Learning by Research

Thesis Title: Predictive mechanisms during the processing of socio-emotional stimuli in autistic adults.

Recent theories suggest that autism characteristics could be explained by atypical predictive processing. However, these theories lack empirical evidence. This thesis considers the existence of atypical visual predictive processing of faces among autistic adults. This work refers to the model of Bar et al. (2006), which proposes that visual predictions triggered by low spatial frequencies improve visual recognition. Sex differences are also examined to better characterize the neurocognitive specificities of autistic men and women. Results from electrophysiological and behavioral experiments suggest that autistic individuals have atypical low-spatial-frequency-dependent predictive processing, but mostly during pre-attentional face processing. During attentional processing, cerebral responses seem more intact. Additionally, autistic women demonstrated an intermediate profile, situated between autistic males and control participants. Finally, our results highlight difficulties among autistic adults in updating their predictions according to the specific context, when engaging in emotion recognition. Overall, these findings suggest that high-functioning autistic adults have atypical predictive processing, but further studies and analyses are required to better understand the mechanisms behind these specificities.

Key Words: autism, predictive brain, spatial frequencies, EEG, sex differences

Doctoral School: ED ISCE - Health, cognition and environmental engineering
Research laboratory: Laboratoire de Psychologie et Neuro Cognition (LPNC - CNRS/UGA/USMB)
Thesis supervision: Martial MERMILLOD, Marie GOMOT and Frédéric DUTHEIL

> To find out about all the 2023 thesis prizes