Wavelet transform analysis of visual evoked potentials: Some preliminary results

VisualEvoked Potentials (VEP) are analyzed with the continuous wavelettransform (CWT). The modulus of the wavelettransform of the signal is calculated and represented in the time-scale domain. Based on the first-order derivative of the Gaussian function chosen as the wavelet, the modulus representation allows a clear-cut discrimination between normal and pathological cases. By processing a set of data composed of VEPs collected from 30 normal subjects and 20 patients, the method proved to be more accurate than the latency method used in hospitals. In addition, the CWT method is found to yield a fine spectral analysis of VEP data even in the presence of the background EEG signal. As a result, preliminary averaging performed on the data will use a much lower number of responses than those required by the latency method. This allows for a shorter recording session which seems to be very suitable for testing elderly people or children and even restless patients with short attention span