摘要
It is well known that averaging is an effective method for extracting the periodic signals which are submerged in the background-noises. The Time-locked Average (TLA) with reference to the stimulated time has been a traditional way to pick up the evoked potentials from the EEG background. The premise of TLA method supposes that the latencies of evoked peaks appear relatively regular as compared with the time of stimulus. But experiment results and clinical observations tell us that the latencies of evoked potentials are not invariant rigidly after stimulations. Hence, the TLA method cannot perfectly reconstruct
