Automated Seizure Onset Zone Approximation Based on Nonharmonic High-Frequency Oscillations in Human Interictal Intracranial EEGs

Abstract

A novel automated algorithm is proposed to approximate the seizure onset zone (SOZ), while providing reproducible output. The SOZ, a surrogate marker for the epileptogenic zone (EZ), was approximated from intracranial electroencephalograms (iEEG) of nine people with temporal lobe epilepsy (TLE), using three methods: (1) Total ripple length (TRL): Manually segmented high-frequency oscillations, (2) Rippleness (R): Area under the curve (AUC) of the autocorrelation functions envelope, and (3) Autoregressive model residual variation (ARR, novel algorithm): Time-variation of residuals from autoregressive models of iEEG windows. TRL, R, and ARR results were compared in terms of separability, using Kolmogorov-Smirnov tests, and performance, using receiver operating characteristic (ROC) curves, to the gold standard for SOZ delineation: visual observation of ictal video-iEEGs. TRL, R, and ARR can distinguish signals from iEEG channels located within the SOZ from those outside it (p < 0.01). The ROC AUC was 0.82 for ARR, while it was 0.79 for TRL, and 0.64 for R. ARR outperforms TRL and R, and may be applied to identify channels in the SOZ automatically in interictal iEEGs of people with TLE. ARR, interpreted as evidence for nonharmonicity of high-frequency EEG components, could provide a new way to delineate the EZ, thus contributing to presurgical workup.

MUSIC-Expected maximization gaussian mixture methodology for clustering and detection of task-related neuronal firing rates

“…This methodology is general and could be used in any situation where multiple neurons are recorded, whether there be a single type of trial or multiple trial types, as considered here. For example, this application could be used in the study of motor commands for gaze control (Takemura et al, 2001), neuronal oscillators (Schwemmer and Lewis, 2011; Geertsema et al, 2015; Aydin et al, 2016) or other brain studies such as deep brain stimulation (Florin et al, 2013). Further, it could be used for research in brain-computer interfaces (Carmena et al, 2003;Ortiz-Rosario and Adeli, 2013).…”

To resect or not to resect? Unbiased performances of single and combined biomarkers in intra-operative corticography for tailoring during epilepsy surgery

“…The auto-regressive residual modulation (ARRm) provides the amount of non-harmonicity in the signal quantified as the high residual variation after auto-regressive modelling 30,31 . It has been shown that brain tissue with high non-harmonicity corresponds to areas with high frequency oscillations (HFOs) which in turn may be an indication of epileptogenic tissue [38][39][40][41][42] .…”

“…Given as a signal we divided the signal in overlapping windows . For each window it is possible to compute the auto-regressive model of order (in Geertsema's work 30 where is the residual variation for window for model order 3. Geertsema et al 31 suggested an improved version of .…”

“…We investigated, in our ground-truth scenario, the performances of different biomarkers that were already used in previous studies [23][24][25][27][28][29][30][31][32][33][34][35] . We defined our biomarker pool with the attempt to be exhaustive according to the three different categories univariate, bivariate and multivariate biomarkers.…”

Unsupervised Detection of High-Frequency Oscillations Using Time-Frequency Maps and Computer Vision

“…In the years that followed, many studies revealed that HF oscillations (HFOs) in the 80-250-Hz frequency range, also known as "ripples, " can be identified in the hippocampal and parahippocampal regions of rodents (Buzsáki et al, 1992;Ylinen et al, 1995), primates (Skaggs et al, 2007), and humans (Bragin et al, 1999;Matsumoto et al, 2013). These studies identified HFOs in healthy subjects or in epileptic patients performing visual or motor tasks (Matsumoto et al, 2013), while other studies found an increased number of HFOs in brain regions that are part of the epileptogenic network; therefore, the HFO was considered to be a potential epilepsy biomarker (Urrestarazu et al, 2007;Jacobs et al, 2009;Brázdil et al, 2010;Kerber et al, 2013;Geertsema et al, 2015). However, in spite of the various existing models of HFO generation (Fink et al, 2015;Helling et al, 2015), it is still a matter of debate how to distinguish physiological and pathological HFOs (Engel et al, 2009;Waldman et al, 2018).…”