Dataset Information

Unsupervised classification of high-frequency oscillations in human neocortical epilepsy and control patients.

ABSTRACT: High-frequency oscillations (HFOs) have been observed in animal and human intracranial recordings during both normal and aberrant brain states. It has been proposed that the relationship between subclasses of these oscillations can be used to identify epileptic brain. Studies of HFOs in epilepsy have been hampered by selection bias arising primarily out of the need to reduce the volume of data so that clinicians can manually review it. In this study, we introduce an algorithm for detecting and classifying these signals automatically and demonstrate the tractability of analyzing a data set of unprecedented size, over 31,000 channel-hours of intracranial electroencephalographic (iEEG) recordings from micro- and macroelectrodes in humans. Using an unsupervised approach that does not presuppose a specific number of clusters in the data, we show direct evidence for the existence of distinct classes of transient oscillations within the 100- to 500-Hz frequency range in a population of nine neocortical epilepsy patients and two controls. The number of classes we find, four (three plus one putative artifact class), is consistent with prior studies that identify "ripple" and "fast ripple" oscillations using human-intensive methods and, additionally, identifies a less examined class of mixed-frequency events.

SUBMITTER: Blanco JA

PROVIDER: S-EPMC2997042 | biostudies-literature | 2010 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Unsupervised classification of high-frequency oscillations in human neocortical epilepsy and control patients.

Blanco Justin A JA Stead Matt M Krieger Abba A Viventi Jonathan J Marsh W Richard WR Lee Kendall H KH Worrell Gregory A GA Litt Brian B

Journal of neurophysiology 20100901 5

High-frequency oscillations (HFOs) have been observed in animal and human intracranial recordings during both normal and aberrant brain states. It has been proposed that the relationship between subclasses of these oscillations can be used to identify epileptic brain. Studies of HFOs in epilepsy have been hampered by selection bias arising primarily out of the need to reduce the volume of data so that clinicians can manually review it. In this study, we introduce an algorithm for detecting and c ...[more]

PMID: 20810694

Similar Datasets

Project description:PurposeTo investigate the characteristics of intracranial ictal high-frequency oscillations (HFOs).MethodsAmong neocortical epilepsy patients who underwent intracranial monitoring and surgery, we studied patients with well-defined, unifocal seizure onsets characterized by discrete HFOs (≥70 Hz). Patients with multifocal or bilateral independent seizure onsets, electroencephalography (EEG) acquired at <1,000 Hz sampling rate, and nonresective surgery were excluded. Based on a prospectively defined protocol, we defined the seizure-onset zone (SOZ) presurgically to include only those channels with HFOs that showed subsequent sustained evolution (HFOs+ channels) but not the channels that lacked evolution (HFOs- channels). We then resected the SOZ as defined above, 1 cm of the surrounding cortex, and immediate spread area, modified by the presence of eloquent cortex in the vicinity. For purposes of this study, we also defined the SOZ based on the conventional frequency activity (CFA, <70 Hz) at seizure onset, although that information was not considered for preoperative determination of the surgical boundary. We investigated the temporal and spatial characteristics of the ictal HFOs post hoc by visual and spectral methods, and also compared them to the seizure onset defined by the CFA.Key findingsOf 14 consecutive neocortical epilepsy patients, six patients met the inclusion criteria. Magnetic resonance imaging (MRI) was normal or showed heterotopia. All had subdural electrodes, with additional intracerebral depth electrodes in some. Electrode coverage was extensive (median 94 channels), including limited contralateral coverage. Seizure onsets were lobar or multilobar. Resections were performed per protocol, except in two patients where complete resection of the SOZ could not be done due to overlap with speech area. Histology was abnormal in all patients. Postoperative outcome was class I/II (n = 5, 83%) or class III over a mean follow-up of 27 months. Post hoc analysis of 15 representative seizures showed that the ictal HFOs were widespread at seizure onset but evolved subsequently with different characteristics. In contrast to HFOs-, the HFOs+ were significantly higher in peak frequency (97.1 vs. 89.1 Hz, p = 0.001), more robust (nearly twofold higher peak power, p < 0.0001), and spatially restricted [mean 12.2 vs. 22.4 channels; odds ratio (OR) 0.51, 95% confidence interval (CI) 0.42-0.62; p < 0.0001]. The seizure onset defined by HFOs+ was earlier (by an average of 0.41 s), and occurred in a significantly different and smaller distribution (OR 0.27, 95% CI 0.21-0.34, p < 0.0001), than the seizure onset defined by the CFA. As intended, the HFOs+ channels were 10 times more likely to have been resected than the HFOs- channels (OR 9.7, 95% CI 5-17, p < 0.0001).SignificanceOur study demonstrates the widespread occurrence of ictal HFOs at seizure onset, outlines a practical method to localize the SOZ based on their restricted pattern of evolution, and highlights the differences between the SOZs defined by HFOs and CFA. We show that smaller resections, restricted mainly to the HFOs channels with evolution, can lead to favorable seizure outcome. Our findings support the notion of widespread epileptic networks underlying neocortical epilepsy.

Project description:High-frequency oscillations >80 Hz (HFOs) have unique features distinguishing them from spikes and artifactual components that can be well-evidenced in the time-frequency representations. We introduce an unsupervised HFO detector that uses computer-vision algorithms to detect HFO landmarks on two-dimensional (2D) time-frequency maps. To validate the detector, we introduce an analytical model of the HFO based on a sinewave having a Gaussian envelope, for which analytical equations in time-frequency space can be derived, allowing us to establish a direct correspondence between common HFO detection criteria in the time domain with the ones in the frequency domain, used by the computer-vision detection algorithm. The detector identifies potential HFO events on the time-frequency representation, which are classified as true HFOs if criteria regarding the HFO's frequency, amplitude, and duration are met. The detector is validated on simulated HFOs according to the analytical model, in the presence of noise, with different signal-to-noise ratios (SNRs) ranging from -9 to 0 dB. The detector's sensitivity was 0.64 at an SNR of -9 dB, 0.98 at -6 dB, and >0.99 at -3 dB and 0 dB, while its positive prediction value was >0.95, regardless of the SNR. Using the same simulation dataset, our detector is benchmarked against four previously published HFO detectors. The F-measure, a combined metric that takes into account both sensitivity and positive prediction value, was used to compare detection algorithms. Our detector surpassed the other detectors at -6, -3, and 0 dB and had the second best F-score at -9 dB SNR after the MNI detector (0.77 vs. 0.83). The ability to detect HFOs in clinical recordings has been tested on a set of 36 intracranial electroencephalogram (EEG) channels in six patients, with 89% of the detections being validated by two independent reviewers. The results demonstrate that the unsupervised detection of HFOs based on their 2D features in time-frequency maps is feasible and has a performance comparable or better than the most used HFO detectors.

Project description:Simultaneous scalp EEG/functional MRI measures non-invasively haemodynamic responses to interictal epileptic discharges, which are related to the epileptogenic zone. High frequency oscillations are also an excellent indicator of this zone, but are primarily recorded from intracerebral EEG. We studied the spatial overlap of these two important markers in patients with drug-resistant epilepsy to assess if their combination could help better define the extent of the epileptogenic zone. We included patients who underwent EEG-functional MRI and later intracerebral EEG. Based on intracerebral EEG findings, we separated patients with unifocal seizures from patients with multifocal or unknown onset seizures. Haemodynamic t-maps were coregistered with the intracerebral electrode positions. Each EEG channel was classified as pertaining to one of the following categories: primary haemodynamic cluster (maximum t-value), secondary cluster (t-value > 90% of the primary cluster) or outside the primary and secondary clusters. We marked high frequency oscillations (ripples: 80-250 Hz; fast ripples: 250-500 Hz) during 1 h of slow wave sleep, and compared their rates in each haemodynamic category. After classifying channels as high- or low-rate, the proportion of high-rate channels within the primary or primary plus secondary clusters was compared to the proportion expected by chance. Twenty-five patients, 11 with unifocal and 14 with multifocal/unknown seizure onsets, were studied. We found a significantly higher median high frequency oscillation rate in the primary cluster compared to secondary cluster and outside these two clusters for the unifocal group (P < 0.0001), but not for the multifocal/unknown group. For the unifocal group, the number of high-rate channels within the primary or primary plus secondary clusters was significantly higher than expected by chance. This held only for the high-ripple-rate channels in the multifocal/unknown group. At the patient level, most patients (18/25, or 72%) had at least one high-rate channel within a primary cluster. In patients with unifocal epilepsy, the maximum haemodynamic response (primary cluster) related to scalp interictal discharges overlaps with the tissue generating high frequency oscillations at high rates. If intracranial EEG is warranted, this response should be explored. As a tentative clinical use of the combination of these techniques we propose that higher high frequency oscillation rates inside than outside the maximum response indicates that the patient has indeed a focal epileptogenic zone demarcated by this response, whereas similar rates inside and outside may indicate a widespread epileptogenic zone or an epileptogenic zone not covered by the implantation.

Dataset Information

Unsupervised classification of high-frequency oscillations in human neocortical epilepsy and control patients.

Publications

Unsupervised classification of high-frequency oscillations in human neocortical epilepsy and control patients.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets