Cardio-respiratory contributions to the magnetoencephalogram

Splitting of the magnetic encephalogram into «brain» and «non-brain» physiological signals based on the joint analysis of frequency-pattern functional tomograms and magnetic resonance images

The article considers the problem of dividing the encephalography data into two time series, that generated by the brain and that generated by other electrical sources located in the human head. The magnetic encephalograms and magnetic resonance images of the head were recorded in the Center for Neuromagnetism at NYU Grossman School of Medicine. Data obtained at McGill University and Montreal University were also used. Recordings were made in a magnetically shielded room and the gradiometers were designed to suppress external noise, making it possible to eliminate them from the data analysis. Magnetic encephalograms were analyzed by the method of functional tomography, based on the Fourier transform and on the solution of inverse problem for all frequencies. In this method, one spatial position is assigned to each frequency component. Magnetic resonance images of the head were evaluated to annotate the space to be included in the analysis. The included space was divided into two parts: «brain» and «non-brain». The frequency components were classified by the feature of their inclusion in one or the other part. The set of frequencies, designated as «brain», represented the partial spectrum of the brain signal, while the set of frequencies designated as «non-brain», represented the partial spectrum of the physiological noise produced by the head. Both partial spectra shared the same frequency band. From the partial spectra, a time series of the «brain» area signal and «non-brain» area head noise were reconstructed. Summary spectral power of the signal was found to be ten times greater than the noise. The proposed method makes it possible to analyze in detail both the signal and the noise components of the encephalogram and to filter the magnetic encephalogram.

Does stroke volume influence heartbeat evoked responses?

We know surprisingly little on how heartbeat-evoked responses (HERs) vary with cardiac parameters. Here, we measured both stroke volume, or volume of blood ejected at each heartbeat, with impedance cardiography, and HER amplitude with magneto-encephalography, in 21 male and female participants at rest with eyes open. We observed that HER co-fluctuates with stroke volume on a beat-to-beat basis, but only when no correction for cardiac artifact was performed. This highlights the importance of an ICA correction tailored to the cardiac artifact. We also observed that easy-to-measure cardiac parameters (interbeat intervals, ECG amplitude) are sensitive to stroke volume fluctuations and can be used as proxies when stroke volume measurements are not available. Finally, interindividual differences in stroke volume were reflected in MEG data, but whether this effect is locked to heartbeats is unclear. Altogether, our results question assumptions on the link between stroke volume and HERs.

Removing Cardiac Artefacts in Magnetoencephalography with Resampled Moving Average Subtraction

Magnetoencephalography (MEG) signals are commonly contaminated by cardiac artefacts (CAs). Principle component analysis and independent component analysis have been widely used for removing CAs, but they typically require a complex procedure for the identification of CA-related components. We propose a simple and efficient method, resampled moving average subtraction (RMAS), to remove CAs from MEG data. Based on an electrocardiogram (ECG) channel, a template for each cardiac cycle was estimated by a weighted average of epochs of MEG data over consecutive cardiac cycles, combined with a resampling technique for accurate alignment of the time waveforms. The template was subtracted from the corresponding epoch of the MEG data. The resampling reduced distortions due to asynchrony between the cardiac cycle and the MEG sampling times. The RMAS method successfully suppressed CAs while preserving both event-related responses and high-frequency (>45 Hz) components in the MEG data.

Optimal evaluation of digital electroencephalograms

Currently available digital EEG equipment provides considerably greater opportunities for clinical data analysis than is generally appreciated especially when appropriate software is used. Data from 7 different laboratories that had been obtained for routine diagnostic evaluations on 7 different EEG instruments and stored on compact disks were investigated. Since the instruments do not filter the data at input, ultra slow activity down to 0.01 Hz is currently being recorded but the attenuation factor is instrument dependent. Nevertheless, relevant clinical information is potentially available in these data and needs to be explored. Several examples in regard to epilepsy are presented. Determination of seizure onset may depend on the frequencies that are examined. The use of appropriate filter settings and viewing windows for the clinical question to be answered is stressed. Differentiation between simple and complex spike wave discharges, as well as spread of spikes, can readily be achieved by expanding the time base to 1 or 2 seconds and placing a cursor on the peak of the negative spike. Latencies in the millisecond range can then become apparent. EEGs co-registered with MEG should be evaluated with the same software in order to allow an adequate assessment of the similarities and differences between electrical and magnetic activity. An example of a comparison of EEG, planar gradiometers and magnetometers for an averaged spike is shown.

Cerebral Electromagnetic Activity in the Subdelta Range

The frequency range between 0.1 and 0.9 Hz was investigated with magnetoelectroencephalography-EEG coregistration in 10 adult patients with epilepsy and five children with other neurologic conditions. In all instances, a dominant rhythm between 0.2 and 0.4 Hz could be observed in the waking and sleeping states. It showed a waxing and waning quality and was unrelated to eye opening or closing but increased in amplitude during sleep. The maximum was usually in the occipital areas but occasionally in the frontal regions. The rhythm was more persistent and better seen in the magnetoelectroencephalogram, but subdelta activity was also discernible in the EEG. The magnetoelectroencephalographic rhythmicity and frequency suggested possible respiration artifact. Two normal control subjects were therefore investigated by electroencephalography while respirations were monitored. A clear relation to respiration was established. It persisted during breath-holding, albeit at lower amplitude. Larger amplitude transients occurred before and at the cessation of breath-holding as well as hyperventilation. An observed frequency increase before voluntary hyperventilation suggests a relation to the readiness potential and event-related desynchronization as well as synchronization. Subdelta frequencies, which can be readily recorded without special DC amplifiers, provide additional information for clinical as well as research data. They may also be an interface between autonomic and voluntary functions, especially in regard to respiration.