Performance of the electroencephalography inverse problem using electric potential gradient measurements

The EEG inverse problem is traditionally solved with measurements from an array of electric potential sensors. In this work we analyze the expected performance if electric potential gradient measurements were used to solve the inverse problem. We use the Cramer-Rao bound (CRB) to analyze and compare the performance of the two cases. The spatial covariance of the noise is estimated from real EEG measurements.

[1]  P. Berg,et al.  A fast method for forward computation of multiple-shell spherical head models. , 1994, Electroencephalography and clinical neurophysiology.

[2]  R. Cohen,et al.  Body surface Laplacian ECG mapping , 1992, IEEE Transactions on Biomedical Engineering.

[3]  D. Geselowitz,et al.  Is accurate recording of the ECG surface Laplacian feasible? , 1999, IEEE Transactions on Biomedical Engineering.

[4]  P C Molenaar,et al.  Ordinary least squares dipole localization is influenced by the reference. , 1996, Electroencephalography and clinical neurophysiology.

[5]  J.C. de Munck,et al.  A random dipole model for spontaneous brain activity , 1992, IEEE Transactions on Biomedical Engineering.

[6]  H. Spekreijse,et al.  Mathematical dipoles are adequate to describe realistic generators of human brain activity , 1988, IEEE Transactions on Biomedical Engineering.

[7]  A. Nehorai,et al.  Magnetoencephalography with diversely oriented and multicomponent sensors , 1997, IEEE Transactions on Biomedical Engineering.

[8]  Z. Zhang,et al.  A fast method to compute surface potentials generated by dipoles within multilayer anisotropic spheres. , 1995, Physics in medicine and biology.