Anomaly Detection in EEG Signals: A Case Study on Similarity Measure

Abstract

Motivation. Anomaly EEG detection is a long-standing problem in analysis of EEG signals. The basic premise of this problem is consideration of the similarity between two nonstationary EEG recordings. A well-established scheme is based on sequence matching, typically including three steps: feature extraction, similarity measure, and decision-making. Current approaches mainly focus on EEG feature extraction and decision-making, and few of them involve the similarity measure/quantification. Generally, to design an appropriate similarity metric, that is compatible with the considered problem/data, is also an important issue in the design of such detection systems. It is however impossible to directly apply those existing metrics to anomaly EEG detection without any consideration of domain specificity. Methodology. The main objective of this work is to investigate the impacts of different similarity metrics on anomaly EEG detection. A few metrics that are potentially available for the EEG analysis have been collected from other areas by a careful review of related works. The so-called power spectrum is extracted as features of EEG signals, and a null hypothesis testing is employed to make the final decision. Two indicators have been used to evaluate the detection performance. One is to reflect the level of measured similarity between two compared EEG signals, and the other is to quantify the detection accuracy. Results. Experiments were conducted on two data sets, respectively. The results demonstrate the positive impacts of different similarity metrics on anomaly EEG detection. The Hellinger distance (HD) and Bhattacharyya distance (BD) metrics show excellent performances: an accuracy of 0.9167 for our data set and an accuracy of 0.9667 for the Bern-Barcelona EEG data set. Both of HD and BD metrics are constructed based on the Bhattacharyya coefficient, implying the priority of the Bhattacharyya coefficient when dealing with the highly noisy EEG signals. In future work, we will exploit an integrated metric that combines HD and BD for the similarity measure of EEG signals.

Identifying Individuals With Mild Cognitive Impairment Using Working Memory-Induced Intra-Subject Variability of Resting-State EEGs

“…It may be possible to improve classification performance further using other types of similarity measures. For example, in a recent study, Hellinger distance and Bhattacharyya distance showed their effectiveness with highly noisy EEG signals (Chen G. et al, 2020 ). Although comparisons between different approaches of similarity measures are beyond the scope of this study, they certainly merit attention in evaluating the effectiveness of using different between-run similarity measures as neurophysiological features for classifying neurodegenerative diseases.…”

“…Since working memory tasks are presumably more cognitively exhausted for the MCI or AD group than the HC group, we hypothesize that the difference in the neurophysiological patterns of the before-task and after-task “resting-state” in the brain will be larger for the MCI group than the HC group, and such difference carries more discriminative information for classification in comparison with the approach using single-run resting-state EEGs that has been adopted in previous studies related to the MCI-HC classification. To achieve this goal, we designed a novel feature extraction framework in which we introduced the delayed matching-to-sample (DSTM) task as a cognitively challenging behavior test, applied a similarity-based approach (Chen G. et al, 2020 ) to quantitatively evaluate the task-induced intra-subject variation of resting-state EEG powers, and used it as a neural marker to classify between the MCI and HC groups. To the best knowledge of the authors, this is the first study that focuses on the analysis of task-induced intra-subject variability between two separate runs of resting-state EEGs for the detection of MCI.…”

A statistical approach for the automatic identification of the start of the chain of events leading to the disruptions at JET

“…In particular, the proposed method can be framed into the anomaly and point change detection research areas. In a broad sense, 'anomaly detection refers to the problem of finding patterns in data that do not conform to expected behavior' [21] and it has found application in various fields and for a long time [22][23][24][25]. Anomaly detection is a highly applicationoriented problem [26] and several approaches have been proposed to solve it, depending on research disciplines (e.g.…”

E2SGAN: EEG-to-SEEG translation with generative adversarial networks

“…The correlation does not vary with distance in a linear trend as is habitually deemed, but rather a non-linear fashion. To demonstrate this counterintuitive relation, the Hellinger Distance (Chen et al, 2020 ) from two SEEG channels and the target EEG channel are compared as shown in Figure 3 . The leftmost column gives an example of what a matching strategy is expected to solve.…”