Experimental tests of the cortical imaging technique-applications to the response to median nerve stimulation and the localization of epileptiform discharges

Abstract

In a previous paper a method for simulating the electric potentials on the surface of the brain was introduced. This method consisted of the construction of a layer of radially oriented current dipoles in a conducting sphere that simulated the head so that the voltages generated by the layer would take the values measured on the surface of the medium (the scalp). The harmonic potential function for this layer was then evaluated in the interior of the medium in an attempt to approximate the potentials that would be generated by the actual neural sources but which could not be observed without recourse to invasive recording techniques. This method, the cortical imaging technique (CIT), has been previously tested by applying it to artificially generated data where the "cortical surface" potentials were known and could be compared with CIT-generated potentials. In this paper the method is tested by applying it to the scalp-recorded potentials evoked by right median nerve stimulation, where direct cortical recordings are available for comparison, and to the scalp-recorded epileptiform discharges from two patients where the spike foci were well defined. The effects of varying the "noise ratio," an input parameter in CIT which allows one to account for noise in scalp-recorded data, is discussed.

High-resolution electro-encephalogram: source estimates of Laplacian-transformed somatosensory-evoked potentials using a realistic subject head model constructed from magnetic resonance images

“…This high-resolution EEG study presents performances of a new cortical source estimate method based on previous cortical imaging (KEARFOTT et al, 1991;SIDMAN et al, 1992;SREBRO, 1994;NUNEZ et al, 1994), linear inverse estimate ) and boundary element (MEIJS et al, 1993) techniques for the modelling of human event-related activity. As an innovation, the linear inverse source estimate of the EEG data was regularised assuming a maximally smoothed spatial distribution of the residuals vector (Ax -b), by means of the surface Laplacian transformation of the potential distribution used as an input for the regularised, weighted minimum norm linear inverse estimate (LT-WMN estimate).…”

“…Importantly, we used the well known Tikhonov regularisation technique, because other regularisation procedures are less standardised. For example, the optimum regularisation 2 value of the truncated singular value decomposition is commonly achieved by setting an arbitrary threshold, to truncate the eigenvalues of the linear system used in the linear inverse source estimate (KEARFOTT et al, 1991;SIDMAN et al, 1992;BABILONI et al, 1997).…”

“…Finally, the cortical imaging technique is commonly based on the assumption that a harmonic function associated with the actual source(s) of brain electrical activity can predict the cortical potential distribution from the scalp potential distribution (linear inverse source estimate) (FREEMAN, 1980;KEARFOTT et al,1991;SIDMAN et al, 1992;LE and GEVINS, 1993;NUNEZ et al, 1994;GEVINS et al, 1999). The linear inverse procedure for the estimation of the dipole strength can be highly underdetermined when the number of dipoles modelling the actual cortical source(s) is higher than the spatial sampling of the potential.…”

Multimodal Imaging from Neuroelectromagnetic and Functional Magnetic Resonance Recordings

“…A three-shell (scalp, inner and outer surface of the skull) realistic head volume conductor is represented, together with the cortical dipole layer. It has been proven that even the use of homogeneous spherical volume conductor for the head and a realistic cortical surface for the dipole layer provided more focused and detailed information than the raw scalp potential s (Sidman et al, 1992, Srebro et aI., 1993, Srebro and Oguz , 1997. However, it must be noted that the condu ctivity ratio between skull and scalp is far than 1 as assumed in homogeneous models.…”

Multimodal Functional Neuroimaging: Integrating Functional MRI and EEG/MEG

“…One may further choose to confine the dipole locations to the cortical gray matter and constrain the dipole orientations to be perpendicular to the cortical surface [26]. Alternatively but less frequently, a bioelectric source unit can also be defined as a current monopole [48], current multipole [49], or extracellular potential [50–58]. Moreover, when brain electrical activity is confined to a few focal regions, one may prefer to use a source model consisting of only a few discrete dipoles, namely the equivalent current dipole (ECD) model [59–62].…”