpretable EEG seizure prediction using a multiobjective evolutionary algorithm

Seizure prediction might be the solution to tackle the apparent unpredictability of seizures in patients with drug-resistant epilepsy, which comprise about a third of all patients with epilepsy. Designing seizure prediction models involves defining the pre-ictal period, a transition stage between inter-ictal brain activity and the seizure discharge. This period is typically a fixed interval, with some recent studies reporting the evaluation of different patient-specific pre-ictal intervals. Recently, researchers have aimed to determine the pre-ictal period, a transition stage between regular brain activity and a seizure. Authors have been using deep learning models given the ability of such models to automatically perform pre-processing, feature extraction, classification, and handling temporal and spatial dependencies. As these approaches create black-box models, clinicians may not have sufficient trust to use them in high-stake decisions. By considering these problems, we developed an evolutionary seizure prediction model that identifies the best set of features while automatically searching for the pre-ictal period and accounting for patient comfort. This methodology provides patient-specific interpretable insights, which might contribute to a better understanding of seizure generation processes and explain the algorithm’s decisions. We tested our methodology on 238 seizures and 3687 h of continuous data, recorded on scalp recordings from 93 patients with several types of focal and generalised epilepsies. We compared the results with a seizure surrogate predictor and obtained a performance above chance for 32% patients. We also compared our results with a control method based on the standard machine learning pipeline (pre-processing, feature extraction, classifier training, and post-processing), where the control marginally outperformed our approach by validating 35% of the patients. In total, 54 patients performed above chance for at least one method: our methodology or the control one. Of these 54 patients, 21 (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\approx$$\end{document}≈38%) were solely validated by our methodology, while 24 (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\approx$$\end{document}≈44%) were only validated by the control method. These findings may evidence the need for different methodologies concerning different patients.

A multimodal AI system for out-of-distribution generalization of seizure identification

Artificial intelligence (AI) and health sensory data-fusion hold the potential to automate many laborious and time-consuming processes in hospitals or ambulatory settings, e.g. home monitoring and telehealth. One such unmet challenge is rapid and accurate epileptic seizure annotation. An accurate and automatic approach can provide an alternative way to label seizures in epilepsy or deliver a substitute for inaccurate patient self-reports. Multimodal sensory fusion is believed to provide an avenue to improve the performance of AI systems in seizure identification. We propose a state-of-the-art performing AI system that combines electroencephalogram (EEG) and electrocardiogram (ECG) for seizure identification, tested on clinical data with early evidence demonstrating generalization across hospitals. The model was trained and validated on the publicly available Temple University Hospital (TUH) dataset. To evaluate performance in a clinical setting, we conducted non-patient-specific pseudo-prospective inference tests on three out-of-distribution datasets, including EPILEPSIAE (30 patients) and the Royal Prince Alfred Hospital (RPAH) in Sydney, Australia (31 neurologists-shortlisted patients and 30 randomly selected). Our multimodal approach improves the area under the receiver operating characteristic curve (AUC-ROC) by an average margin of 6.71% and 14.42% for deep learning techniques using EEG-only and ECG-only, respectively. Our model's state-of-the-art performance and robustness to out-of-distribution datasets show the accuracy and efficiency necessary to improve epilepsy diagnoses. To the best of our knowledge, this is the first pseudo-prospective study of an AI system combining EEG and ECG modalities for automatic seizure annotation achieved with fusion of two deep learning networks.

A Comparison of Energy-Efficient Seizure Detectors for Implantable Neurostimulation Devices

Introduction

About 30% of epilepsy patients are resistant to treatment with antiepileptic drugs, and only a minority of these are surgical candidates. A recent therapeutic approach is the application of electrical stimulation in the early phases of a seizure to interrupt its spread across the brain. To accomplish this, energy-efficient seizure detectors are required that are able to detect a seizure in its early stages.

Methods

Three patient-specific, energy-efficient seizure detectors are proposed in this study: (i) random forest (RF); (ii) long short-term memory (LSTM) recurrent neural network (RNN); and (iii) convolutional neural network (CNN). Performance evaluation was based on EEG data (n = 40 patients) derived from a selected set of surface EEG electrodes, which mimic the electrode layout of an implantable neurostimulation system. As for the RF input, 16 features in the time- and frequency-domains were selected. Raw EEG data were used for both CNN and RNN. Energy consumption was estimated by a platform-independent model based on the number of arithmetic operations (AOs) and memory accesses (MAs). To validate the estimated energy consumption, the RNN classifier was implemented on an ultra-low-power microcontroller.

Results

The RNN seizure detector achieved a slightly better level of performance, with a median area under the precision-recall curve score of 0.49, compared to 0.47 for CNN and 0.46 for RF. In terms of energy consumption, RF was the most efficient algorithm, with a total of 67k AOs and 67k MAs per classification. This was followed by CNN (488k AOs and 963k MAs) and RNN (772k AOs and 978k MAs), whereby MAs contributed more to total energy consumption. Measurements derived from the hardware implementation of the RNN algorithm demonstrated a significant correlation between estimations and actual measurements.

Discussion

All three proposed seizure detection algorithms were shown to be suitable for application in implantable devices. The applied methodology for a platform-independent energy estimation was proven to be accurate by way of hardware implementation of the RNN algorithm. These findings show that seizure detection can be achieved using just a few channels with limited spatial distribution. The methodology proposed in this study can therefore be applied when designing new models for responsive neurostimulation.

Ensemble Deep Neural Network for Automatic Classification of EEG Independent Components

OBJECTIVE

Independent component analysis (ICA) is commonly used to remove noisy artifacts from multi-channel scalp electroencephalogram (EEG) signals. ICA decomposes EEG into different independent components (ICs) and then, experts remove the noisy ones. This process is highly time-consuming and experts are not always available. To surpass this drawback, research is going on to develop models to automatically conduct IC classification. Current state-of-the-art models use power spectrum densities (PSDs) and topoplots to classify ICs. The performance of these methods may be limited by disregarding the IC time-series that would contribute to fully simulate the visual inspection performed by experts.

METHODS

We present a novel ensemble deep neural network that combines time-series, PSDs, and topoplots to classify ICs. Moreover, we study the ability to use our model in transfer learning approaches.

RESULTS

Experimental results showed that using time-series improves IC classification. Results also indicated that transfer learning obtained higher performance than simply training a new model from scratch.

CONCLUSION

Researchers should develop IC classifiers using the three sources of information. Moreover, transfer learning approaches should be considered when producing new deep learning models.

SIGNIFICANCE

This work improves IC classification, enhancing the automatic removal of EEG artifacts. Additionally, since labelled ICs are generally not publicly available, the possibility of using our model in transfer learning studies may motivate other researchers to develop their own classifiers.

An Overview of EEG-based Machine Learning Methods in Seizure Prediction and Opportunities for Neurologists in this Field

The unpredictability of epileptic seizures is one of the most problematic aspects of the field of epilepsy. Methods or devices capable of detecting seizures minutes before they occur may help prevent injury or even death and significantly improve the quality of life. Machine learning (ML) is an emerging technology that can markedly enhance algorithm performance by interpreting data. ML has gained increasing attention from medical researchers in recent years. Its epilepsy applications range from the localization of the epileptic region, predicting the medical or surgical outcome of epilepsy, and automated electroencephalography (EEG) analysis to seizure prediction. While ML has good prospects with regard to detecting epileptic seizures via EEG signals, many clinicians are still unfamiliar with this field. This work briefly summarizes the history and recent significant progress made in this field and clarifies the essential components of the automatic seizure detection system using ML methodologies for clinicians. This review also proposes how neurologists can actively contribute to ensure improvements in seizure prediction using EEG-based ML.

Seizure Susceptibility Prediction in Uncontrolled Epilepsy

Epileptic seizure forecasting, combined with the delivery of preventative therapies, holds the potential to greatly improve the quality of life for epilepsy patients and their caregivers. Forecasting seizures could prevent some potentially catastrophic consequences such as injury and death in addition to several potential clinical benefits it may provide for patient care in hospitals. The challenge of seizure forecasting lies within the seemingly unpredictable transitions of brain dynamics into the ictal state. The main body of computational research on determining seizure risk has been focused solely on prediction algorithms, which involves a challenging issue of balancing sensitivity and false alarms. There have been some studies on identifying potential biomarkers for seizure forecasting; however, the questions of “What are the true biomarkers for seizure prediction” or even “Is there a valid biomarker for seizure prediction?” are yet to be fully answered. In this paper, we introduce a tool to facilitate the exploration of the potential biomarkers. We confirm using our tool that interictal slowing activities are a promising biomarker for epileptic seizure susceptibility prediction.

Wavelet Ridges in EEG Diagnostic Features Extraction: Epilepsy Long-Time Monitoring and Rehabilitation after Traumatic Brain Injury

Interchannel EEG synchronization, as well as its violation, is an important diagnostic sign of a number of diseases. In particular, during an epileptic seizure, such synchronization occurs starting from some pairs of channels up to many pairs in a generalized seizure. Additionally, for example, after traumatic brain injury, the destruction of interneuronal connections occurs, which leads to a violation of interchannel synchronization when performing motor or cognitive tests. Within the framework of a unified approach to the analysis of interchannel EEG synchronization using the ridges of wavelet spectra, two problems were solved. First, the segmentation of the initial data of long-term monitoring of scalp EEG with various artifacts into fragments suspicious of epileptic seizures in order to reduce the total duration of the fragments analyzed by the doctor. Second, assessments of recovery after rehabilitation of cognitive functions in patients with moderate traumatic brain injury. In the first task, the initial EEG was segmented into fragments in which at least two channels were synchronized, and by the adaptive threshold method into fragments with a high value of the EEG power spectral density. Overlapping in time synchronized fragments with fragments of high spectral power density was determined. As a result, the total duration of the fragments for analysis by the doctor was reduced by more than 60 times. In the second task, the network of phase-related EEG channels was determined during the cognitive test before and after rehabilitation. Calculation-logical and spatial-pattern cognitive tests were used. The positive dynamics of rehabilitation was determined during the initialization of interhemispheric connections and connections in the frontal cortex of the brain.

Forecasting Seizure Likelihood With Wearable Technology

The unpredictability of epileptic seizures exposes people with epilepsy to potential physical harm, restricts day-to-day activities, and impacts mental well-being. Accurate seizure forecasters would reduce the uncertainty associated with seizures but need to be feasible and accessible in the long-term. Wearable devices are perfect candidates to develop non-invasive, accessible forecasts but are yet to be investigated in long-term studies. We hypothesized that machine learning models could utilize heart rate as a biomarker for well-established cycles of seizures and epileptic activity, in addition to other wearable signals, to forecast high and low risk seizure periods. This feasibility study tracked participants' (n = 11) heart rates, sleep, and step counts using wearable smartwatches and seizure occurrence using smartphone seizure diaries for at least 6 months (mean = 14.6 months, SD = 3.8 months). Eligible participants had a diagnosis of refractory epilepsy and reported at least 20 seizures (mean = 135, SD = 123) during the recording period. An ensembled machine learning and neural network model estimated seizure risk either daily or hourly, with retraining occurring on a weekly basis as additional data was collected. Performance was evaluated retrospectively against a rate-matched random forecast using the area under the receiver operating curve. A pseudo-prospective evaluation was also conducted on a held-out dataset. Of the 11 participants, seizures were predicted above chance in all (100%) participants using an hourly forecast and in ten (91%) participants using a daily forecast. The average time spent in high risk (prediction time) before a seizure occurred was 37 min in the hourly forecast and 3 days in the daily forecast. Cyclic features added the most predictive value to the forecasts, particularly circadian and multiday heart rate cycles. Wearable devices can be used to produce patient-specific seizure forecasts, particularly when biomarkers of seizure and epileptic activity cycles are utilized.

The power of ECG in multimodal patient-specific seizure monitoring: Added value to an EEG-based detector using limited channels

Objective

Wearable seizure detection devices could provide more reliable seizure documentation outside the hospital compared to seizure self‐reporting by patients, which is the current standard. Previously, during the SeizeIT1 project, we studied seizure detection based on behind‐the‐ear electroencephalography (EEG). However, the obtained sensitivities were too low for practical use, because not all seizures are associated with typical ictal EEG patterns. Therefore, in this paper, we aim to develop a multimodal automated seizure detection algorithm integrating behind‐the‐ear EEG and electrocardiography (ECG) for detecting focal seizures. In this framework, we quantified the added value of ECG to behind‐the‐ear EEG.

Methods

This study analyzed three multicenter databases consisting of 135 patients having focal epilepsy and a total of 896 seizures. A patient‐specific multimodal automated seizure detection algorithm was developed using behind‐the‐ear/temporal EEG and single‐lead ECG. The EEG and ECG data were processed separately using machine learning methods. A late integration approach was applied for fusing those predictions.

Results

The multimodal algorithm outperformed the EEG‐based algorithm in two of three databases, with an increase of 11% and 8% in sensitivity for the same false alarm rate.

Significance

ECG can be of added value to an EEG‐based seizure detection algorithm using only behind‐the‐ear/temporal lobe electrodes for patients with focal epilepsy.

Prediction of Seizure Recurrence. A Note of Caution

Great strides have been made recently in documenting that machine-learning programs can predict seizure occurrence in people who have epilepsy. Along with this progress have come claims that appear to us to be a bit premature. We anticipate that many people will benefit from seizure prediction. We also doubt that all will benefit. Although machine learning is a useful tool for aiding discovery, we believe that the greatest progress will come from deeper understanding of seizures, epilepsy, and the EEG features that enable seizure prediction. In this essay, we lay out reasons for optimism and skepticism.

Heart rate variability analysis for the identification of the preictal interval in patients with drug-resistant epilepsy

Electrocardiogram (ECG) recordings, lasting hours before epileptic seizures, have been studied in the search for evidence of the existence of a preictal interval that follows a normal ECG trace and precedes the seizure's clinical manifestation. The preictal interval has not yet been clinically parametrized. Furthermore, the duration of this interval varies for seizures both among patients and from the same patient. In this study, we performed a heart rate variability (HRV) analysis to investigate the discriminative power of the features of HRV in the identification of the preictal interval. HRV information extracted from the linear time and frequency domains as well as from nonlinear dynamics were analysed. We inspected data from 238 temporal lobe seizures recorded from 41 patients with drug-resistant epilepsy from the EPILEPSIAE database. Unsupervised methods were applied to the HRV feature dataset, thus leading to a new perspective in preictal interval characterization. Distinguishable preictal behaviour was exhibited by 41% of the seizures and 90% of the patients. Half of the preictal intervals were identified in the 40 min before seizure onset. The results demonstrate the potential of applying clustering methods to HRV features to deepen the current understanding of the preictal state.

Zero-crossing patterns reveal subtle epileptiform discharges in the scalp EEG

Clinical diagnosis of epilepsy depends heavily on the detection of interictal epileptiform discharges (IEDs) from scalp electroencephalographic (EEG) signals, which by purely visual means is far from straightforward. Here, we introduce a simple signal analysis procedure based on scalp EEG zero-crossing patterns which can extract the spatiotemporal structure of scalp voltage fluctuations. We analyzed simultaneous scalp and intracranial EEG recordings from patients with pharmacoresistant temporal lobe epilepsy. Our data show that a large proportion of intracranial IEDs manifest only as subtle, low-amplitude waveforms below scalp EEG background and could, therefore, not be detected visually. We found that scalp zero-crossing patterns allow detection of these intracranial IEDs on a single-trial level with millisecond temporal precision and including some mesial temporal discharges that do not propagate to the neocortex. Applied to an independent dataset, our method discriminated accurately between patients with epilepsy and normal subjects, confirming its practical applicability.

A personalized and evolutionary algorithm for interpretable EEG epilepsy seizure prediction

Seizure prediction may improve the quality of life of patients suffering from drug-resistant epilepsy, which accounts for about 30% of the total epileptic patients. The pre-ictal period determination, characterized by a transitional stage between normal brain activity and seizure, is a critical step. Past approaches failed to attain real-world applicability due to lack of generalization capacity. More recently, deep learning techniques may outperform traditional classifiers and handle time dependencies. However, despite the existing efforts for providing interpretable insights, clinicians may not be willing to make high-stake decisions based on them. Furthermore, a disadvantageous aspect of the more usual seizure prediction pipeline is its modularity and significant independence between stages. An alternative could be the construction of a search algorithm that, while considering pipeline stages' synergy, fine-tunes the selection of a reduced set of features that are widely used in the literature and computationally efficient. With extracranial recordings from 19 patients suffering from temporal-lobe seizures, we developed a patient-specific evolutionary optimization strategy, aiming to generate the optimal set of features for seizure prediction with a logistic regression classifier, which was tested prospectively in a total of 49 seizures and 710 h of continuous recording and performed above chance for 32% of patients, using a surrogate predictor. These results demonstrate the hypothesis of pre-ictal period identification without the loss of interpretability, which may help understanding brain dynamics leading to seizures and improve prediction algorithms.

Automatic seizure detection based on imaged-EEG signals through fully convolutional networks

Seizure detection is a routine process in epilepsy units requiring manual intervention of well-trained specialists. This process could be extensive, inefficient and time-consuming, especially for long term recordings. We proposed an automatic method to detect epileptic seizures using an imaged-EEG representation of brain signals. To accomplish this, we analyzed EEG signals from two different datasets: the CHB-MIT Scalp EEG database and the EPILEPSIAE project that includes scalp and intracranial recordings. We used fully convolutional neural networks to automatically detect seizures. For our best model, we reached average accuracy and specificity values of 99.3% and 99.6%, respectively, for the CHB-MIT dataset, and corresponding values of 98.0% and 98.3% for the EPILEPSIAE patients. For these patients, the inclusion of intracranial electrodes together with scalp ones increased the average accuracy and specificity values to 99.6% and 58.3%, respectively. Regarding the other metrics, our best model reached average precision of 62.7%, recall of 58.3%, F-measure of 59.0% and AP of 54.5% on the CHB-MIT recordings, and comparatively lowers performances for the EPILEPSIAE dataset. For both databases, the number of false alarms per hour reached values less than 0.5/h for 92% of the CHB-MIT patients and less than 1.0/h for 80% of the EPILEPSIAE patients. Compared to recent studies, our lightweight approach does not need any estimation of pre-selected features and demonstrates high performances with promising possibilities for the introduction of such automatic methods in the clinical practice.

From Seizure Detection to Smart and Fully Embedded Seizure Prediction Engine: A Review

Recent review papers have investigated seizure prediction, creating the possibility of preempting epileptic seizures. Correct seizure prediction can significantly improve the standard of living for the majority of epileptic patients, as the unpredictability of seizures is a major concern for them. Today, the development of algorithms, particularly in the field of machine learning, enables reliable and accurate seizure prediction using desktop computers. However, despite extensive research effort being devoted to developing seizure detection integrated circuits (ICs), dedicated seizure prediction ICs have not been developed yet. We believe that interdisciplinary study of system architecture, analog and digital ICs, and machine learning algorithms can promote the translation of scientific theory to a more realistic intelligent, integrated, and low-power system that can truly improve the standard of living for epileptic patients. This review explores topics ranging from signal acquisition analog circuits to classification algorithms and dedicated digital signal processing circuits for detection and prediction purposes, to provide a comprehensive and useful guideline for the construction, implementation and optimization of wearable and integrated smart seizure prediction systems.

Big data in epilepsy: Clinical and research considerations. Report from the Epilepsy Big Data Task Force of the International League Against Epilepsy

Epilepsy is a heterogeneous condition with disparate etiologies and phenotypic and genotypic characteristics. Clinical and research aspects are accordingly varied, ranging from epidemiological to molecular, spanning clinical trials and outcomes, gene and drug discovery, imaging, electroencephalography, pathology, epilepsy surgery, digital technologies, and numerous others. Epilepsy data are collected in the terabytes and petabytes, pushing the limits of current capabilities. Modern computing firepower and advances in machine and deep learning, pioneered in other diseases, open up exciting possibilities for epilepsy too. However, without carefully designed approaches to acquiring, standardizing, curating, and making available such data, there is a risk of failure. Thus, careful construction of relevant ontologies, with intimate stakeholder inputs, provides the requisite scaffolding for more ambitious big data undertakings, such as an epilepsy data commons. In this review, we assess the clinical and research epilepsy landscapes in the big data arena, current challenges, and future directions, and make the case for a systematic approach to epilepsy big data.

Machine learning and wearable devices of the future

Machine learning (ML) is increasingly recognized as a useful tool in healthcare applications, including epilepsy. One of the most important applications of ML in epilepsy is seizure detection and prediction, using wearable devices (WDs). However, not all currently available algorithms implemented in WDs are using ML. In this review, we summarize the state of the art of using WDs and ML in epilepsy, and we outline future development in these domains. There is published evidence for reliable detection of epileptic seizures using implanted electroencephalography (EEG) electrodes and wearable, non-EEG devices. Application of ML using the data recorded with WDs from a large number of patients could change radically the way we diagnose and manage patients with epilepsy.

Comparing Different Methods for Named Entity Recognition in Portuguese Neurology Text

Electronic Medical Records (EMRs) are written in an unstructured way, often using natural language. Information Extraction (IE) may be used for acquiring knowledge from such texts, including the automatic recognition of meaningful entities, through models for Named Entity Recognition (NER). However, while most work on the previous was made for English, this experience aimed at testing different methods in Portuguese text, more precisely, on the domain of Neurology, and take some conclusions. This paper comprised the comparison between Conditional Random Fields (CRF), bidirectional Long Short-term Memory - Conditional Random Fields (BiLSTM-CRF) and a BiLSTM-CRF with residual learning connections, using not only Portuguese texts from medical journals but also texts from the Coimbra Hospital and Universitary Centre (CHUC) Neurology Service. Furthermore, the performances of BiLSTM-CRF models using word embeddings (WEs) trained with clinical text and WEs trained with general language texts were compared. Deep learning models achieved F1-Scores of nearly 83% and 75%, respectively for relaxed and strict evaluation, on texts extracted from the medical journal. For texts collected from the Hospital, the same achieved F1-Scores of nearly 71% and 62%. This work concludes that deep learning models outperform the shallow learning models and that in-domain WEs get better results than general language WEs, even when the latter are trained with much more text than the former. Furthermore, the results show that it is possible to extract information from Hospital clinical texts with models trained with clinical cases extracted from medical journals, and thus openly available. Nevertheless, such results still require a healthcare technician to check if the information is well extracted.

Preictal Time Assessment using Heart Rate Variability Features in Drug-resistant Epilepsy Patients

Epileptic seizures are by default associated with the occurrence of EEG changes as a consequence of alterations in brain activity seconds to hours before the seizure onset. Additionally, the influence of the autonomic nervous system may also be reflected in the electrocardiogram (ECG) trace. In both biosignals (EEG and ECG), differences have been reported between normal epochs, known as interictal periods and the interval preceding seizures, preictal period. However, the existence of a preictal state indicating the transition between epileptic brain states has not yet been clinically defined. In fact, some studies report differences in preictal location and duration among patients and among seizures occurring in the same patient as well. Based on the above, a study was designed in order to investigate the existence of a preictal interval specific for each seizure, using heart rate variability (HRV) features. Time and frequency domain features (linear and non-linear) were extracted from ECG data acquired in 37 drug-resistant epilepsy patients, comprised in EPILEPSIAE database. A total of 209 seizures occurring in the temporal lobe were analysed. The existence of a transition period before the seizure onset was inspected using a linear discriminant analysis classifier. The overall best performance (88.04%±12.30% of accuracy) was obtained by combining RRMean, NN50 and SD₂ features when discriminating the 50 min of interictal located farthest from seizure onset from the nearest preictal 50 min.

A review on the pattern detection methods for epilepsy seizure detection from EEG signals

Over several years, research had been conducted for the detection of epileptic seizures to support an automatic diagnosis system to comfort the clinicians' encumbrance. In this regard, a number of research papers have been published for the identification of epileptic seizures. A thorough review of all these papers is required. So, an attempt has been made to review on the pattern detection methods for epilepsy seizure detection from EEG signals. More than 150 research papers have been discussed to determine the techniques for detecting epileptic seizures. Further, the literature review confirms that the pattern recognition techniques required to detect epileptic seizures varies across the electroencephalogram (EEG) datasets of different conditions. This is mostly owing to the fact that EEG detected under different conditions have different characteristics. This consecutively necessitates the identification of the pattern recognition technique to efficiently differentiate EEG epileptic data from the EEG data of various conditions.

[Mobile seizure monitoring in epilepsy patients]

Wearables are receiving much attention from both epilepsy patients and treating physicians, for monitoring of seizure frequency and warning of seizures. They are also of interest for the detection of seizure-associated risks of patients, for differential diagnosis of rare seizure types and prediction of seizure-prone periods. Accelerometry, electromyography (EMG), heart rate and further autonomic parameters are recorded to capture clinical seizure manifestations. Currently, a clinical use to document nocturnal motor seizures is feasible. In this review the available devices, data on the performance in the documentation of seizures, current options for clinical use and developments in data analysis are presented and critically discussed.

Hardware Implementation of a Performance and Energy-optimized Convolutional Neural Network for Seizure Detection

We present for the first time a μW-power convolutional neural network for seizure detection running on a low-power microcontroller. On a dataset of 22 patients a median sensitivity of 100% is achieved. With a false positive rate of 20.7 fp/h and a short detection delay of 3.4 s it is suitable for the application in an implantable closed-loop device.

Seizure self-prediction in a randomized controlled trial of stress management

OBJECTIVE

Using electronic diaries as part of a randomized controlled trial of stress reduction for epilepsy, we evaluated factors associated with successful seizure self-prediction.

METHODS

Adults with medication-resistant focal epilepsy were recruited from 3 centers and randomized to treatment with progressive muscle relaxation or control focused attention. An 8-week baseline was followed by 12 weeks of double-blind treatment. Twice daily, participants rated the likelihood of a seizure in the next 24 hours on a 5-point scale from very unlikely to almost certain, along with mood, premonitory symptoms, stress ratings, and seizure counts. We analyzed the association of mood, premonitory symptoms, stress, and circadian influences on seizure self-prediction.

RESULTS

Sixty-four participants completed the trial (3,126 seizures). Diary entry adherence was >82%. Participant self-prediction was associated with seizure occurrence at 6, 12, and 24 hours (p < 0.0001). Odds ratio (OR) of seizure prediction increased systematically with participants' prediction of seizure likelihood (p < 0.0001, all levels of prediction and all time intervals). For the 12-hour prediction window, median specificity for seizure prediction was 0.94 and negative predictive value 0.94; median sensitivity was 0.10 and positive predictive value 0.13. A subset of 13 participants (20% of sample) met criteria for good predictors (median OR for seizure prediction 5.25). Mood, stress, premonitory symptoms, seizure time, and randomized group were not associated with seizure occurrence.

CONCLUSION

In this prospective study, participants' prediction of a high probability of seizure was significantly associated with subsequent seizure occurrence within 24 hours. Future studies should focus on understanding factors that drive self-prediction.

CLINICALTRIALSGOV IDENTIFIER

NCT01444183.

Phase Synchronization Operator for On-Chip Brain Functional Connectivity Computation

This paper presents an integer-based digital processor for the calculation of phase synchronization between two neural signals. It is based on the measurement of time periods between two consecutive minima. The simplicity of the approach allows for the use of elementary digital blocks, such as registers, counters, and adders. The processor, fabricated in a 0.18- μm CMOS process, only occupies 0.05 mm ² and consumes 15 nW from a 0.5 V supply voltage at a signal input rate of 1024 S/s. These low-area and low-power features make the proposed processor a valuable computing element in closed-loop neural prosthesis for the treatment of neural disorders, such as epilepsy, or for assessing the patterns of correlated activity in neural assemblies through the evaluation of functional connectivity maps.