Prediction of brain electrical activity in epilepsy using a higher-dimensional prediction algorithm for discrete time cellular neural networks (DTCNN)

Several investigations have shown that a higher-dimensional nonlinear signal analysis can contribute to the problem of detecting precursors for impending epileptic seizures in electroencephalographic recordings. In previous work we analyzed brain electrical activity using Volterra systems as stated in M. Schetzen (1980) and CNN in L. O. Chua (1998). The outline of this paper is to propose a higher-dimensional DTCNN prediction algorithm. First results are given for the long term recording of brain electrical activity.