LATENCY CHANGE DETECTION IN EVOKED POTENTIALS ESTIMATION BY DIRECT LEAST MEAN p-NORM ADAPTIVE TIME DELAY

Evoked potentials (EP) have been widely used to quantify neurological system properties. Changes in EP latency may be linked to neurological in- juries. Many traditional EP analysis methods are devel- oped under the condition that the additive background electroencephalograph (EEG) noises are Gaussian. This paper proposes a latency change detection and estima- tion algorithm under the a-stable noise condition) a generalization of Gaussian noise assumption. An analy- sis shows that the a-stable model fits the noises from an impact acceleration experiment better than the Gaus- sian model. The robustness of the proposed algorithm is demonstrated through computer simulations and ex- perimental data analysis, while traditional time delay estimation algorithm fails to yield an accurate estimate of the latency changes under lower-order a-stable noise conditions.