scholarly journals Refining epileptogenic high-frequency oscillations using deep learning: a reverse engineering approach

2021 ◽  
Author(s):  
Yipeng Zhang ◽  
Qiujing Lu ◽  
Tonmoy Monsoor ◽  
Shaun A Hussain ◽  
Joe X Qiao ◽  
...  
Keyword(s):  
Deep Learning ◽  
Reverse Engineering ◽  
High Frequency ◽  
Time Window ◽  
Current Knowledge ◽  
Brain Regions ◽  
Seizure Freedom ◽  
Epileptogenic Zone ◽  
Artifact Detection ◽  
High Frequency Oscillations

Intracranially-recorded interictal high-frequency oscillations (HFOs) have been proposed as a promising spatial biomarker of the epileptogenic zone. However, visual verification of HFOs is time-consuming and exhibits poor inter-rater reliability. Furthermore, no method is currently available to distinguish HFOs generated from the epileptogenic zone (epileptogenic HFOs: eHFOs) from those generated from other areas (non-epileptogenic HFOs: non-eHFOs). To address these issues, we constructed a deep learning (DL)-based algorithm using HFO events from chronic intracranial electroencephalogram (iEEG) data via subdural grids from 19 children with medication-resistant neocortical epilepsy to: 1) replicate human expert annotation of artifacts and HFOs with or without spikes, and 2) discover eHFOs by designing a novel weakly supervised model (HFOs from the resected brain regions are initially labeled as eHFOs, and those from the preserved brain regions as non-eHFOs). The "purification power" of DL is then used to automatically relabel the HFOs to distill eHFOs. Using 12,958 annotated HFO events from 19 patients, the model achieved 96.3% accuracy on artifact detection (F1 score = 96.8%) and 86.5% accuracy on classifying HFOs with or without spikes (F1 score = 80.8%) using patient-wise cross-validation. Based on the DL-based algorithm trained from 84,602 HFO events from nine patients who achieved seizure-freedom after resection, the majority of such DL-discovered eHFOs were found to be HFOs with spikes (78.6%, p < 0.001). While the resection ratio of detected HFOs (number of resected HFOs/number of detected HFOs) did not correlate significantly with post-operative seizure freedom (the area under the curve [AUC]=0.76, p=0.06), the resection ratio of eHFOs positively correlated with post-operative seizure freedom (AUC=0.87, p=0.01). We discovered that the eHFOs had a higher signal intensity associated with ripple (80-250 Hz) and fast ripple (250-500 Hz) bands at the HFO onset and with a lower frequency band throughout the event time window (the inverted T-shaped), compared to non-eHFOs. We then designed perturbations on the input of the trained model for non-eHFOs to determine the model's decision-making logic. The model probability significantly increased towards eHFOs by the artificial introduction of signals in the inverted T-shaped frequency bands (mean probability increase: 0.285, p < 0.001), and by the artificial insertion of spike-like signals into the time domain (mean probability increase: 0.452, p < 0.001). With this DL-based framework, we reliably replicated HFO classification tasks by human experts. Using a reverse engineering technique, we distinguished eHFOs from others and identified salient features of eHFOs that aligned with current knowledge.

High-Frequency Oscillations (HFO)

2018 ◽  
pp. 131-148
Author(s):  
Julia Jacobs
Keyword(s):  
High Frequency ◽  
Automatic Detection ◽  
Brain Regions ◽  
Seizure Freedom ◽  
Seizure Onset ◽  
Clinical Value ◽  
Brain Areas ◽  
High Frequency Oscillations ◽  
Epileptic Spikes ◽  
Presurgical Assessment

High-frequency oscillations (HFO) are new EEG biomarkers for epileptic tissue. These oscillations range in frequencies from 80 to 500 Hz and can be recorded with standard intracranial macroelectrodes. During the presurgical assessment of patients with refractory epilepsy, HFO have been found to occur mainly over seizure onset areas. HFO might co-occur with epileptic spikes, but are more specific to epileptic tissue than epileptic spikes. Several retrospective studies showed a correlation between the removal of brain areas generating HFO and postsurgical seizure freedom. In addition to demonstrating the clinical value of HFO analysis, this chapter provides a detailed introduction to the techniques for analysing HFO, including recording techniques and visual and automatic detection tools, and to interpretation of the results. It also reviews methodological challenges such as the occurrence of physiological HFO and the variability of HFO rates between patients and brain regions.

scholarly journals A spiking neural network (SNN) for detecting high frequency oscillations (HFOs) in the intraoperative ECoG

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Karla Burelo ◽  
Mohammadali Sharifshazileh ◽  
Niklaus Krayenbühl ◽  
Georgia Ramantani ◽  
Giacomo Indiveri ◽  
...  
Keyword(s):  
Neural Network ◽  
Real Time ◽  
Epilepsy Surgery ◽  
High Frequency ◽  
Seizure Outcome ◽  
Seizure Freedom ◽  
Epileptogenic Zone ◽  
Spiking Neural Network ◽  
High Frequency Oscillations ◽  
Neuromorphic Hardware

AbstractTo achieve seizure freedom, epilepsy surgery requires the complete resection of the epileptogenic brain tissue. In intraoperative electrocorticography (ECoG) recordings, high frequency oscillations (HFOs) generated by epileptogenic tissue can be used to tailor the resection margin. However, automatic detection of HFOs in real-time remains an open challenge. Here we present a spiking neural network (SNN) for automatic HFO detection that is optimally suited for neuromorphic hardware implementation. We trained the SNN to detect HFO signals measured from intraoperative ECoG on-line, using an independently labeled dataset (58 min, 16 recordings). We targeted the detection of HFOs in the fast ripple frequency range (250-500 Hz) and compared the network results with the labeled HFO data. We endowed the SNN with a novel artifact rejection mechanism to suppress sharp transients and demonstrate its effectiveness on the ECoG dataset. The HFO rates (median 6.6 HFO/min in pre-resection recordings) detected by this SNN are comparable to those published in the dataset (Spearman’s $$\rho$$ ρ = 0.81). The postsurgical seizure outcome was “predicted” with 100% (CI [63 100%]) accuracy for all 8 patients. These results provide a further step towards the construction of a real-time portable battery-operated HFO detection system that can be used during epilepsy surgery to guide the resection of the epileptogenic zone.

scholarly journals Interictal high frequency background activity as a biomarker of epileptogenic tissue

2021 ◽  
Author(s):  
Truman Stovall ◽  
Brian Hunt ◽  
Simon Glynn ◽  
William C Stacey ◽  
Stephen V Gliske
Keyword(s):  
Logistic Regression ◽  
Confidence Interval ◽  
High Frequency ◽  
High Frequency Oscillation ◽  
Epileptogenic Zone ◽  
Odds Ratios ◽  
Seizure Onset ◽  
High Frequency Oscillations ◽  
Seizure Onset Zone ◽  
Background Data

Abstract High Frequency Oscillations are very brief events that are a well-established biomarker of the epileptogenic zone, but are rare and comprise only a tiny fraction of the total recorded EEG. We hypothesize that the interictal high frequency “background” data, which has received little attention but represents the majority of the EEG record, also may contain additional, novel information for identifying the epileptogenic zone. We analyzed intracranial EEG (30–500 Hz frequency range) acquired from 24 patients who underwent resective surgery. We computed 38 quantitative features based on all usable, interictal data (63–307 hours per subject), excluding all detected high frequency oscillations. We assessed association between each feature and the seizure onset zone and resected volume using logistic regression. A pathology score per channel was also created via principle component analysis and logistic regression, using hold-out-one-patient cross validation to avoid in-sample training. Association of the pathology score with the seizure onset zone and resected volume was quantified using an asymmetry measure. Many features were associated with the seizure onset zone: 23/38 features had odds ratios &gt;1.3 or &lt; 0.7 and 17/38 had odds ratios different than zero with high significance (p &lt; 0.001/39, logistic regression with Bonferroni Correction). The pathology score, the rate of high frequency oscillations, and their channel-wise product were each strongly associated with the seizure onset zone (median asymmetry &gt; =0.44, good surgery outcome patients; median asymmetry &gt; =0.40, patients with other outcomes; 95% confidence interval &gt; 0.27 in both cases). The pathology score and the channel-wise product also had higher asymmetry with respect to the seizure onset zone than the high frequency oscillation rate alone (median difference in asymmetry &gt; =0.18, 95% confidence interval &gt;0.05). These results support that the high frequency background data contains useful information for determining the epileptogenic zone, distinct and complementary to information from detected high frequency oscillations. The concordance between the high frequency activity pathology score and the rate of high frequency oscillations appears to be a better biomarker of epileptic tissue than either measure alone.

scholarly journals Nasal respiration is necessary for ketamine-dependent high frequency network oscillations and behavioral hyperactivity in rats

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
Jacek Wróbel ◽  
Władysław Średniawa ◽  
Gabriela Jurkiewicz ◽  
Jarosław Żygierewicz ◽  
Daniel K. Wójcik ◽  
...  
Keyword(s):  
High Frequency ◽  
Ventral Striatum ◽  
Behavioral Activation ◽  
Brain Regions ◽  
Basal Respiration ◽  
Oscillatory Activity ◽  
Nasal Airflow ◽  
High Frequency Oscillations ◽  
Theta Frequency ◽  
Freely Moving

Abstract Changes in oscillatory activity are widely reported after subanesthetic ketamine, however their mechanisms of generation are unclear. Here, we tested the hypothesis that nasal respiration underlies the emergence of high-frequency oscillations (130–180 Hz, HFO) and behavioral activation after ketamine in freely moving rats. We found ketamine 20 mg/kg provoked “fast” theta sniffing in rodents which correlated with increased locomotor activity and HFO power in the OB. Bursts of ketamine-dependent HFO were coupled to “fast” theta frequency sniffing. Theta coupling of HFO bursts were also found in the prefrontal cortex and ventral striatum which, although of smaller amplitude, were coherent with OB activity. Haloperidol 1 mg/kg pretreatment prevented ketamine-dependent increases in fast sniffing and instead HFO coupling to slower basal respiration. Consistent with ketamine-dependent HFO being driven by nasal respiration, unilateral naris blockade led to an ipsilateral reduction in ketamine-dependent HFO power compared to the control side. Bilateral nares blockade reduced ketamine-induced hyperactivity and HFO power and frequency. These findings suggest that nasal airflow entrains ketamine-dependent HFO in diverse brain regions, and that the OB plays an important role in the broadcast of this rhythm.

scholarly journals The morphology of high frequency oscillations (HFO) does not improve delineating the epileptogenic zone

2016 ◽  
Vol 127 (4) ◽  
pp. 2140-2148 ◽  
Author(s):  
Sergey Burnos ◽  
Birgit Frauscher ◽  
Rina Zelmann ◽  
Claire Haegelen ◽  
Johannes Sarnthein ◽  
...  
Keyword(s):  
High Frequency ◽  
Epileptogenic Zone ◽  
High Frequency Oscillations

scholarly journals Blinded study: prospectively defined high frequency oscillations predict seizure outcome in individual patients

2020 ◽  
Author(s):  
V Dimakopoulos ◽  
P Mégevand ◽  
E Boran ◽  
S Momjian ◽  
M Seeck ◽  
...  
Keyword(s):  
Epilepsy Surgery ◽  
High Frequency ◽  
Dwell Time ◽  
Nrem Sleep ◽  
Seizure Outcome ◽  
Seizure Freedom ◽  
University Hospitals ◽  
High Frequency Oscillations ◽  
Depth Electrodes

AbstractBackgroundInterictal high frequency oscillations (HFO) are discussed as biomarkers for epileptogenic brain tissue that should be resected in epilepsy surgery to achieve seizure freedom. The prospective classification of tissue sampled by individual electrode contacts remains a challenge. We have developed an automated, prospective definition of clinically relevant HFO in intracranial EEG (iEEG) from MNI Montreal and tested it in iEEG from Zurich. We here validate the algorithm on iEEG recorded in an independent epilepsy center so that HFO analysis was blinded to seizure outcome.MethodsWe selected consecutive patients from Geneva University Hospitals who underwent resective epilepsy surgery with postsurgical follow-up > 12 months. We analyzed long-term iEEG recordings during non-rapid eye movement (NREM) sleep that we segmented into intervals of 5 min. HFOs were defined in the ripple (80-250 Hz) and the fast ripple (FR, 250-500 Hz) frequency band. Contacts with the highest rate of ripples co-occurring with FR (FRandR) designated the HFO area. If the HFO area was not fully resected and the patient suffered from recurrent seizures (ILAE 2-6), this was classified as a true positive (TP) prediction.ResultsWe included iEEG recordings from 16 patients (median age 32 y, range [18-53]) with stereotactic depth electrodes and/or with subdural electrode grids (median follow-up 27 mo, range [12-55]). The HFO area had high test-retest reliability across intervals (median dwell time 95%). We excluded two patients with dwell time < 50% from further analysis.The HFO area was fully included in the resected volume in 2/4 patients who achieved postoperative seizure freedom (ILAE 1, specificity 50%) and was not fully included in 9/10 patients with recurrent seizures (ILAE > 1, sensitivity 90%), leading to an accuracy of 79%.ConclusionsWe validated the automated procedure to delineate the clinical relevant HFO area in individual patients of an independently recorded dataset and achieved the same good accuracy as in our previous studies.SignificanceThe reproducibility of our results across datasets is promising for a multicienter study testing the clinical application of HFO detection to guide epilepsy surgery.

scholarly journals Seizure Freedom After Epilepsy Surgery and Higher Baseline Cognition May Be Associated With a Negatively Correlated Epilepsy Network in Temporal Lobe Epilepsy

2021 ◽  
Vol 14 ◽  
Author(s):  
Elliot G. Neal ◽  
Mike R. Schoenberg ◽  
Stephanie Maciver ◽  
Yarema B. Bezchlibnyk ◽  
Fernando L. Vale
Keyword(s):  
Temporal Lobe Epilepsy ◽  
Temporal Lobe ◽  
Epilepsy Surgery ◽  
Memory Function ◽  
Brain Regions ◽  
Seizure Freedom ◽  
Epileptogenic Zone ◽  
Temporal Lobes ◽  
Scalp Eeg ◽  
The Brain

Background: Brain regions positively correlated with the epileptogenic zone in patients with temporal lobe epilepsy vary in spread across the brain and in the degree of correlation to the temporal lobes, thalamus, and limbic structures, and these parameters have been associated with pre-operative cognitive impairment and seizure freedom after epilepsy surgery, but negatively correlated regions have not been as well studied. We hypothesize that connectivity within a negatively correlated epilepsy network may predict which patients with temporal lobe epilepsy will respond best to surgery.Methods: Scalp EEG and resting state functional MRI (rsfMRI) were collected from 19 patients with temporal lobe epilepsy and used to estimate the irritative zone. Using patients’ rsfMRI, the negatively correlated epilepsy network was mapped by determining all the brain voxels that were negatively correlated with the voxels in the epileptogenic zone and the spread and average connectivity within the network was determined.Results: Pre-operatively, connectivity within the negatively correlated network was inversely related to the spread (diffuseness) of that network and positively associated with higher baseline verbal and logical memory. Pre-operative connectivity within the negatively correlated network was also significantly higher in patients who would go on to be seizure free.Conclusion: Patients with higher connectivity within brain regions negatively correlated with the epilepsy network had higher baseline memory function, narrower network spread, and were more likely to be seizure free after surgery.

scholarly journals Long-range phase synchronization of high-frequency oscillations in human cortex

2020 ◽  
Vol 11 (1) ◽  
Author(s):  
G. Arnulfo ◽  
S. H. Wang ◽  
V. Myrov ◽  
B. Toselli ◽  
J. Hirvonen ◽  
...  
Keyword(s):  
High Frequency ◽  
Large Scale ◽  
Phase Synchronization ◽  
Brain Activity ◽  
Brain Regions ◽  
Spike Timing ◽  
Human Cortex ◽  
Neuronal Communication ◽  
High Frequency Oscillations ◽  
Connectivity Patterns

Abstract Inter-areal synchronization of neuronal oscillations at frequencies below ~100 Hz is a pervasive feature of neuronal activity and is thought to regulate communication in neuronal circuits. In contrast, faster activities and oscillations have been considered to be largely local-circuit-level phenomena without large-scale synchronization between brain regions. We show, using human intracerebral recordings, that 100–400 Hz high-frequency oscillations (HFOs) may be synchronized between widely distributed brain regions. HFO synchronization expresses individual frequency peaks and exhibits reliable connectivity patterns that show stable community structuring. HFO synchronization is also characterized by a laminar profile opposite to that of lower frequencies. Importantly, HFO synchronization is both transiently enhanced and suppressed in separate frequency bands during a response-inhibition task. These findings show that HFO synchronization constitutes a functionally significant form of neuronal spike-timing relationships in brain activity and thus a mesoscopic indication of neuronal communication per se.

scholarly journals Stable high frequency background EEG activity distinguishes epileptic from healthy brain regions

2020 ◽  
Vol 2 (2) ◽  
Author(s):  
Annika Minthe ◽  
Wibke G Janzarik ◽  
Daniel Lachner-Piza ◽  
Peter Reinacher ◽  
Andreas Schulze-Bonhage ◽  
...  
Keyword(s):  
High Frequency ◽  
Background Activity ◽  
Sleep Stage ◽  
Strong Association ◽  
Brain Regions ◽  
High Frequency Oscillation ◽  
Seizure Onset ◽  
High Frequency Oscillations ◽  
Seizure Onset Zone ◽  
Low Entropy

Abstract High-frequency oscillations are markers of epileptic tissue. Recently, different patterns of EEG background activity were described from which high-frequency oscillations occur: high-frequency oscillations with continuously oscillating background were found to be primarily physiological, those from quiet background were linked to epileptic tissue. It is unclear, whether these interactions remain stable over several days and during different sleep-wake stages. High-frequency oscillation patterns (oscillatory vs. quiet background) were analysed in 23 patients implanted with depth and subdural grid electrodes. Pattern scoring was performed on every channel in 10 s intervals in three separate day- and night-time EEG segments. An entropy value, measuring variability of patterns per channel, was calculated. A low entropy value indicated a stable occurrence of the same pattern in one channel, whereas a high value indicated pattern instability. Differences in pattern distribution and entropy were analysed for 143 280 10 s intervals with allocated patterns from inside and outside the seizure onset zone, different electrode types and brain regions. We found a strong association between high-frequency oscillations out of quiet background activity, and channels of the seizure onset zone (35.2% inside versus 9.7% outside the seizure onset zone, P &lt; 0.001), no association was found for high-frequency oscillations from continuous oscillatory background (P = 0.563). The type of background activity remained stable over the same brain region over several days and was independent of sleep stage and recording technique. Stability of background activity was significantly higher in channels of the seizure onset zone (entropy mean value 0.56 ± 0.39 versus 0.64 ± 0.41; P &lt; 0.001). This was especially true for the presumed epileptic high-frequency oscillations out of quiet background (0.57 ± 0.39 inside versus 0.72 ± 0.37 outside the seizure onset zone; P &lt; 0.001). In contrast, presumed physiological high-frequency oscillations from continuous oscillatory backgrounds were significantly more stable outside the seizure onset zone (0.72 ± 0.45 versus 0.48 ± 0.53; P &lt; 0.001). The overall low entropy values suggest that interactions between high-frequency oscillations and background activity are a stable phenomenon specific to the function of brain regions. High-frequency oscillations occurring from a quiet background are strongly linked to the seizure onset zone whereas high-frequency oscillations from an oscillatory background are not. Pattern stability suggests distinct underlying mechanisms. Analysing short time segments of high-frequency oscillations and background activity could help distinguishing epileptic from physiologically active brain regions.

Sign in / Sign up

Export Citation Format

Share Document