Brief Report: Classification of Autistic Traits According to Brain Activity Recoded by fNIRS Using ε-Complexity Coefficients.

Dahan et al. (2021) asked adults to tap in time with a partner while wearing fNIRS caps. The caps shine light through the scalp to track brain activity.

They used a math tool called ε-complexity to turn the light signals into a single score. The goal was to see if this score could sort people into high or low autistic-trait groups.

The ε-complexity score cleanly split the two groups. High-trait brains looked simpler and more repetitive during the sync task.

A computer could tell the groups apart with high accuracy. The whole test took under five minutes.

Vabalas et al. (2016) and KAgiovlasitis et al. (2025) saw the same social-attention drop using eye trackers. Anat’s fNIRS result adds brain data, showing the trait marker holds across cultures and tools.

Chien et al. (2025) also used fNIRS but found less frontal activity in autistic adults. Their result seems opposite—Anat saw simpler signals, Yi-Ling saw weaker signals. The difference is task: tapping together versus hard word lists. Social sync lights up different circuits than solo word work.

Sabatino et al. (2013) mapped these circuits years ago with fMRI. Anat’s quick fNIRS metric now gives clinicians a cheap, five-minute update of that early picture.

You can’t bring an fMRI scanner to clinic, but you can keep a portable fNIRS kit in a backpack. If a five-minute tapping game gives a reliable trait read-out, you can screen adults while they wait, adjust social-skills goals on the spot, or track progress after intervention without long questionnaires.