Clinical neuroscienceTime-frequency analysis of resting state and evoked EEG data recorded at higher magnetic fields up to 9.4Â T

- Pulse artefact removal in EEG at ultra-high fields still requires improvement.
- Combining mutual information with ICA improves PA rejection.
- Time-frequency analysis of resting state and evoked EEG at 9.4 T becomes possible.
- Our approach revealed reliable results w.r.t. EEG hybrid imaging at higher fields.

BackgroundCombining both high temporal and spatial resolution by means of simultaneous electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) is of relevance to neuroscientists. This combination, however, leads to a distortion of the EEG signal by the so-called cardio-ballistic artefacts. The aim of the present study was developing an approach to restore meaningful physiological EEG data from recordings at different magnetic fields.New methodsThe distortions introduced by the magnetic field were corrected using a combination of concepts from independent component analysis (ICA) and mutual information (MI). Thus, the components were classified as either related to the cardio-ballistic artefacts or to the signals of interest. EEG data from two experimental paradigms recorded at different magnetic field strengths up to 9.4 T were analyzed: (i) spontaneous activity using an eyes-open/eyes-closed alternation, and (ii) responses to auditory stimuli, i.e. auditory evoked potentials.ResultsEven at ultra-high magnetic fields up to 9.4 T the proposed artefact rejection approach restored the physiological time-frequency information contained in the signal of interest and the data were suitable for subsequent analyses.Comparison with existing methodsBlind source separation (BSS) has been used to retrieve information from EEG data recorded inside the MR scanner in previous studies. After applying the presented method on EEG data recorded at 4 T, 7 T, and 9.4 T, we could retrieve more information than from data cleaned with the BSS method.ConclusionsThe present work demonstrates that EEG data recorded at ultra-high magnetic fields can be used for studying neuroscientific research question related to oscillatory activity.