Dimitrov L I
Institute of Information Processing, Austrian Academy of Sciences, Vienna, Austria.
Hum Brain Mapp. 1998;6(4):189-202. doi: 10.1002/(SICI)1097-0193(1998)6:4<189::AID-HBM1>3.0.CO;2-#.
A method for representing the results of statistical EEG analysis in a visually comprehensive and concentrated way is presented. It consists of a combination of scattered data interpolation, texture mapping, and volume rendering techniques. First, starting with an EEG study of the proband in question, a smoothly interpolated rectangular map of the EEG findings in the (phi, theta)-domain (phi and theta denote spherical coordinates) is calculated. Then, using the data from an accompanying MRI scan, volume-rendered representations of the brain with a pseudo-colored cortical surface, vividly depicting the precalculated EEG-activities, are produced. Custom-developed software was used for interpolating the EEG maps from the scattered data samples, as well as for mapping and rendering them together with the brain volume data. The algorithm which achieves the scattered data interpolation is a genuine development employing Delaunay triangulation and a modification of the well-known rotation-invariant Phong interpolation. The pictures produced by the presented method are currently being utilized in neurophysiology for detecting correlations between EEG-parameters, cortical morphology, and underlying mental activities.