LOCAL ESTIMATE OF SURFACE LAPLACIAN DERIVATION ON A REALISTICALLY SHAPED SCALP SURFACE AND ITS PERFORMANCE ON NOISY DATA

A new implementation of the surface Laplacian derivation (SLD) method is described which reconstructs a realistically shaped, local scalp su rface geometry using measured electrode positions, generates a local s pectral-interpolated potential distribution function, and estimates th e surface Laplacian values through a local planar parametric space usi ng a stable numerical method combining Taylor expansions with the leas t-squares technique. The implementation is modified for efficient repe ated SLD operations on a time series. Examples are shown of applicatio ns to evoked potential data. The resolving power of the SLD is examine d as a function of the spatial signal-to-noise (SNR) ratio. The analys is suggests that the Laplacian is effective when the spatial SNR is gr eater than 3. It is shown that spatial low-pass filtering with a Gauss ian filter can be used to reduce the effect of noise and recover usefu l signal if the noise is spatially incoherent.