engleski

Evaluation of MUSIC-based initial source modeling assumptions for spatio-temporal fitting of simulated multi-source neuromagnetic data

A series of simulations was conducted to examine effects of a) multi-source configurations ; b)temporal correlations ; c) noise ; and d) assumed signal subspace dimensionality on the number of the observed peaks and their peak values, in order to evaluate the capabilities and limitations of a MUSIC based approach to localize multiple dipoles. The MUSIC algorithm in which multiple dipole locations are found by scanning potential locations using a simple 1-dipole model was inadequate for several of our multi-source configurations. MUSIC tends to provide fewer that the actual number of sources due to the inherent inability of SVD based appraches to resolve synchronous activity and some asynchronous activity under noisy conditions.

magnetoencephalography (MEG); MUSIC algoritham; numerical simulations; source modeling; multi-source resolution

nije evidentirano