Auditory Evoked Potential (AEP) is an electrophysiological signal elicited from the brain while an auditory stimulus is presented in a time-locked manner. Previous studies reported the occurrence of AEPs in response to different audio stimuli during the time interval up to 600 ms from the onset of the stimulus. In this paper, we report the occurrence and detection of long-latency AEPs, around 1.5 s after the onset of the stimulus, from the Electroencephalogram (EEG) acquired when the participants listen to voice clips of their own-name. The trial averaged AEPs of 24 participants who listened to audio stimuli of their own-name and familiar-names clearly showed the occurrence of a distinct negative peak around 1.5 s after the onset of the own-name stimulus in Fp1 and F7 EEG electrodes. No such long-latency AEPs were observed when the participants listened to familiar-name stimulus. We classified the own-name and familiar-name averaged AEP responses using a quadratic kernel support vector machine (SVM) classifier. The leave-one-participant-out cross-validation was performed, and the classification accuracy achieved is 80.56 % (corresponding to averaging of 8AEP responses) with a sensitivity of 68.52 % and a specificity of 92.59 %. Our study reveals the occurrence of a distinct long-latency AEP around 1.5 s for own-name stimulus and a higher discrimination ability of participants in discarding non-self-referential stimuli (familiar-name) compared to self-referential stimuli (own-name). The method can be applied for the off-line analysis of own-name AEPs to diagnose mild cognitive impairments and also developmental disorders like autism.
Read full abstract