Adaptive estimation of evoked potential latency under low SNR conditions

The authors investigate the effectiveness of an adaptive time delay estimation algorithm when applied to estimate latency changes of somatosensory evoked potentials (SEPs) under poor signal to noise ratio (SNR) conditions. Such an estimation technique is to be used to study the effects of whole-body indirect impact acceleration of humans and to develop safe-exposure limits. Using an SEP template obtained through human experiments, the authors artificially delay the SEP signal according to the latency change pattern observed in animal experiments. Background EEG signals are added to simulate the low SNR conditions. Simulation results show that the adaptive time delay estimation algorithm is very robust to the additive noise. Latency changes in the SEP signal can be reliably detected and estimated with SNR as low as -40 dB