Follow our projects in real time on Research Gate ! Get reprints! Click here.
Characterization of networks
Presurgical evaluation of epilepsy consists in defining the extent of the epileptogenic tissues. In some cases, one can define an ‘epileptic focus’, i.e. a single well delimited regions. In many other situations, epilepsy involves large, distributed networks. We have show that independent component analysis can extract such networks non-invasively on MEG signal, and help defining a “leading” region (Malinowska et al 2014, pubmed, RG full text).
Interictal Networks on MEG and SEEG signals (Malinowska et al 2014).
Simultaneous recordings of MEG, EEG and SEEG
Most previous work on the comparison of MEG, EEG and intracerebral EEG was performed on separate recordings. However, this is far from being optimal. Indeed, only simultaneous recordings of surface and depth activity permit to investigate the exact same activity, independently of spontaneous fluctuations that can occur from a session to another (depending on subject state, medication etc…). Moreover, simultaneous recordings allow using interevent fluctuations as a source of information on the link between signals, as was done on simultaneous EEG-fMRI.
Activity evoked by visual stimulation (checkerboard) on MEG, EEG and intracerebral EEG (SEEG) recorded simultaneously (Dubarry et al 2014).
Time frequency methods for high-frequency activity
High frequency oscillations (HFOs) have been proposed as a new marker for delineating epileptic tissues. One difficulty in characterizing these activity is that they are of much smaller amplitude than classivle spike markers. We have proposed a new method to whiten the SEEG signals in order to better visualize and detect HFOs (Roehri et al TBME 2016, pubmed, RG full text).
Effect of whitening on SEEG traces (see Roehri et al 2016).
The Anywave software
We are developping within the team a software for visualizing electrophysiological traces (main developper B Colombet). This software is multiplatform and modular, and in particular permits to add easily plugins in Python or Matlab (Colombet et al 2015, Pubmed, RG full text). It is available at http://meg.univ-amu.fr/AnyWave/download.html.
Example of ICA decomposition in Anywave (MEG traces). See http://meg.univ-amu.fr/AnyWave/download.html for download.