Neurosurgery Clinics of North America, Journal Year: 2023, Volume and Issue: 35(1), P. 27 - 48
Published: Sept. 22, 2023
Language: Английский
Royal Society Open Science, Journal Year: 2022, Volume and Issue: 9(8)
Published: Aug. 1, 2022
This paper proposes an artificial intelligence system that continuously improves over time at event prediction using initially unlabelled data by self-supervised learning. Time-series are inherently autocorrelated. By a detection model to generate weak labels on the fly, which concurrently used as targets train time-shifted input stream, this autocorrelation can effectively be harnessed reduce burden of manual labelling. is critical in medical patient monitoring, it enables development personalized forecasting models without demanding annotation long sequences physiological signal recordings. We perform feasibility study seizure prediction, identified ideal test case, pre-ictal brainwaves patient-specific, and tailoring individual patients known improve performance significantly. Our approach individualized for 10 patients, showing average relative improvement sensitivity 14.30% reduction false alarms 19.61% early forecasting. proof-of-concept continuous stream time-series neurophysiological paves way towards low-power neuromorphic neuromodulation system.
Language: Английский
Citations
21
Epilepsia, Journal Year: 2023, Volume and Issue: 64(S3)
Published: Feb. 13, 2023
A lot of mileage has been made recently on the long and winding road toward seizure forecasting. Here we briefly review some selected milestones passed along way, which were discussed at International Conference for Technology Analysis Seizures-ICTALS 2022-convened University Bern, Switzerland. Major impetus was gained from wearable implantable devices that record not only electroencephalography, but also data motor behavior, acoustic signals, various signals autonomic nervous system. This multimodal monitoring can be performed ultralong timescales covering months or years. Accordingly, features metrics extracted these now assess dynamics with a greater degree completeness. Most prominently, this allowed confirmation long-suspected cyclical nature interictal epileptiform activity, risk, seizures. The cover daily, multi-day, yearly cycles. Progress fueled by approaches originating interdisciplinary field network science. Considering epilepsy as large-scale disorder yielded novel perspectives pre-ictal evolving epileptic brain. In addition to discrete predictions will take place in specified prediction horizon, community broadened scope probabilistic forecasts risk continuously time. shift gears triggered incorporation additional quantify performance forecasting algorithms, should compared chance constrained stochastic null models. An imminent task utmost importance is find optimal ways communicate output seizure-forecasting algorithms patients, caretakers, clinicians, so they have socioeconomic impact improve patients' well-being.
Language: Английский
Citations
12
Epilepsia, Journal Year: 2023, Volume and Issue: 64(5), P. 1125 - 1174
Published: Feb. 15, 2023
Abstract Antiseizure medication (ASM) is the primary treatment for epilepsy. In clinical practice, methods to assess ASM efficacy (predict seizure freedom or reduction), during any phase of drug lifecycle, are limited. This scoping review identifies and appraises prognostic electroencephalographic (EEG) biomarkers models that use EEG features, which associated with outcomes following initiation, dose adjustment, withdrawal. We also aim summarize population context in these were identified described, understand how they could be used practice. Between January 2021 October 2022, four databases, references, citations systematically searched studies investigating changes interictal using features outcomes. Study bias was appraised modified Quality Prognosis Studies criteria. Results synthesized into a qualitative review. Of 875 identified, 93 included. Biomarkers classed as (visually by wave morphology) quantitative. Qualitative include identifying hypsarrhythmia, centrotemporal spikes, epileptiform discharges (IED), classifying normal/abnormal/epileptiform, photoparoxysmal response. Quantitative statistics applied IED, high‐frequency activity, frequency band power, current source density estimates, pairwise statistical interdependence between channels, measures complexity. Prognostic Cox proportional hazards machine learning models. There promise some quantitative efficacy, but further research required. insufficient evidence conclude specific biomarker can particular prognosticate efficacy. potential battery biomarkers, combined However, many confounders need addressed translation
Language: Английский
Citations
11
Epileptic Disorders, Journal Year: 2024, Volume and Issue: 26(2), P. 199 - 208
Published: Feb. 9, 2024
Abstract Objective Automated seizure detection of focal epileptic seizures is needed for objective quantification to optimize the treatment patients with epilepsy. Heart rate variability (HRV)‐based using patient‐adaptive threshold logistic regression machine learning (LRML) methods has presented promising performance in a study Danish patient cohort. The this was assess generalizability novel LRML algorithm by validating it dataset recorded from long‐term video‐EEG monitoring (LTM) Brazilian Methods Ictal and inter‐ictal ECG‐data epochs during LTM were analyzed retrospectively. Thirty‐four had 107 (79 focal, 28 generalized tonic–clonic [GTC] including focal‐to‐bilateral‐tonic–clonic seizures) eligible analysis, total 185.5 h recording. Because HRV‐based only suitable marked ictal autonomic change, >50 beats/min change heart selected as responders. applied all elected ECG data, results computed separately responders non‐responders. Results yielded sensitivity 84.8% (95% CI: 75.6–93.9) false alarm .25/24 responder group (22 patients, 59 seizures). Twenty‐five 26 GTC detected (96.2%), 25 33 without bilateral convulsions (75.8%). Significance confirms new, independent external good previous suggests that method generalizable. This seems useful detecting both seizures. can be embedded wearable system alert caregivers generate counts helping patients.
Language: Английский
Citations
4
Epilepsia, Journal Year: 2024, Volume and Issue: unknown
Published: June 5, 2024
Abstract Wearable devices have attracted significant attention in epilepsy research recent years for their potential to enhance patient care through improved seizure monitoring and forecasting. This narrative review presents a detailed overview of the current clinical state art while addressing how that assess autonomic nervous system (ANS) function reflect seizures central (CNS) changes. includes description interactions between CNS ANS, including physiological epilepsy‐related changes affecting dynamics. We first discuss technical aspects measuring biosignals considerations using ANS sensors practice. then detection forecasting studies, highlighting performance capability biomarkers. Finally, we address field's challenges provide an outlook future developments.
Language: Английский
Citations
4
Epilepsia, Journal Year: 2025, Volume and Issue: unknown
Published: Feb. 24, 2025
Abstract Objective Epileptic seizures occurring in cyclical patterns is increasingly recognized as a significant opportunity to advance epilepsy management. Current methods for detecting seizure cycles rely on intrusive techniques or specialized biomarkers, thereby limiting their accessibility. This study evaluates non‐invasive cycle detection method using diaries and compares its accuracy with identified from intracranial electroencephalography (iEEG) interictal epileptiform discharges (IEDs). Methods Using data previously published first in‐human iEEG device trial ( n = 10), we analyzed through diary reports, seizures, IEDs. Cycle similarities across IEDs were evaluated at periods of 1 45 days spectral coherence, accuracy, precision, recall, the false‐positive rate. Results A coherence analysis raw signals showed moderately similar periodic components between seizures/day (median .43, IQR .68). In contrast, there was low IEDs/day .11, .18) .12, .19). Accuracy, recall scores, rates significantly higher than chance all participants (accuracy (mean ± standard deviation): .95 .02; precision: .56 .19; recall: rate: .02 .01). However, scores IED both did not perform above chance, average. Recall compared good reporters, under‐reporters, over‐reporters, generally performing better reporters under‐reporters over‐reporters. Significance These findings suggest that can be even individuals who under‐ over‐report seizures. approach offers an accessible alternative monitoring more invasive methods.
Language: Английский
Citations
0
Revue Neurologique, Journal Year: 2025, Volume and Issue: unknown
Published: March 1, 2025
Language: Английский
Citations
0
Epilepsia, Journal Year: 2025, Volume and Issue: unknown
Published: April 26, 2025
Abstract Objective The circadian rhythm synchronizes physiological and behavioral patterns with the 24‐h light–dark cycle. Disruption to is linked various health conditions, although optimal methods describe these disruptions remain unclear. An emerging approach examine intraindividual variability in measurable properties of over extended periods. Epileptic seizures are modulated by rhythms, but relevance disruption epilepsy remains unexplored. Our study investigates its relationship seizures. Methods We retrospectively analyzed >70 000 h wearable smartwatch data (Fitbit) from 143 people (PWE) 31 healthy controls. Circadian oscillations heart rate time series were extracted, daily estimates period, acrophase, amplitude produced, an entire recording calculated. Results PWE exhibited greater period (76 vs. 57 min, d = .66, p < .001) acrophase (64 48 .49, .004) compared controls, not (2 beats per minute, −.15, .49). Variability showed no correlation seizure frequency nor any differences between weeks without Significance For first time, we show that rhythms more variable PWE, detectable via consumer devices. However, association or occurrence was found, suggesting this might be underpinned etiology rather than being a seizure‐driven effect.
Language: Английский
Citations
0
Journal of Clinical Neurophysiology, Journal Year: 2024, Volume and Issue: 41(3), P. 207 - 213
Published: March 1, 2024
Among the many fears associated with seizures, patients epilepsy are greatly frustrated and distressed over seizure's apparent unpredictable occurrence. However, increasing evidence have emerged years to support that seizure occurrence is not a random phenomenon as previously presumed; it has cyclic rhythm oscillates multiple timescales. The pattern in rises falls of rate varies 24 hours, weeks, months, become target for development innovative devices intend detect, predict, forecast seizures. This article will review different tools available or been studied detection, prediction, forecasting, well challenges limitations utilization these devices. Although there strong rhythmicity occurrence, very little known about mechanism behind this oscillation. concludes early insights into regulations may potentially drive cyclical variability future directions.
Language: Английский
Citations
3
Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)
Published: Feb. 4, 2025
Language: Английский
Citations
0