|
|
|
| e-Pub |
Previous | 
Home
Section: New Results
Coupling functional and structural models
Propagation of epileptic spikes revealed by diffusion-based constrained MEG source reconstruction
Participants : Anne-Charlotte Philippe, Théodore Papadopoulo, Christian Bénar [Hospital "La Timone", Marseille] , Jean-Michel Badier [Hospital "La Timone", Marseille] , Maureen Clerc, Rachid Deriche.
In this work, we study the propagation of an epileptic spike (from single event data). As in the two previous sections, a cortex parcellation is performed using structural information coming from diffusion MRI Then, a MEG inverse problem is defined on a parcellated source space which imposes constant activity on each parcel. This inverse problem is applied separately for measurements obtained in a given time range. The most active parcels over the time range are located and their time course are displayed. This allowed the study of the propagation of an epileptic spike via those active parcels. Results on real data shows varying spatial propagations of an epileptic spike for the same subject.
This work has been published in [41] .
Using diffusion MRI information in the Maximum Entropy on Mean framework to solve MEG/EEG inverse problem
Participants : Brahim Belaoucha, Jean-Marc Lina [Centre de Recherches Mathématique, Montréal] , Maureen Clerc, Anne-Charlotte Philippe, Christophe Grova [McGill University] , Théodore Papadopoulo.
Magnetoencephalography (MEG) and Electroencephalography (EEG) inverse problem is well-known to require regularization in order to avoid ill-posedness. Usually, regularization is based on mathematical criteria (minimum norm, ...). Physiologically, the brain is organized in functional parcels and imposing a certain homogeneity of the activity within these parcels was proven to be an efficient way to analyze the MEG/EEG data. The parcels information can be computed from diffusion Magnetic Resonances Imaging (dMRI) by grouping together source positions shared the same connectivity profile (computed as tractograms from diffusion images). In this work, three parcel-based inverse problem approaches have been tested. The first two approaches are based on minimum norm with added regularization terms to account for the parcel information. They differ by the use of a hard/soft constraint in the way they impose that the activity is constant within each parcel [74] . The third approach is based on the Maximum Entropy on Mean (MEM) framework [42] . It models source activity with a random variable and parcels are also used as a regularization. Several tests have been conducted with synthetic and real data that encompass the MEG/EEG and the diffusion magnetic resonance signals to compare these three approaches in terms of active region-detection accuracy.
This work has been published in [36] .