Objective: We studied the event-related potentials elicited by categorical matching of faces. The purpose was to find cortical sources responsible for face recognition and comparison.Methods: Nineteen healthy volunteers participated in the study. Each trial began with one of the two cues (S1) followed by consecutive pictures (S2 and S3). Each picture was a photograph of a familiar face with a superimposed abstract dot pattern. One cue directed attention to compare faces and another to compare patterns. 128-channel electroencephalogram was recorded. Spatio-temporal multiple dipole source models were generated using Brain Electromagnetic Source Analysis 2000, for the window of 80 -600 ms from S3 onset.Results: The obtained model for face recognition and comparison contained 8 dipoles explaining 97% of grand average and about 90% of individual data and showing temporal and spatial separation of sources: in the frontal region, in the occipital cortex, and in the bilateral medial temporal and inferotemporal regions. Different faces elicited larger components than same person's faces around 400 ms, mainly explained by frontal dipoles.Conclusions: The sources in our models estimate the activity common for both Face task conditions (the recognition of a familiar person) and also differential activity, related to the match/mismatch item processing.

No abstract

The contingent negative variation (CNV) is known as an electrical manifestation of expectancy, readiness and attention. The purpose of this study was to evaluate whether slow potentials recorded from the human scalp during the expectation to perform different tasks or during the expectation of the second stimulus in the pair demonstrate specificity regarding the type of visual stimuli. Participants were nineteen healthy adults. Each trial began with one of the two cues (S1) followed by consecutive pictures (S2 and S3). Each picture of this pair was a photograph of a familiar face with superimposed abstract dot pattern. One cue directed attention to compare faces and another to compare patterns. EEG was recorded with a 128-channel net. Slow potentials following S1 and S2 presentation (SP1 and SP2) are reported here for Face and Pattern tasks. The duration of each segment of the trial was 1500 ms from the stimulus onset, and here the last 1000 ms was analyzed. Statistical comparisons of amplitudes and current density mapping were performed. Within the same task, SP1-SP2 differences were complex, implying multiple dissimilarities in the processes during the two expectation periods. Slow potentials in SP1 differed between the tasks mostly in the anterior recording sites in 700-1500 ms. These latencies corresponded to the late CNV (or stimulus preceding negativity, SPN) showing task-specificity during SP1. For SP2, the differences were located centrally in 500-900 ms, the period that reflects the superimposition of processes generating the early CNV and the late positive complex. Task-specific differences appeared only in SP1.