Jérémy Lamontagne¹ (), Francis Gingras^1,2, Laurianne Côté¹, Alexis Bellerose¹, Chanelle Demeule¹, Caroline Blais¹, Daniel Fiset; ¹Université du Québec en Outaouais, ²Université du Québec à Montréal

A previous study from our laboratory revealed the N170 amplitude correlates with diagnostic information for face recognition. The present study aims to replicate these results and ascertain the task specificity of the effect by extending it to a repetition suppression design. Six participants (3 women) completed 6000 trials of a same/different identity task while EEG data was being recorded. The stimuli were grayscale faces of 84 celebrities (42 women), with 3 different images per identity. Each trial presented two successive faces,one of which was filtered using Bubbles, both presented for 1s and preceded by a fixation cross shown for 1-1.4s. A classification image (CI) representing the information leading to accurate categorization was computed, and trials were then grouped in 6 levels of 1000 trials, based on their amount of available diagnostic information. Group analyses reveal a linear positive relationship between the quantity of diagnostic information and the N170 amplitude on the first face, F(4) = 72.16, p = .001 as well as a negative relationship between the diagnostic information group and the amplitude of the N170 for the second face, F(4) = 15.52, p = .017 and F(4)=6.56, p = .062 respectively for same and different identities. Additionally, group CIs showed that the N170 amplitude was driven by the presence of the eyes and mouth in trials containing highly diagnostic information. This study suggests a direct link between featural information processing and the facial repetition suppression found on the N170. The results also suggest that the N170 is sensitive to identity, demonstrated by a better fit of the regression model for trials in which the identity was repeated.