Improving the ripple classification in focal pediatric epilepsy: identifying pathological high-frequency oscillations by Gaussian mixture model clustering

Abstract

Objective. High-frequency oscillations (HFOs) have emerged as a promising clinical biomarker for presurgical evaluation in childhood epilepsy. HFOs are commonly classified in stereo-encephalography as ripples (80–200 Hz) and fast ripples (200–500 Hz). Ripples are less specific and not so directly associated with epileptogenic activity because of their physiological and pathological origin. The aim of this paper is to distinguish HFOs in the ripple band and to improve the evaluation of the epileptogenic zone (EZ). Approach. This study constitutes a novel modeling approach evaluated in ten patients from Sant Joan de Deu Pediatric Hospital (Barcelona, Spain), with clearly-defined seizure onset zones (SOZ) during presurgical evaluation. A subject-by-subject basis analysis is proposed: a probabilistic Gaussian mixture model (GMM) based on the combination of specific ripple features is applied for estimating physiological and pathological ripple subpopulations. Main Results. Clear pathological and physiological ripples are identified. Features differ considerably among patients showing within-subject variability, suggesting that individual models are more appropriate than a traditional whole-population approach. The difference in rates inside and outside the SOZ for pathological ripples is significantly higher than when considering all the ripples. These significant differences also appear in signal segments without epileptiform activity. Pathological ripple rates show a sharp decline from SOZ to non-SOZ contacts and a gradual decrease with distance. Significance. This novel individual GMM approach improves ripple classification and helps to refine the delineation of the EZ, as well as being appropriate to investigate the interaction of epileptogenic and propagation networks.

Exploring patient medication adherence and data mining methods in clinical big data: A contemporary review

“…255 This showcases the interdependency between samples and feature clustering outcomes. The biomedical signal processing realm has seen these clustering models grow prominent in signal classification [256][257][258][259][260][261][262] and medical image segmentation. [263][264][265] As deep learning has progressed, an increasing number of researchers have harnessed neural networks for feature extraction and optimized the process of feature extraction and clustering by combining deep neural networks (DNNs) with clustering algorithms.…”

EMHapp: a pipeline for the automatic detection, localization and visualization of epileptic magnetoencephalographic high-frequency oscillations

“…In this study, HFOs that spatially and temporally co-occurred with IEDs [13] were automatically detected by using a two-stage-based approach (see figure 1). HFO candidate events were first detected via a threshold-based detector [25,26], and false-positive HFO events were further removed with a clusteringbased detector [13,41,42]. The details of the detection method are described below.…”

“…HFO segments exhibiting at least 50% overlap in time were considered one HFO candidate event [25]. The HFO segments were clustered with a Gaussian mixture model [13,41,42] based on these five features, and the optimal number of clusters was determined by the elbow method [44] (see text S4 in the supplementary materials for more details). The HFO segments derived from clusters with less apparent HFO features were manually discarded.…”

Unsupervised Multivariate Feature-Based Adaptive Clustering Analysis of Epileptic EEG Signals

“…Wan et al combined several signal analysis methods, and the epileptic EEG signals were processed by Stockwell's positive inverse transform and singular value decomposition, respectively, and four features were extracted and analyzed by clustering with the improved Fuzzy C-means algorithm [18]. Carolina et al applied the Gaussian Mixture Model clustering method to evaluate the EEG waveforms of pediatric epileptic patients [19].…”