Integrating Artificial Intelligence with Real-time Intracranial EEG
  Monitoring to Automate Interictal Identification of Seizure Onset Zones in
  Focal Epilepsy

There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

Abstract

An ability to map seizure-generating brain tissue, i.e., the seizure onset zone (SOZ), without recording actual seizures could reduce the duration of invasive EEG monitoring for patients with drug-resistant epilepsy. A widely-adopted practice in the literature is to compare the incidence (events/time) of putative pathological electrophysiological biomarkers associated with epileptic brain tissue with the SOZ determined from spontaneous seizures recorded with intracranial EEG, primarily using a single biomarker. Clinical translation of the previous efforts suffers from their inability to generalize across multiple patients because of (a) the inter-patient variability and (b) the temporal variability in the epileptogenic activity. Here, we report an artificial intelligence-based approach for combining multiple interictal electrophysiological biomarkers and their temporal characteristics as a way of accounting for the above barriers and show that it can reliably identify seizure onset zones in a study cohort of 82 patients who underwent evaluation for drug-resistant epilepsy. Our investigation provides evidence that utilizing the complementary information provided by multiple electrophysiological biomarkers and their temporal characteristics can significantly improve the localization potential compared to previously published single-biomarker incidence-based approaches, resulting in an average area under ROC curve (AUC) value of 0.73 in a cohort of 82 patients. Our results also suggest that recording durations between ninety minutes and two hours are sufficient to localize SOZs with accuracies that may prove clinically relevant. The successful validation of our approach on a large cohort of 82 patients warrants future investigation on the feasibility of utilizing intra-operative EEG monitoring and artificial intelligence to localize epileptogenic brain tissue.

Related collections

Most cited references 50

Record: found
Abstract: found
Article: not found

Long-term seizure outcomes following epilepsy surgery: a systematic review and meta-analysis.

Jose Téllez-Zenteno, Raj K. Dhar, Samuel Wiebe (2005)

Assessment of long-term outcomes is essential in brain surgery for epilepsy, which is an irreversible intervention for a chronic condition. Excellent short-term results of resective epilepsy surgery have been established, but less is known about long-term outcomes. We performed a systematic review and meta-analysis of the evidence on this topic. To provide evidence-based estimates of long-term results of various types of epilepsy surgery and to identify sources of variation in results of published studies, we searched Medline, Index Medicus, the Cochrane database, bibliographies of reviews, original articles and book chapters to identify articles published since 1991 that contained > or =20 patients of any age, undergoing resective or non-resective epilepsy surgery, and followed for a mean/median of > or =5 years. Two reviewers independently assessed study eligibility and extracted data, resolving disagreements through discussion. Seventy-six articles fulfilled our eligibility criteria, of which 71 reported on resective surgery (93%) and five (7%) on non-resective surgery. There were no randomized trials and only six studies had a control group. Some articles contributed more than one study, yielding 83 studies of which 78 dealt with resective surgery and five with non-resective surgery. Forty studies (51%) of resective surgery referred to temporal lobe surgery, 25 (32%) to grouped temporal and extratemporal surgery, seven (9%) to frontal surgery, two (3%) to grouped extratemporal surgery, two (3%) to hemispherectomy, and one (1%) each to parietal and occipital surgery. In the non-resective category, three studies reported outcomes after callosotomy and two after multiple subpial transections. The median proportion of long-term seizure-free patients was 66% with temporal lobe resections, 46% with occipital and parietal resections, and 27% with frontal lobe resections. In the long term, only 35% of patients with callosotomy were free of most disabling seizures, and 16% with multiple subpial transections remained free of all seizures. The year of operation, duration of follow-up and outcome classification system were most strongly associated with outcomes. Almost all long-term outcome studies describe patient cohorts without controls. Although there is substantial variation in outcome definition and methodology among the studies, consistent patterns of results emerge for various surgical interventions after adjusting for sources of heterogeneity. The long-term (> or =5 years) seizure free rate following temporal lobe resective surgery was similar to that reported in short-term controlled studies. On the other hand, long-term seizure freedom was consistently lower after extratemporal surgery and palliative procedures.

0 comments Cited 204 times – based on 0 reviews      Review now

Bookmark

Record: found
Abstract: found
Article: not found

High-frequency oscillations as a new biomarker in epilepsy.

Rina Zelmann, Jean Gotman, Maeike Zijlmans … (2012)

The discovery that electroencephalography (EEG) contains useful information at frequencies above the traditional 80Hz limit has had a profound impact on our understanding of brain function. In epilepsy, high-frequency oscillations (HFOs, >80Hz) have proven particularly important and useful. This literature review describes the morphology, clinical meaning, and pathophysiology of epileptic HFOs. To record HFOs, the intracranial EEG needs to be sampled at least at 2,000Hz. The oscillatory events can be visualized by applying a high-pass filter and increasing the time and amplitude scales, or EEG time-frequency maps can show the amount of high-frequency activity. HFOs appear excellent markers for the epileptogenic zone. In patients with focal epilepsy who can benefit from surgery, invasive EEG is often required to identify the epileptic cortex, but current information is sometimes inadequate. Removal of brain tissue generating HFOs has been related to better postsurgical outcome than removing the seizure onset zone, indicating that HFOs may mark cortex that needs to be removed to achieve seizure control. The pathophysiology of epileptic HFOs is challenging, probably involving populations of neurons firing asynchronously. They differ from physiological HFOs in not being paced by rhythmic inhibitory activity and in their possible origin from population spikes. Their link to the epileptogenic zone argues that their study will teach us much about the pathophysiology of epileptogenesis and ictogenesis. HFOs show promise for improving surgical outcome and accelerating intracranial EEG investigations. Their potential needs to be assessed by future research. Copyright © 2012 American Neurological Association.

0 comments Cited 138 times – based on 0 reviews      Review now

Bookmark

Record: found
Abstract: found
Article: not found

Interictal high-frequency oscillations (80-500 Hz) are an indicator of seizure onset areas independent of spikes in the human epileptic brain.

Jeffery Hall, R. Chander, Julia Jacobs … (2008)

High-frequency oscillations (HFOs) known as ripples (80-250 Hz) and fast ripples (250-500 Hz) can be recorded from macroelectrodes inserted in patients with intractable focal epilepsy. They are most likely linked to epileptogenesis and have been found in the seizure onset zone (SOZ) of human ictal and interictal recordings. HFOs occur frequently at the time of interictal spikes, but were also found independently. This study analyses the relationship between spikes and HFOs and the occurrence of HFOs in nonspiking channels. Intracerebral EEGs of 10 patients with intractable focal epilepsy were studied using macroelectrodes. Rates of HFOs within and outside spikes, the overlap between events, event durations, and the percentage of spikes carrying HFOs were calculated and compared according to anatomical localization, spiking activity, and relationship to the SOZ. HFOs were found in all patients, significantly more within mesial temporal lobe structures than in neocortex. HFOs could be seen in spiking as well as nonspiking channels in all structures. Rates and durations of HFOs were significantly higher in the SOZ than outside. It was possible to establish a rate of HFOs to identify the SOZ with better sensitivity and specificity than with the rate of spikes. HFOs occurred to a large extent independently of spikes. They are most frequent in mesial temporal structures. They are prominent in the SOZ and provide additional information on epileptogenicity independently of spikes. It was possible to identify the SOZ with a high specificity by looking at only 10 min of HFO activity.