Complementary Methods for Interpreting Brain Signals: Linear versus Nonlinear Techniques

Magnetoencephalography (MEG) brain signals are characterized using both linear and nonlinear dynamical methods. The linear approach employs the power analysis in a spatial visualization. The nonlinear approach estimates the value of d_infin (d-infinite) to characterize the system´s asymptotic chaotic behavior using a computationally less onerous method than the conventional one for d_infin. Both methods are applied here to study a female patient with obsessive compulsive disorder and an age-sex matched normal subject. MEG time series were obtained using dual 37-channel bio-magnetometers (4-D Neuroimaging, San Diego, CA).