Inferring spatiotemporal network patterns from intracranial EEG data

Statistical Model-Based Classification to Detect Patient-Specific Spike-and-Wave in EEG Signals

Spike-and-wave discharge (SWD) pattern detection in electroencephalography (EEG) is a crucial signal processing problem in epilepsy applications. It is particularly important for overcoming time-consuming, difficult, and error-prone manual analysis of long-term EEG recordings. This paper presents a new method to detect SWD, with a low computational complexity making it easily trained with data from standard medical protocols. Precisely, EEG signals are divided into time segments for which the continuous Morlet 1-D wavelet decomposition is computed. The generalized Gaussian distribution (GGD) is fitted to the resulting coefficients and their variance and median are calculated. Next, a k-nearest neighbors (k-NN) classifier is trained to detect the spike-and-wave patterns, using the scale parameter of the GGD in addition to the variance and the median. Experiments were conducted using EEG signals from six human patients. Precisely, 106 spike-and-wave and 106 non-spike-and-wave signals were used for training, and 96 other segments for testing. The proposed SWD classification method achieved 95% sensitivity (True positive rate), 87% specificity (True Negative Rate), and 92% accuracy. These promising results set the path for new research to study the causes underlying the so-called absence epilepsy in long-term EEG recordings.

Tailor-Made Surgery Based on Functional Networks for Intractable Epilepsy

Normal and pathological networks related to seizure propagation have got attention to elucide complex seizure semiology and contribute to diagnosis and surgical monitoring in epilepsy treatment. Since focal and generalized epileptogenic syndromes abnormalities might involve multiple foci and large-scale networks, we applied electrophysiolpgy (cortco-cortico evoked potential; CCEP), and tractography to make detailed diagnosis for complex syndrome. All 14 epilepsy patients with no or little abnormality on images investigations underwent subdural grid implantation for epilepsy diagnosis. To perform quick network analysis, we recorded and analyzed high gamma activity (HGA) of epileptogenic activity and CCEPs to identify pathological activity distribution and network connectivity. [Results] Pathological CCEPs showed two negative deflections consisting of early (>40 ms) and late (>150 ms) components in electrically stable circumstance at bed side and early CCEPs appeared in 57% of the patients. On the basis of the CCEP findings, tractography detected anatomical connections. Early components of pathological CCEPs diminished after complete disconnection of tractoography-based fibers between the foci in seven of eight cases. One case with residual pathological CCEPs showed poorer outcome. Thirteen (92.8%) patients with or without CCEPs who underwent network surgery had favorable prognosis except for a case with wide traumatic epilepsy. Intraoperative CCEP measurements and HGA mapping enabled visualization of pathological networks and clinical impotence as a biomarker to improve functional prognosis. HGA/CCEP recording should shed light on pathological and complex propagation for epilepsy surgery.

Mining EEG with SVM for Understanding Cognitive Underpinnings of Math Problem Solving Strategies

We have developed a new methodology for examining and extracting patterns from brain electric activity by using data mining and machine learning techniques. Data was collected from experiments focused on the study of cognitive processes that might evoke different specific strategies in the resolution of math problems. A binary classification problem was constructed using correlations and phase synchronization between different electroencephalographic channels as characteristics and, as labels or classes, the math performances of individuals participating in specially designed experiments. The proposed methodology is based on using well-established procedures of feature selection, which were used to determine a suitable brain functional network size related to math problem solving strategies and also to discover the most relevant links in this network without including noisy connections or excluding significant connections.

Display of consistent ictal networks in refractory mesial temporal lobe epilepsy

OBJECTIVE

Exploration of emergent ictal networks was performed in homogeneous subjects with refractory medial temporal lobe epilepsy.

METHODS

Maximal Synchrony Index (SI) values were calculated for all electrode pairs for each second during 25 seizures and displayed as connectivity animations. Consistent temporal patterns of SI value and spatial connectivity were observed across seizures and subjects, and used to define a sequence of network stages.

RESULTS

Highest SI values were found in electrodes within the area of surgical resection. Analysis of these electrodes by network stage demonstrated lateral temporal cortex dominance at seizure initiation, giving way to hippocampal synchrony during the major portion of the seizure, with lateral temporal regions re-emerging as the seizure terminated. SI values also corresponded to behavioral severity of seizures, and lower SI values were associated with post-surgical seizure freedom.

CONCLUSION

SI based methods of network characterization consistently display the intrinsic MTLE ictal network and may be sensitive to clinical features.

SIGNIFICANCE

Consistency of EEG-derived network patterns is an important step as network features are applied towards improvement of clinical management. These data confirm consistency of network patterns within and across subjects and support the potential for these methods to distinguish relevant clinical variables.

A note on the phase locking value and its properties

We investigate the properties of the Phase Locking Value (PLV) and the Phase Lag Index (PLI) as metrics for quantifying interactions in bivariate local field potential (LFP), electroencephalography (EEG) and magnetoencephalography (MEG) data. In particular we describe the relationship between nonparametric estimates of PLV and PLI and the parameters of two distributions that can both be used to model phase interactions. The first of these is the von Mises distribution, for which the sample PLV is a maximum likelihood estimator. The second is the relative phase distribution associated with bivariate circularly symmetric complex Gaussian data. We derive an explicit expression for the PLV for this distribution and show that it is a function of the cross-correlation between the two signals. We compare the bias and variance of the sample PLV and the PLV computed from the cross-correlation. We also show that both the von Mises and Gaussian models are suitable for representing relative phase in application to LFP data from a visually-cued motor study in macaque. We then compare results using the two different PLV estimators and conclude that, for this data, the sample PLV provides equivalent information to the cross-correlation of the two complex time series.

Characteristic profiles of high gamma activity and blood oxygenation level-dependent responses in various language areas

High gamma activity (HGA) has been shown to be positively correlated with blood oxygenation level-dependent (BOLD) responses in the primary cortices with simple tasks. It is, however, an open question whether the correlation is simply applied to the association areas related to higher cognitive functions. The aim of this study is to investigate quantitative correlation between HGA and BOLD and their spatial and temporal profiles during semantic processing. Thirteen patients with intractable epilepsy underwent fMRI and electrocorticography (ECoG) with a word interpretation task to evoke language-related responses. Percent signal change of BOLD was calculated at each site of ECoG electrode, which has power amplification of high gamma band (60-120 Hz) activity. We transformed locations of individual electrodes and brains to a universal coordination using SPM8 and made the quantitative comparisons on a template brain. HGAs were increased in several language-related areas such as the inferior frontal and middle temporal gyri and were positively correlated with BOLD responses. The most striking finding was different temporal dynamics of HGAs in the different brain regions. Whereas the frontal lobe showed longer-lasting HGA, the HGA-intensity on the temporal lobe quickly declined. The different temporal dynamics of HGA might explain why routine language-fMRI hardly detected BOLD in the temporal lobe. This study clarified different neural oscillation and BOLD response in various brain regions during semantic processing and will facilitate practical utilization of fMRI for evaluating higher-order cognitive functions not only in basic neuroscience, but also in clinical practice.

Connectivity measures applied to human brain electrophysiological data

Connectivity measures are (typically bivariate) statistical measures that may be used to estimate interactions between brain regions from electrophysiological data. We review both formal and informal descriptions of a range of such measures, suitable for the analysis of human brain electrophysiological data, principally electro- and magnetoencephalography. Methods are described in the space-time, space-frequency, and space-time-frequency domains. Signal processing and information theoretic measures are considered, and linear and nonlinear methods are distinguished. A novel set of cross-time-frequency measures is introduced, including a cross-time-frequency phase synchronization measure.