Nicole Ille

Artifact correction of the ongoing EEG using spatial filters based on artifact and brain signal topographies.

Summary Review and analysis of continuous EEG recordings may be impeded by physiological artifacts such as blinks, eye movements, or cardiac activity. Spatial filters based on artifact and brain signal topographies can remove artifacts completely without distortion of relevant brain activity. The authors describe the basic principle of artifact correction by spatial filtering and they review different approaches to estimate artifact and brain signal topographies. The main focus is on the preselection approach, which is fast enough to be applied while paging through the segments of a digital EEG recording. Examples of real EEG segments, containing epileptic seizure activity or interictal spikes contaminated by artifacts, show that spatial filtering by preselection can be a useful tool during EEG review. Advantages and disadvantages of the different spatial filter approaches are discussed.

BESA Source Coherence: A New Method to Study Cortical Oscillatory Coupling

This paper introduces source coherence, a new method for the analysis of cortical coherence using noninvasive EEG and MEG data. Brain electrical source analysis (BESA) is applied to create a discrete multiple source model. This model is used as a source montage to transform the recorded data from sensor level into brain source space. This provides source waveforms of the modeled brain regions as a direct measure for their activities on a single trial basis. The source waveforms are transformed into time-frequency space using complex demodulation. Magnitude-squared coherence between the brain sources reveals oscillatory coupling between sources. This procedure allows one to separate the time-frequency content of different brain regions even if their activities severely overlap at the surface. Thus, source coherence overcomes problems of localization and interpretation that are inherent to coherence analysis at sensor level. The principle of source coherence is illustrated using an EEG recording of an error-related negativity as an example. In this experiment the subject performed a visuo-motor task. Source coherence analysis revealed dynamical linking between posterior and central areas within the gamma-band around the time of button press at a post-stimulus latency of 200-300 ms.

Deconvolution of 40 Hz steady-state fields reveals two overlapping source activities of the human auditory cortex

Steady-state auditory evoked fields were recorded from 15 subjects using a whole head MEG system. Stimuli were 800 ms trains of binaural clicks with constant stimulus onset asynchrony (SOA). Seven different SOA settings (19, 21, 23, 25, 27, 29 and 31 ms) were used to give click rates near 40 Hz. Transient responses to each click were reconstructed using a new algorithm that deconvoluted the averaged responses to the different trains. Spatio-temporal multiple dipole modelling in relation to 3D MRI scans revealed two overlapping source components in both the left and right auditory cortex. The primary sources in the medial part of Heschls gyrus exhibited a N19-P30-N40 m pattern. The secondary, weaker sources at more lateral sites on Heschls gyrus showed a N24-P36-N46 m pattern. When applied to transient middle latency auditory evoked fields (MAEFs) recorded at SOAs of 95-135 ms, the primary sources imaged activities similar to the deconvoluted steady-state responses, but the secondary source activities were inconsistent. Linear summation of the deconvoluted source waveforms accounted for more than 96% of the steady-state variance. This indicates that the primary activity of the auditory cortex remains constant up to high stimulation rates and is not specifically enhanced around 40 Hz.

Advanced tools for digital EEG review: Virtual source montages, whole-head mapping, correlation, and phase analysis

Summary Digital EEG allows one to combine recorded EEG channels into new montages without the need to record new data. Using spherical splines, voltages can be estimated at any point on the head. This allows one to generate various montages with the recorded or virtual electrodes at standardized locations, to interpolate bad electrodes, and to generate topographic maps over the whole head. Simulations of EEG activity originating in various brain regions are used to illustrate the effects of known generators on various montages and on whole-head maps. Some properties of spatial filters are introduced, and it is shown how they can be used to develop source montages with signals that estimate the activity in specific brain regions. The usefulness and validity of a source montage designed to focus on temporal lobe activity is illustrated with simulations and examples of temporal lobe spikes and seizures. Additional tools such as cross-correlation among channels, fast Fourier transform, and phase maps are described. These tools are useful in estimating time lags between source channels and in interpreting propagating spike and seizure activity. In combination, these tools help to analyze and to enhance activities that may be hard to detect from the background scalp EEG in traditional montages.

The correction of ocular artifacts: a topographic perspective

OBJECTIVE To evaluate the scalp topography of the potentials related to saccades and blinks. METHODS The scalp topographies of the potentials associated with saccades and blinks were recorded in 60 subjects. The topographies were analyzed using both source components and attenuation factors, with each factor representing the fraction of the potential recorded in peri-ocular electrodes that contributes to the EEG recorded from a particular scalp location. RESULTS Blinks and upward saccades generated potentials with very different topographies. Left and right saccades and up and down saccades generated equal but inverted fields except at peri-ocular locations where subtle inequalities occurred. The potentials associated with lateral saccades were consistently larger in female subjects than in male subjects. CONCLUSIONS The differences in the scalp topographies between blinks and vertical saccades can be explained by the different ways in which they are generated. Blink potentials are caused by the eyelids sliding down over the positively charged cornea, whereas saccade potentials are caused by changes in the orientation of the corneoretinal dipole. Any compensation procedure for ocular artifacts must take into account the topographic differences between blinks and upward saccades.

MEG versus EEG: influence of background activity on interictal spike detection.

The comparative sensitivity of EEG and magnetoencephalography (MEG) in the visual detection of focal epileptiform activity in simultaneous interictal sleep recordings were investigated. The authors examined 14 patients aged 3.5 to 17 years with localization-related epilepsy. Simultaneous 122-channel whole-head MEG and 33-channel EEG were recorded for 20 to 40 minutes during spontaneous sleep. The EEG and MEG data were separated and four blinded independent reviewers marked the presence and timing of epileptic discharges (ED) in the 28 data segments. EEG and MEG data were matched and spikes identified by at least three reviewers were classified in three categories according to the following criteria: type 1 MEG > EEG, type 2 EEG > MEG (type 1/2: difference of three or more raters), and type 3 EEG = MEG (three or more raters each). The presence of simultaneous sleep changes was visually determined for every single EEG-segment. Spikes with high spatiotemporal correlation were averaged and subjected to single dipole analysis of peak activity in EEG. Out of 4704 marked patterns, 1387 spikes fulfilled the above criteria. In fact, more spikes were unique to MEG (689) than to EEG (136) and to the combination of both modalities (562). ED were detected predominantly by MEG in eight patients and by EEG in two patients. The presence of vertex waves and spindles lead to a significantly higher number of spikes identified only in MEG. Averaging of type 1 spikes produced clear spike activity in EEG in 9 of 12 cases. On the contrary, only 2 of 10 type 2 spikes were visible in MEG after averaging. Dipoles of spikes visible in MEG showed a more tangential orientation compared with more radial dipoles of type 2 spikes. Spike characteristics, e.g., dipole orientation, are a key factor for a sole EEG representation. Exclusive MEG detection is more likely influenced by overlapping background activity in EEG. Because MEG is indifferent to radial activity, i.e., sleep changes, a higher ratio of spikes unique to MEG compared with EEG is detected in the case of overlapping sleep changes.

Multimodal functional imaging of prolonged neurological deficits in a patient suffering from familial hemiplegic migraine

The case of a patient with familial hemiplegic migraine (FHM) suffering from prolonged right sided hemiparesis and aphasia that persisted for more than 10 days is reported. The symptoms were accompanied by slowing of the magnetoencephalogram over the left hemisphere, which normalized parallel to the clinical improvement. Positron emission tomography obtained on the 6th day revealed glucose-hypometabolism (hemispheric difference > or =10%) in left hemispheres fronto-basal cortex, caudate nucleus, and thalamus. In contrast, magnetic resonance imaging including perfusion and diffusion weighted imaging was normal and did not show significant alterations of cortical perfusion or water mobility during the episode. We hypothesize that this finding provides evidence for a primary neuronal dysfunction causing the prolonged neurological deficits in FHM.

Fast evaluation of interictal spikes in long-term EEG by hyper-clustering

Purpose:  The burden of reviewing long‐term scalp electroencephalography (EEG) is not much alleviated by automated spike detection if thousands of events need to be inspected and mentally classified by the reviewer. This study investigated a novel technique of clustering and 24‐h hyper‐clustering on top of automated detection to assess whether fast review of focal interictal spike types was feasible and comparable to the spikes types observed during routine EEG review in epilepsy monitoring.

P6-14 BESA epilepsy: A new clinical tool for fast evaluation of interictal spikes in long-term EEG

Objective: BESA Epilepsy uses a new hypercluster technique and a file independent data management to combine similar events over 24 hours of EEG. Each day, the physician inspects the hyperclusters and decides whether they are epileptiform or not. Optimized source waveform and EEG segment displays, 3D maps, and localization in a head scheme allow for fast decision and assessment of the region of origin. Methods: A new spike detection and clustering algorithm based on an EEG transformation into 29 regional brain sources was developed. Clusters were calculated in 2 hour epochs and combined into daily hyperclusters using empirical rules on similarity in waveshape and topography. 24 hour EEG data of 44 epilepsy patients (21 children) were evaluated by independent raters using traditional visual inspection versus fast hypercluster evaluation. Results: Visual rating yielded 107 epileptiform spike types. Hypercluster rating agreed in 85% (temporal lobe spikes 94%, extratemporal 78%). In a 24 h recording, about 15 25 hyperclusters had to be inspected to decide whether they reflected epileptiform discharges, artifacts, or other EEG patterns. Due to the rapid inspection tools, the decision and reporting process was typically completed within 5 minutes by experienced physicians. Conclusion: The traditional hourly evaluation of 2 5 minute epochs of long-term EEG can be readily supplemented by a computer-based hypercluster evaluation. This adds a fast, comprehensive overview and report, an independent control of the existence of one or multiple spike foci, and an estimation of their origin. The involvement of the physician in the decision process allowed to increase sensitivity at the cost of reviewing and rejecting more artifact hyperclusters. Longterm EEG evaluation is facilitated by a file independent graphical data management. This enables the rapid clinical inspection of the same, fixed 24 h interval every day without the time-consuming need to open and close different EEG files.