Yale-led EEG study finds less refined face-processing signals in autistic children

Researchers led by Yale's James McPartland used whole-scalp EEG and machine learning to study face perception in autism, in work recently published in Nature Mental Health and summarized by SciTechDaily. The team drew on Autism Biomarkers Consortium for Clinical Trials data and recorded activity from 128 electrodes placed around the scalp as children viewed faces and objects. SciTechDaily reported measurements from approximately 400 autistic children and found face-related neural signals were less distinct than in neurotypical peers. In neurotypical children, face-related signals became increasingly specialized with age, while autistic children showed little comparable refinement. The study frames that altered trajectory as one clue to why interpreting faces can remain difficult for some autistic people. The paper lists DOI 10.1038/s44220-026-00672-y.