Interictal Epileptiform Discharge Detection Using Multi-Head Deep Convolutional Neural Network.

Interictal epileptiform discharges (IEDs) are intermittent electrophysiological events that occur in patients with epilepsy between seizures. Automated detection of IEDs helps clinician to identify cortical irritations and relations to seizure recurrence. It also reduces the necessity of visual inspection by physicians interpreting the EEG. This paper presents a novel deep learning-based approach that combines one-dimensional local binary pattern symbolization method with a regularized multi-head one-dimensional convolutional neural network to learn unique morphological patterns from different EEG sub-bands for IED detection. Experimentation using the Temple University Events corpus scalp EEG data shows promising performance, e.g. F1-score of 87.18%.

SEEG-based epileptic seizure network modeling and analysis for pre-surgery evaluation.

Stereo-electroencephalography is a minimally invasive technique for patients with refractory epilepsy pursuing surgery to reduce or control seizures. Electrodes are implanted based on pre-surgery evaluations and can collect deep brain activities for surgery decisions. This paper presents a methodology to analyze stereo-electroencephalography and assist clinicians by recommending the optimal surgical option and target areas for focal epilepsy patients. A seizure network (graph) model is proposed to characterize the spatial distribution and temporal changes of ictal events. The network nodes and edges correspond to specific epileptogenic regions and propagation/impact pathways (weighted by directed transfer function), respectively. We then employ a K-means clustering strategy to group nodes into a few clusters, from which the target surgical areas can be identified. Ten patients with different types of focal seizures were thoroughly analyzed. Promising consistency between results of our method's recommendations, clinical decisions and surgery outcomes were observed.

Seizure onset patterns predict outcome after stereo-electroencephalography-guided laser amygdalohippocampotomy.

OBJECTIVE: Stereotactic laser amygdalohippocampotomy (SLAH) is an appealing option for patients with temporal lobe epilepsy, who often require intracranial monitoring to confirm mesial temporal seizure onset. However, given limited spatial sampling, it is possible that stereotactic electroencephalography (stereo-EEG) may miss seizure onset elsewhere. We hypothesized that stereo-EEG seizure onset patterns (SOPs) may differentiate between primary onset and secondary spread and predict postoperative seizure control. In this study, we characterized the 2-year outcomes of patients who underwent single-fiber SLAH after stereo-EEG and evaluated whether stereo-EEG SOPs predict postoperative seizure freedom. METHODS: This retrospective five-center study included patients with or without mesial temporal sclerosis (MTS) who underwent stereo-EEG followed by single-fiber SLAH between August 2014 and January 2022. Patients with causative hippocampal lesions apart from MTS or for whom the SLAH was considered palliative were excluded. An SOP catalogue was developed based on literature review. The dominant pattern for each patient was used for survival analysis. The primary outcome was 2-year Engel I classification or recurrent seizures before then, stratified by SOP category. RESULTS: Fifty-eight patients were included, with a mean follow-up duration of 39 ± 12 months after SLAH. Overall 1-, 2-, and 3-year Engel I seizure freedom probability was 54%, 36%, and 33%, respectively. Patients with SOPs, including low-voltage fast activity or low-frequency repetitive spiking, had a 46% 2-year seizure freedom probability, compared to 0% for patients with alpha or theta frequency repetitive spiking or theta or delta frequency rhythmic slowing (log-rank test, p = .00015). SIGNIFICANCE: Patients who underwent SLAH after stereo-EEG had a low probability of seizure freedom at 2 years, but SOPs successfully predicted seizure recurrence in a subset of patients. This study provides proof of concept that SOPs distinguish between hippocampal seizure onset and spread and supports using SOPs to improve selection of SLAH candidates.

A randomized phase 2b efficacy study in patients with seizure episodes with a predictable pattern using Staccato® alprazolam for rapid seizure termination.

OBJECTIVE: Alprazolam administered via the Staccato® breath-actuated device is delivered into the deep lung for rapid systemic exposure and is a potential therapy for rapid epileptic seizure termination (REST). We conducted an inpatient study (ENGAGE-E-001 [NCT03478982]) in patients with stereotypic seizure episodes with prolonged or repetitive seizures to determine whether Staccato alprazolam rapidly terminates seizures in a small observed population after administration under direct supervision. METHODS: Adult patients with established diagnosis of focal and/or generalized epilepsy with a documented history of seizure episodes with a predictable pattern were enrolled. They were randomized 1:1:1 to double-blind treatment of a single seizure event with one dose of Staccato alprazolam 1.0 mg or 2.0 mg, or Staccato placebo in an inpatient unit. The primary end point of the study was the proportion of responders in each treatment group achieving seizure activity cessation within 2 min after administration of study drug and no recurrence of seizure activity within 2 h. RESULTS: A total of 273 patients were screened, and 116 randomized patients received treatment with the study drug in the double-blind part. The proportion of treated patients who were responders was 65.8% for each of Staccato alprazolam 1.0 mg (n = 38; p = .0392) and 2.0 mg (n = 38; p = .0392), compared with 42.5% for Staccato placebo (n = 40). Staccato alprazolam was well tolerated when administered as a single dose of 1.0 or 2.0 mg: cough and somnolence were the most common adverse events (AEs) (both 14.5%), followed by dysgeusia (13.2%). AEs were mostly mild or moderate in intensity; there were no treatment-related serious AEs. SIGNIFICANCE: Both 1.0 mg and 2.0 mg doses of Staccato alprazolam demonstrated efficacy in rapidly terminating seizures in an inpatient setting and were well tolerated. The next step is a Phase 3 confirmatory study to demonstrate efficacy and safety of Staccato alprazolam for rapid cessation of seizures in an outpatient setting.

Single center outcomes of intracranial evaluation and surgical intervention in the elderly population.

Objective: The object of this case series is to report the effectiveness and complication rates of presurgical evaluation and surgical treatment among elderly epilepsy patients in our clinic. Methods: We reviewed patients charts from 2016 to 2020 and identified patients over the age of 55 years of age who underwent intracranial EEG, resection, and device placement. We compared the complications and post-intervention outcomes of 14 different patients. Results: The mean age of patients was 63.6 ± 4.13 years, and 6 (42.9 %) patients were female. 8 (57.1 %) patients underwent intracranial evaluation; 6 patients underwent SEEG, and 2 patients underwent SDE placement. 5/11 (45 %) patients reached seizure freedom with at least one year follow-up. Discussion: Intracranial evaluation could play an important role in the success of surgical intervention in the elderly epilepsy population. Prospective, multicenter studies are needed to determine the ideal candidate for a safe and effective intracranial evaluation and resection.

Emergency department visits and readmissions in patients with psychogenic nonepileptic seizures (PNES) at a safety net hospital.

INTRODUCTION: Readmissions and emergency department (ED) visits after an index admission have been become a quality measure due to associations with poor outcomes and increased healthcare costs. Readmissions and ED encounters have been studied in a variety of conditions including epilepsy but have not been examined exclusively in psychogenic nonepileptic seizures (PNES). In this study we examined the rate of readmissions and ED visits after a discharge from an Epilepsy Monitoring Unit (EMU) in a safety net hospital. We also determined patient phenotypes that are associated with readmissions. MATERIAL AND METHODS: This was a retrospective chart review study with index admission being a discharge from an EMU between January 1 and December 31 2016 with follow-up until August 31 2020. We obtained data regarding demographics, medical and psychiatric history, and social history and treatment interventions. Our outcome variables were both all-cause and seizure-related hospital readmissions and ED visits 30 days following the index discharge and readmissions and ED visits 30 days thereafter. RESULTS: Eleven of 122 patients (9%) had a non-seizure-related ED visit and/or hospitalization within 30 days of index discharge while 45 (37%) had re-contact with the health system thereafter for non-seizure-related issues. Seven of 122 patients (6%) had a seizure-related ED visit or hospital readmission within 30 days of discharge. Twenty-eight (23%) had a seizure-related readmission or ED visit after 30 days. Of these 28, 4 patients had been to an ER within 7 days of EMU discharge. The majority of subsequent encounters with the healthcare system were through the ED (n = 38) as compared to hospital (n = 10) and EMU readmissions (n = 9). On bivariate statistical analysis, charity or self-pay insurance status (p < 0.01), homelessness (p < 0.01), emergent EMU admission on index admission (p < 0.01), history of a psychiatric diagnosis (p < 0.02), and ED encounters 12 months prior to admission (p < 0.01) were significantly associated with readmission; however, on multivariate analysis only charity insurance status was a significant predictor. CONCLUSIONS: In this study of readmissions and ED visits after discharge with a diagnosis of PNES at a safety net hospital, we found a seizure-related readmission rate of approximately 6% in 30 days and 23% thereafter with the majority of re-contact with the hospital being in the ED. On multi-variate analysis insurance status was a significant factor associated with readmission and ED visits. Our future research directions include examining referrals and treatment completion at the hospital's PNES clinic as well as creating a risk score to better identify patients with PNES at risk of readmission.

Beyond the Wada: An updated approach to pre-surgical language and memory testing: An updated review of available evaluation techniques and recommended workflow to limit Wada test use to essential clinical cases.

The Intracarotid amobarbital test (IAT), also called Wada test, is considered the "gold standard" for lateralizing language dominance in the pre-surgical evaluation of patients with epilepsy. In addition, it has been further modified to assess the postoperative risk of amnesia in patients undergoing temporal lobectomy. Since then it has been utilized to lateralize language and assess pre-surgical memory function. Over the years, its popularity has declined due to several limitations and availability of alternative procedures like fMRI and MEG. A survey of its use in the pre-surgical evaluation for epilepsy surgery has not been performed since the 2008 international survey by Baxendale et al. and it was heavily skewed due to data from European and North American countries. Only approximately 12% of the epilepsy centers indicated that they used the Wada test in every patient to assess preoperative memory function and language lateralization before temporal lobectomy. Nowadays, we have many functional mapping tools at our disposal. It has become somewhat unsuitable to have epilepsy patients undergo an invasive test such as the Wada test for the risks associated with it outweigh the benefits. Our objective is to review the Wada Test and alternative methods of assessing language and memory dominance, as it is past its prime and should only be used in specific circumstances.

Emergent Admissions to the Epilepsy Monitoring Unit in the Setting of COVID-19 Pandemic-related, State-mandated Restrictions: Clinical Decision Making and Outcomes.

Due to the coronavirus disease 2019 (COVID-19) pandemic, the state of Texas-limited elective procedures to conserve beds and personal protective equipment (PPE); therefore, between March 22 and May 18, 2020, admission to the epilepsy monitoring unit (EMU) was limited only to urgent and emergent cases. We evaluated clinical characteristics and outcomes of these patients who were admitted to the EMU. Nineteen patients were admitted (one patient twice) with average age of 36.26 years (11 female) and average length of stay 3 days (range: 2-9 days). At least one event was captured on continuous EEG (cEEG) and video monitoring in all 20 admissions (atypical in one). One patient had both epileptic (ES) and psychogenic non-epileptic seizures (PNES) while 10 had PNES and 9 had ES. In 8 of 9 patients with ES, medications were changed, while in 5 patients with PNES, anti-epileptic drugs (AED) were stopped; the remaining 5 were not on medications. Of the 14 patients who had seen an epileptologist pre-admission, 13 (or 93%) had their diagnosis confirmed by EMU stay; a statistically significant finding. While typically an elective admission, in the setting of the COVID-19 pandemic, urgent and emergent EMU admissions were required for increased seizure or event frequency. In the vast majority of patients (13 of 19), admission lead to medication changes to either better control seizures or to change therapeutics as appropriate when PNES was identified.

Non-lesional mesial temporal lobe epilepsy requires bilateral invasive evaluation.

PURPOSE: Mesial temporal lobe epilepsy (MTLE) usually responds well to surgical treatment, although in non-lesional cases up to 50% of patients experience seizure relapse. The possibility of bilateral independent seizure onset should be considered as a reason for epilepsy surgery failure. METHODS: In a cohort of 177 patients who underwent invasive presurgical evaluation with stereo-tactically placed electrodes in two level four epilepsy centers, 29 had non-lesional MTLE. Invasive evaluation results are described. RESULTS: Among 29 patients with non-lesional MRI and mesial temporal lobe seizure onset recorded during stereo-EEG (SEEG) evaluation, four patients with unilateral preimplantation hypothesis had independent bilateral mesial temporal seizures on SEEG despite of unilateral non-invasive evaluation data. Three of these patients were treated with bitemporal responsive neurostimulator system (RNS). Independent bilateral mesial temporal seizures have been confirmed on RNS ECoG (electrocorticography). The fourth patient underwent right anterior temporal lobectomy. CONCLUSION: We propose that patients with non-lesional mesial temporal lobe epilepsy would benefit from bilateral invasive evaluation of mesial temporal structures to predict those patients who would be at most risk for surgical failure. Neurostimulaiton could be an initial treatment option for patients with independent bitemporal seizure onset.

Personalized EEG Feature Selection for Low-Complexity Seizure Monitoring.

Approximately, one third of patients with epilepsy are refractory to medical therapy and thus can be at high risk of injuries and sudden unexpected death. A low-complexity electroencephalography (EEG)-based seizure monitoring algorithm is critically important for daily use, especially for wearable monitoring platforms. This paper presents a personalized EEG feature selection approach, which is the key to achieve a reliable seizure monitoring with a low computational cost. We advocate a two-step, personalized feature selection strategy to enhance monitoring performances for each patient. In the first step, linear discriminant analysis (LDA) is applied to find a few seizure-indicative channels. Then in the second step, least absolute shrinkage and selection operator (LASSO) method is employed to select a discriminative subset of both frequency and time domain features (spectral powers and entropy). A personalization strategy is further customized to find the best settings (number of channels and features) that yield the highest classification scores for each subject. Experimental results of analyzing 23 subjects in CHB-MIT database are quite promising. We have achieved an average F-1 score of 88% with excellent sensitivity and specificity using not more than 7 features extracted from at most 3 channels.

Feature Selection Using F-statistic Values for EEG Signal Analysis.

Electroencephalography (EEG) is a highly complex and non-stationary signal that reflects the cortical electric activity. Feature selection and analysis of EEG for various purposes, such as epileptic seizure detection, are highly in demand. This paper presents an approach to enhance classification performance by selecting discriminative features from a combined feature set consisting of frequency domain and entropy based features. For each EEG channel, nine different features are extracted, including six sub-band spectral powers and three entropy values (sample, permutation and spectral entropy). Features are then ranked across all channels using F-statistic values and selected for SVM classification. Experimentation using CHB-MIT dataset shows that our method achieves average sensitivity, specificity and F-1 score of 92.63%, 99.72% and 91.21%, respectively.

Comparison of psychiatric comorbidities and impact on quality of life in patients with epilepsy or psychogenic nonepileptic spells.

OBJECTIVES: Psychiatric comorbidity is common in people with epilepsy (PWE) and psychogenic nonepileptic spells (PNES). These comorbidities can be detrimental to quality of life (QOL) and are often underdiagnosed and undertreated. Some types of epilepsy, such as focal temporal lobe epilepsy (TLE), have been associated with higher rates of psychiatric comorbidity. This study examined the impact of psychiatric comorbidity on QOL in patients admitted to two level 4 epilepsy monitoring units (EMUs). METHODS: In this prospective observational study, 200 patients admitted to two level 4 EMUs completed standardized surveys including the Quality of Life in Epilepsy (QOLIE-31-P), Generalized Anxiety Disorder 7-item (GAD-7), Patient Health Questionnaire (PHQ-9), and Beck Depression Inventory-II (BDI-II). Hierarchal multiple regression was performed to assess impact on QOL. RESULTS: Of the 200 participants, 113 had a diagnosis of epilepsy, 36 had a diagnosis of PNES, and 51 were excluded for nondiagnostic evaluation or dual diagnosis. Of those with epilepsy, 65 had TLE, 28 had focal extratemporal lobe epilepsy (ETLE), and 20 had nonfocal epilepsy. Patients with PNES had higher self-reported anxiety and depression levels (GAD-7: p = 0.04, PHQ-9: p < 0.01; BDI-II: p < 0.01) but similar QOL to PWE (p = 0.78). Using hierarchal multiple regression, symptoms of anxiety and depression were significant predictors of lower QOL in PWE but not in patients with PNES. There was no difference in QOL in those with ETLE and TLE. CONCLUSIONS: Our findings suggest that self-reported anxiety and depression symptoms are common in patients admitted to level 4 EMUs regardless of diagnosis and play an important role in predicting QOL in PWE. Our findings emphasize the importance of routinely screening all EMU patients for psychiatric comorbidity.

Provocative induction of psychogenic nonepileptic seizures: Noninferiority of an induction technique without versus with placebo.

We aim to demonstrate, in a sufficiently powered and standardized study, that the success rate of inducing psychogenic nonepileptic seizures (PNES) without placebo (saline infusion) is noninferior to induction with placebo. The clinical data of 170 consecutive patients with suspected PNES who underwent induction with placebo from January 21, 2009 to March 31, 2013 were pair-matched with 170 consecutive patients with suspected PNES who underwent the same induction technique but without addition of placebo from April 1, 2013 to February 7, 2018 at the same center. The success rates of induction were 79.4% (135/170) without placebo and 73.5% (125/170) with placebo. The difference of these two proportions was 5.9%, with two-sided 95% confidence interval ranging from -3.6% to 15.2%, indicating a non-statistically significant difference. The lower bound of the 95% confidence interval (-3.6%) was above the noninferiority margin (δ = -5%), hence inferring noninferiority of induction without versus with placebo. The greater cumulative induction experiences of the clinician performer (influencing the manner/presentation of induction) may supplant the potential advantage from addition of placebo (the means utilized). Among experienced performers, provocative induction without placebo should be the preferred diagnostic approach, given more ethically acceptable transparency and the noninferior success rate when compared to the same induction technique with placebo.

Propofol-ketamine combination therapy for effective control of super-refractory status epilepticus.

Retrospective analysis was conducted of patients with SRSE who were treated simultaneously with propofol and ketamine. Sixty-seven patients were identified from 2012 to 2015, and outcomes documented were resolution and mortality. The duration of combined ketamine and propofol use ranged from 1 to 28 days (mean - 3.6 days). Infusion rates ranged up to 145 and 175 mcg/kg/min. Vasopressors were used in 53 patients (79%), and were given within the first 5 days of the ICU admission in 48 (91%) patients. The overall SRSE resolution rate was 91%, and the overall mortality including patients with anoxic brain injury was 39%. Of the 13 patients with SRSE as a result of anoxic brain injury, SRSE was controlled in 5 (56%). The primary determinant of mortality was family withdrawing care related to the presence of severe medical/neurological diseases.

Pyridoxine deficiency in adult patients with status epilepticus.

An 8-year-old girl treated at our facility for superrefractory status epilepticus was found to have a low pyridoxine level at 5 µg/L. After starting pyridoxine supplementation, improvement in the EEG for a 24-hour period was seen. We decided to look at the pyridoxine levels in adult patients admitted with status epilepticus. We reviewed the records on patients admitted to the neurological ICU for status epilepticus (SE). Eighty-one adult patients were identified with documented pyridoxine levels. For comparison purposes, we looked at pyridoxine levels in outpatients with epilepsy (n=132). Reported normal pyridoxine range is >10 ng/mL. All but six patients admitted for SE had low normal or undetectable pyridoxine levels. A selective pyridoxine deficiency was seen in 94% of patients with status epilepticus (compared to 39.4% in the outpatients) which leads us to believe that there is a relationship between status epilepticus and pyridoxine levels.

Safety & pK of IV loading dose of lacosamide in the ICU.

A restrospective review of patients treated in the ICU for refractory status epilepticus who had received an initial IV loading dose of lacosamide (LCS) was performed. A total of 142 patients were identified. The first 34 patients received 400mg which by weight-based measurement ranged from 2 to 11 mg/kg. Higher mg/kg dosing had been used subsequently with doses up to 13 mg/kg. No patient required reduction in rate or cessation of infusion. Initiation of pressor agents was not needed during the infusion of the loading dose. Postinfusion LCS blood levels were drawn, and dosing of 10-12 mg/kg and higher resulted in blood levels above 15 µg/ml while doses of 2-6 mg/kg resulted in levels below 10 µg/ml. We conclude that a weight-based loading dose of 10-12 mg/kg at an infusion rate of 0.4 mg/kg/min is safe and will produce levels of 15 µg/ml and higher. This article is part of a Special Issue entitled "Status Epilepticus".