Spyglass: a framework for reproducible and shareable neuroscience research.

Scientific progress depends on reliable and reproducible results. Progress can also be accelerated when data are shared and re-analyzed to address new questions. Current approaches to storing and analyzing neural data typically involve bespoke formats and software that make replication, as well as the subsequent reuse of data, difficult if not impossible. To address these challenges, we created Spyglass, an open-source software framework that enables reproducible analyses and sharing of data and both intermediate and final results within and across labs. Spyglass uses the Neurodata Without Borders (NWB) standard and includes pipelines for several core analyses in neuroscience, including spectral filtering, spike sorting, pose tracking, and neural decoding. It can be easily extended to apply both existing and newly developed pipelines to datasets from multiple sources. We demonstrate these features in the context of a cross-laboratory replication by applying advanced state space decoding algorithms to publicly available data. New users can try out Spyglass on a Jupyter Hub hosted by HHMI and 2i2c: https://spyglass.hhmi.2i2c.cloud/.

Four State Sleep Staging From a Multilayered Algorithm Using Electrocardiographic and Actigraphic Data.

PURPOSE: Sleep studies are important to evaluate sleep and sleep-related disorders. The standard test for evaluating sleep is polysomnography, during which several physiological signals are recorded separately and simultaneously with specialized equipment that requires a technologist. Simpler recordings that can model the results of a polysomnography would provide the benefit of expanding the possibilities of sleep recordings. METHODS: Using the publicly available sleep data set from the multiethnic study of atherosclerosis and 1769 nights of sleep, we extracted a distinct data subset with engineered features of the biomarkers collected by actigraphic, oxygenation, and electrocardiographic sensors. We then applied scalable models with recurrent neural network and Extreme Gradient Boosting (XGBoost) with a layered approach to produce an algorithm that we then validated with a separate data set of 177 nights. RESULTS: The algorithm achieved an overall performance of 0.833 accuracy and 0.736 kappa in classifying into four states: wake, light sleep, deep sleep, and rapid eye movement (REM). Using feature analysis, we demonstrated that heart rate variability is the most salient feature, which is similar to prior reports. CONCLUSIONS: Our results demonstrate the potential benefit of a multilayered algorithm and achieved higher accuracy and kappa than previously described approaches for staging sleep. The results further the possibility of simple, wearable devices for sleep staging. Code is available at https://github.com/NovelaNeuro/nEureka-SleepStaging.

Spinal epidural arteriovenous fistula with nerve root enhancement mimicking myeloradiculitis: a case report.

BACKGROUND: Gadolinium enhancement of spinal nerve roots on magnetic resonance imaging (MRI) has rarely been reported in spinal dural arteriovenous fistula (SDAVF). Nerve root enhancement and cerebrospinal fluid (CSF) pleocytosis can be deceptive and lead to a misdiagnosis of myeloradiculitis. We report a patient who was initially diagnosed with neurosarcoid myeloradiculitis due to spinal nerve root enhancement, mildly inflammatory cerebrospinal fluid, and pulmonary granulomas, who ultimately was found to have an extensive symptomatic SDAVF. CASE PRESENTATION: A 52-year-old woman presented with a longitudinally extensive spinal cord lesion with associated gadolinium enhancement of the cord and cauda equina nerve roots, and mild lymphocytic pleocytosis. Pulmonary lymph node biopsy revealed non-caseating granulomas and neurosarcoid myeloradiculitis was suspected. She had rapid and profound clinical deterioration after a single dose of steroids. Further work-up with spinal angiography revealed a thoracic SDAVF, which was surgically ligated leading to clinical improvement. CONCLUSIONS: This case highlights an unexpected presentation of SDAVF with nerve root enhancement and concurrent pulmonary non-caseating granulomas, leading to an initial misdiagnosis with neurosarcoidosis. Nerve root enhancement has only rarely been described in cases of SDAVF; however, as this case highlights, it is an important consideration in the differential diagnosis of non-inflammatory causes of longitudinally extensive myeloradiculopathy with nerve root enhancement. This point is highly salient due to the importance of avoiding misdiagnosis of SDAVF, as interventions such as steroids or epidural injections used to treat inflammatory or infiltrative mimics may worsen symptoms in SDAVF. We review the presentation, diagnosis, and management of SDAVF as well as a proposed diagnostic approach to differentiating SDAVF from inflammatory myeloradiculitis.

Development of a core outcome set for quality of life for adults with drug-resistant epilepsy: A multistakeholder Delphi consensus study.

OBJECTIVE: In 2017, the American Academy of Neurology (AAN) convened the AAN Quality Measurement Set working group to define the improvement and maintenance of quality of life (QOL) as a key outcome measure in epilepsy clinical practice. A core outcome set (COS), defined as an accepted, standardized set of outcomes that should be minimally measured and reported in an area of health care research and practice, has not previously been defined for QOL in adult epilepsy. METHODS: A cross-sectional Delphi consensus study was employed to attain consensus from patients and caregivers on the QOL outcomes that should be minimally measured and reported in epilepsy clinical practice. Candidate items were compiled from QOL scales recommended by the AAN 2017 Quality Measurement Set. Inclusion criteria to participate in the Delphi study were adults with drug-resistant epilepsy diagnosed by a physician, no prior diagnosis of psychogenic nonepileptic seizures or a cognitive and/or developmental disability, or caregivers of patients meeting these criteria. RESULTS: A total of 109 people satisfied inclusion/exclusion criteria and took part in Delphi Round 1 (patients, n = 95, 87.2%; caregivers, n = 14, 12.8%), and 55 people from Round 1 completed Round 2 (patients, n = 43, 78.2%; caregivers, n = 12, 21.8%). One hundred three people took part in the final consensus round. Consensus was attained by patients/caregivers on a set of 36 outcomes that should minimally be included in the QOL COS. Of these, 32 of the 36 outcomes (88.8%) pertained to areas outside of seizure frequency and severity. SIGNIFICANCE: Using patient-centered Delphi methodology, this study defines the first COS for QOL measurement in clinical practice for adults with drug-resistant epilepsy. This set highlights the diversity of factors beyond seizure frequency and severity that impact QOL in epilepsy.

BAYESIAN NON-HOMOGENEOUS HIDDEN MARKOV MODEL WITH VARIABLE SELECTION FOR INVESTIGATING DRIVERS OF SEIZURE RISK CYCLING.

A major issue in the clinical management of epilepsy is the unpredictability of seizures. Yet, traditional approaches to seizure forecasting and risk assessment in epilepsy rely heavily on raw seizure frequencies, which are a stochastic measurement of seizure risk. We consider a Bayesian non-homogeneous hidden Markov model for unsupervised clustering of zero-inflated seizure count data. The proposed model allows for a probabilistic estimate of the sequence of seizure risk states at the individual level. It also offers significant improvement over prior approaches by incorporating a variable selection prior for the identification of clinical covariates that drive seizure risk changes and accommodating highly granular data. For inference, we implement an efficient sampler that employs stochastic search and data augmentation techniques. We evaluate model performance on simulated seizure count data. We then demonstrate the clinical utility of the proposed model by analyzing daily seizure count data from 133 patients with Dravet syndrome collected through the Seizure Tracker™ system, a patient-reported electronic seizure diary. We report on the dynamics of seizure risk cycling, including validation of several known pharmacologic relationships. We also uncover novel findings characterizing the presence and volatility of risk states in Dravet syndrome, which may directly inform counseling to reduce the unpredictability of seizures for patients with this devastating cause of epilepsy.

A Bayesian switching linear dynamical system for estimating seizure chronotypes.

Epilepsy is a disorder characterized by paroxysmal transitions between multistable states. Dynamical systems have been useful for modeling the paroxysmal nature of seizures. At the same time, intracranial electroencephalography (EEG) recordings have recently discovered that an electrographic measure of epileptogenicity, interictal epileptiform activity, exhibits cycling patterns ranging from ultradian to multidien rhythmicity, with seizures phase-locked to specific phases of these latent cycles. However, many mechanistic questions about seizure cycles remain unanswered. Here, we provide a principled approach to recast the modeling of seizure chronotypes within a statistical dynamical systems framework by developing a Bayesian switching linear dynamical system (SLDS) with variable selection to estimate latent seizure cycles. We propose a Markov chain Monte Carlo algorithm that employs particle Gibbs with ancestral sampling to estimate latent cycles in epilepsy and apply unsupervised learning on spectral features of latent cycles to uncover clusters in cycling tendency. We analyze the largest database of patient-reported seizures in the world to comprehensively characterize multidien cycling patterns among 1,012 people with epilepsy, spanning from infancy to older adulthood. Our work advances knowledge of cycling in epilepsy by investigating how multidien seizure cycles vary in people with epilepsy, while demonstrating an application of an SLDS to frame seizure cycling within a nonlinear dynamical systems framework. It also lays the groundwork for future studies to pursue data-driven hypothesis generation regarding the mechanistic drivers of seizure cycles.

Seizure count forecasting to aid diagnostic testing in epilepsy.

OBJECTIVE: Epilepsy monitoring unit (EMU) admissions are critical for presurgical evaluation of drug-resistant epilepsy but may be nondiagnostic if an insufficient number of seizures are recorded. Seizure forecasting algorithms have shown promise for estimating the likelihood of seizures as a binary event in individual patients, but methods to predict how many seizures will occur remain elusive. Such methods could increase the diagnostic yield of EMU admissions and help patients mitigate seizure-related morbidity. Here, we evaluated the performance of a state-space method that uses prior seizure count data to predict future counts. METHODS: A Bayesian negative-binomial dynamic linear model (DLM) was developed to forecast daily electrographic seizure counts in 19 patients implanted with a responsive neurostimulation (RNS) device. Holdout validation was used to evaluate performance in predicting the number of electrographic seizures for forecast horizons ranging 1-7 days ahead. RESULTS: One-day-ahead prediction of the number of electrographic seizures using a negative-binomial DLM resulted in improvement over chance in 73.1% of time segments compared to a random chance forecaster and remained >50% for forecast horizons of up to 7 days. Superior performance (mean error = .99) was obtained in predicting the number of electrographic seizures in the next day compared to three traditional methods for count forecasting (integer-valued generalized autoregressive conditional heteroskedasticity model or INGARCH, 1.10; Croston, 1.06; generalized linear autoregressive moving average model or GLARMA, 2.00). Number of electrographic seizures in the preceding day and laterality of electrographic pattern detections had highest predictive value, with greater number of electrographic seizures and RNS magnet swipes in the preceding day associated with a higher number of electrographic seizures the next day. SIGNIFICANCE: This study demonstrates that DLMs can predict the number of electrographic seizures a patient will experience days in advance with above chance accuracy. This study represents an important step toward the translation of seizure forecasting methods into the optimization of EMU admissions.

Bilateral temporal lobe epilepsy: How many seizures are required in chronic ambulatory electrocorticography to estimate the laterality ratio?

OBJECTIVE: This study was undertaken to measure the duration of chronic electrocorticography (ECoG) needed to attain stable estimates of the seizure laterality ratio in patients with drug-resistant bilateral temporal lobe epilepsy (BTLE). METHODS: We studied 13 patients with drug-resistant BTLE who were implanted for at least 1 year with a responsive neurostimulation device (RNS System) that provides chronic ambulatory ECoG. Bootstrap analysis and nonlinear regression were applied to model the relationship between chronic ECoG duration and the probability of capturing at least one seizure. Laterality of electrographic seizures in chronic ECoG was compared with the seizure laterality ratio from Phase 1 scalp video-electroencephalographic (vEEG) monitoring. The Kaplan-Meier estimator was used to evaluate time to seizure laterality ratio convergence. RESULTS: Seizure laterality ratios from Phase 1 scalp vEEG monitoring correlated poorly with those from RNS chronic ECoG (r = .31, p = .30). Across the 13 patients, average electrographic seizure frequencies ranged from 1.4 seizures/month to 5.1 seizures/day. A 50% probability of recording at least one electrographic seizure required 9.1 days of chronic ECoG, and 90% probability required 44.3 days of chronic ECoG. A median recording duration of 150.9 days (5 months), corresponding to a median of 16 seizures, was needed before confidence intervals for the seizure laterality ratio reliably contained the long-term value. The median recording duration before the point estimate of the seizure laterality ratio converged to a stationary value was 236.8 days (7.9 months). SIGNIFICANCE: RNS chronic ECoG overcomes temporal sampling limitations intrinsic to inpatient Phase 1 vEEG evaluations. In patients with drug-resistant BTLE, approximately 8 months of chronic RNS ECoG are needed to precisely estimate the seizure laterality ratio, with 75% of people with BTLE achieving convergence after 1 year of RNS recording. For individuals who are candidates for unilateral resection based on seizure laterality, optimized recording duration may help avert morbidity associated with delay to definitive treatment.

Impact of intellectual and developmental disability on quality-of-life priorities in adults with epilepsy.

OBJECTIVE: Adults living with intellectual and developmental disability (IDD) and epilepsy (IDD-E) face challenges in addition to those faced by the general population of adults with epilepsy, which may be associated with distinct priorities for improving health-related quality of life (HR-QOL). This study sought to (1) conduct a survey of HR-QOL priorities identified by adults with IDD-E and caregivers, and (2) perform an exploratory cross-sectional comparison to adults with epilepsy who do not have IDD. METHODS: This cross-sectional study recruited 65 adults with IDD-E and 134 adults with epilepsy without IDD and caregivers. Using a three-step development process, 256 items from existing quality-of-life scales recommended by the American Academy of Neurology (AAN) were rated by patients/caregivers for their importance as HR-QOL priorities. HR-QOL items identified as critical to the majority of the sample of adults with IDD-E were reported. Health-related quality of life priorities were compared between adults with IDD-E and adults with epilepsy without IDD. RESULTS: Health-related quality of life was significantly lower in adults with IDD-E. Health-related quality of life domains identified as critical priorities by adults with IDD-E included seizure burden, anti-seizure medication side effects, seizure unpredictability, and family impact. Priorities for improving HR-QOL differed between adults with and without IDD-E, with concerns about family impact, difficulty finding appropriate living conditions, inadequate assistance, and difficulty transitioning from pediatric-to-adult care valued significantly more among those with IDD-E. SIGNIFICANCE: Intellectual and developmental disability is an important determinant of HR-QOL among adults with epilepsy. We report HR-QOL priorities identified by adults with IDD-E and their caregivers. These results may help epilepsy clinicians and researchers develop tailored strategies to address priorities of the patient with IDD-E/caregiver community.

Can machine learning improve randomized clinical trial analysis?

PURPOSE: Recently a realistic simulator of patient seizure diaries was developed that can reproduce effects seen in randomized clinical trials (RCTs). RCTs suffer from high costs and statistical inefficiencies. Using realistic simulation and machine learning this study aimed to identify a more statistically efficient outcome metric. METHODS: Five candidate deep learning architectures with 54 permutations of hyperparameters were compared to the traditional standard, median percent change (MPC). Each were also tested for type 1 error. All models had similar outcomes, with appropriate low levels of type 1 error. RESULTS: The simplest model was equivalent to a logistic regression of a histogram of individual percentage changes in seizure rate, requiring 21-22% less patients to discriminate drug from placebo at 90% power. This model was referred to as LPC. CONCLUSION: Future studies to validate LPC may enable faster, cheaper and more efficient clinical trials.

Guidelines for Conducting Ethical Artificial Intelligence Research in Neurology: A Systematic Approach for Clinicians and Researchers.

Preemptive recognition of the ethical implications of study design and algorithm choices in artificial intelligence (AI) research is an important but challenging process. AI applications have begun to transition from a promising future to clinical reality in neurology. As the clinical management of neurology is often concerned with discrete, often unpredictable, and highly consequential events linked to multimodal data streams over long timescales, forthcoming advances in AI have great potential to transform care for patients. However, critical ethical questions have been raised with implementation of the first AI applications in clinical practice. Clearly, AI will have far-reaching potential to promote, but also to endanger, ethical clinical practice. This article employs an anticipatory ethics approach to scrutinize how researchers in neurology can methodically identify ethical ramifications of design choices early in the research and development process, with a goal of preempting unintended consequences that may violate principles of ethical clinical care. First, we discuss the use of a systematic framework for researchers to identify ethical ramifications of various study design and algorithm choices. Second, using epilepsy as a paradigmatic example, anticipatory clinical scenarios that illustrate unintended ethical consequences are discussed, and failure points in each scenario evaluated. Third, we provide practical recommendations for understanding and addressing ethical ramifications early in methods development stages. Awareness of the ethical implications of study design and algorithm choices that may unintentionally enter AI is crucial to ensuring that incorporation of AI into neurology care leads to patient benefit rather than harm.

Evidence for long memory in focal seizure duration.

Objective: A major source of disability for people with epilepsy involves uncertainty surrounding seizure timing and severity. Although patients often report that long seizure-free intervals are followed by more severe seizures, there is little experimental evidence supporting this observation. Optimal characterization of seizure severity is debated; however, seizure duration is associated with seizure type and can be quantified in electrographic recordings as a limited proxy of clinical seizure severity. Here, using chronic intracranial electroencephalography (cEEG), we investigate the relationship between interseizure interval (ISI) and duration of the subsequent seizure. Methods: We performed a retrospective analysis of 14 subjects implanted with a responsive neurostimulation device (RNS System) that provides cEEG, including timestamps of electrographic seizures. We determined seizure durations for isolated seizures and for representative seizures from clusters determined through unsupervised methods. For each subject, the median ISI preceding long-duration seizures, defined as the top quintile of seizure durations, was compared with the median ISI preceding seizures with durations in the residual quintiles. In a group analysis, the mean seizure duration and the proportion of long-duration seizures were compared across ISI categories representing different lengths. Results: For 5 out of 14 subjects (36%), the median ISI preceding long-duration seizures was significantly greater than the median ISI preceding shorter-duration seizures. In the group analysis, when ISI was categorized by length, the proportion of long-duration seizures within the high ISI category was significantly higher than that of the low ISI category (P < 0.001). Significance: By leveraging cEEG and accounting for seizure clusters, we found that the likelihood of long-duration seizures positively correlates with ISI length, in a subset of individuals. These findings corroborate anecdotal clinical observations and support the existence of capacitor-like long memory processes governing the dynamics of focal seizures.

Evaluation and recommendations for effective data visualization for seizure forecasting algorithms.

OBJECTIVE: Seizure forecasting algorithms have become increasingly accurate and may reduce the morbidity and mortality caused by seizure unpredictability. Translating these benefits into meaningful health outcomes for people with epilepsy requires effective data visualization of algorithm outputs. To date, no studies have investigated patient and physician perspectives on effective translation of algorithm outputs into data visualizations through health information technology. MATERIALS AND METHODS: We developed front-end data visualizations as part of a Seizure Forecast Visualization Toolkit. We surveyed 627 people living with epilepsy and caregivers, and 28 epilepsy healthcare providers. Respondents scored each visualization in terms of international standardized software quality criteria for functionality, appropriateness, and usability. RESULTS: People with epilepsy and caregivers ranked hourly radar charts highest for protecting against errors in interpreting forecasts, reducing anxiety from seizure unpredictability, and understanding seizure patterns. Accuracy in interpreting visuals, such as a risk gauge, was dependent on seizure frequency. Visuals showing hourly/daily forecasts were more useful for patients who experienced seizure cycling than those who did not. Hourly line graphs and monthly heat maps were rated highest among clinicians for ease of understanding, anticipated integration into clinical practice, and the likelihood of clinical usage. Epilepsy providers indicated that daily heat maps, daily line graphs, and hourly line graphs were most useful for interpreting seizure diary patterns, assessing therapy impact, and counseling on seizure safety. DISCUSSION: The choice of data visualization impacts the effective translation of seizure forecast algorithms into meaningful health outcomes. CONCLUSION: This effort underlines the importance of incorporating standardized, quantitative methods for assessing the effectiveness of data visualization to translate seizure forecast algorithms into clinical practice.

Evidence of state-dependence in the effectiveness of responsive neurostimulation for seizure modulation.

BACKGROUND: An implanted device for brain-responsive neurostimulation (RNS® System) is approved as an effective treatment to reduce seizures in adults with medically-refractory focal epilepsy. Clinical trials of the RNS System demonstrate population-level reduction in average seizure frequency, but therapeutic response is highly variable. HYPOTHESIS: Recent evidence links seizures to cyclical fluctuations in underlying risk. We tested the hypothesis that effectiveness of responsive neurostimulation varies based on current state within cyclical risk fluctuations. METHODS: We analyzed retrospective data from 25 adults with medically-refractory focal epilepsy implanted with the RNS System. Chronic electrocorticography was used to record electrographic seizures, and hidden Markov models decoded seizures into fluctuations in underlying risk. State-dependent associations of RNS System stimulation parameters with changes in risk were estimated. RESULTS: Higher charge density was associated with improved outcomes, both for remaining in a low seizure risk state and for transitioning from a high to a low seizure risk state. The effect of stimulation frequency depended on initial seizure risk state: when starting in a low risk state, higher stimulation frequencies were associated with remaining in a low risk state, but when starting in a high risk state, lower stimulation frequencies were associated with transition to a low risk state. Findings were consistent across bipolar and monopolar stimulation configurations. CONCLUSION: The impact of RNS on seizure frequency exhibits state-dependence, such that stimulation parameters which are effective in one seizure risk state may not be effective in another. These findings represent conceptual advances in understanding the therapeutic mechanism of RNS, and directly inform current practices of RNS tuning and the development of next-generation neurostimulation systems.

An optimized QF-binary expression system for use in zebrafish.

The zebrafish model organism has been of exceptional utility for the study of vertebrate development and disease through the application of tissue-specific labelling and overexpression of genes carrying patient-derived mutations. However, there remains a need for a binary expression system that is both non-toxic and not silenced over animal generations by DNA methylation. The Q binary expression system derived from the fungus Neurospora crassa is ideal, because the consensus binding site for the QF transcription factor lacks CpG dinucleotides, precluding silencing by CpG-meditated methylation. To optimize this system for zebrafish, we systematically tested several variants of the QF transcription factor: QF full length; QF2, which lacks the middle domain; QF2w, which is an attenuated version of QF2; and chimeric QFGal4. We found that full length QF and QF2 were strongly toxic to zebrafish embryos, QF2w was mildly toxic, and QFGal4 was well tolerated, when injected as RNA or expressed ubiquitously from stable transgenes. In addition, QFGal4 robustly activated a Tg(QUAS:GFPNLS) reporter transgene. To increase the utility of this system, we also modified the QF effector sequence termed QUAS, which consists of five copies of the QF binding site. Specifically, we decreased both the CpG dinucleotide content, as well as the repetitiveness of QUAS, to reduce the risk of transgene silencing via CpG methylation. Moreover, these modifications to QUAS removed leaky QF-independent neural expression that we detected in the original QUAS sequence. To demonstrate the utility of our QF optimizations, we show how the Q-system can be used for lineage tracing using a Cre-dependent Tg(ubi:QFGal4-switch) transgene. We also demonstrate that QFGal4 can be used in transient injections to tag and label endogenous genes by knocking in QFGal4 into sox2 and ubiquitin C genes.

Natural history of generalized motor seizures: A retrospective analysis.

PURPOSE: This study aims to characterize the natural history of generalized motor seizures through longitudinal stratification of patient-reported clinical seizures into high, medium and low rates of generalized motor seizures (also known as generalized tonic-clonic seizures or GTCs). METHODS: From 2007 to 2018, 1.4 million seizures were recorded by 12,402 SeizureTracker.com users that met inclusion/exclusion criteria. The number of GTCs per year since the first seizure diary entry was calculated for each user and categorized as: Low (0 GTCs/year), Medium (1-2 GTCs/year), or High (>3 GTCs/year) GTC rates. RESULTS: Kaplan-Meier survival curves for the time until exiting the initial category were computed. There was a global difference between risk groups (p < 0.01). Further pairwise log rank tests revealed a difference between each pair of risk groups (p < 0.01). At 3 years, 40.8% of people initially presenting with high GTC rates remained in their initial category, while 77.3% of people initially presenting with low GTC rates remained in their initial category. CONCLUSION: A patient with a low rate of GTCs is likely to remain at low risk for future GTCs, whereas higher GTC rate patients (at least one GTC/year) may leave their initial risk stratification. Thus, yearly re-assessment may be prudent when considering risk of further GTCs. Given the association between higher yearly rates of GTCs with increased SUDEP risk and morbidity in epilepsy, further validation of these findings is important for prognostication.

Individualizing the definition of seizure clusters based on temporal clustering analysis.

OBJECTIVE: Seizure clusters are often encountered in people with poorly controlled epilepsy. Detection of seizure clusters is currently based on simple clinical rules, such as two seizures separated by four or fewer hours or multiple seizures in 24 h. Current definitions fail to distinguish between statistically significant clusters and those that may result from natural variation in the person's seizures. Ability to systematically define when a seizure cluster is significant for the individual carries major implications for treatment. However, there is no uniform consensus on how to define seizure clusters. This study proposes a principled statistical approach to defining seizure clusters that addresses these issues. METHODS: A total of 533,968 clinical seizures from 1,748 people with epilepsy in the Seizure Tracker™ seizure diary database were used for algorithm development. We propose an algorithm for automated individualized seizure cluster identification combining cumulative sum change-point analysis with bootstrapping and aberration detection, which provides a new approach to personalized seizure cluster identification at user-specified levels of clinical significance. We develop a standalone user interface to make the proposed algorithm accessible for real-time seizure cluster identification (ClusterCalc™). Clinical impact of systematizing cluster identification is demonstrated by comparing empirically-defined clusters to those identified by routine seizure cluster definitions. We also demonstrate use of the Hurst exponent as a standardized measure of seizure clustering for comparison of seizure clustering burden within or across patients. RESULTS: Seizure clustering was present in 26.7 % (95 % CI, 24.5-28.7 %) of people with epilepsy. Empirical tables were provided for standardizing inter- and intra-patient comparisons of seizure cluster tendency. Using the proposed algorithm, we found that 37.7-59.4 % of seizures identified as clusters based on routine definitions had high probability of occurring by chance. Several clusters identified by the algorithm were missed by conventional definitions. The utility of the ClusterCalc algorithm for individualized seizure cluster detection is demonstrated. SIGNIFICANCE: This study proposes a principled statistical approach to individualized seizure cluster identification and demonstrates potential for real-time clinical usage through ClusterCalc. Using this approach accounts for individual variations in baseline seizure frequency and evaluates statistical significance. This new definition has the potential to improve individualized epilepsy treatment by systematizing identification of unrecognized seizure clusters and preventing unnecessary intervention for random events previously considered clusters.