Applying Eye Tracking to Identify Autism Spectrum Disorder in Children.

The team showed the kids a 10-second clip of a woman talking. Half of the kids had autism, half were typical.

An eye-tracking camera recorded where each child looked. No extra equipment, no long test.

Kids with autism looked less at the mouth and body. A computer used those two numbers alone.

It picked out autism vs typical with 85 % accuracy, 87 % sensitivity, and 84 % specificity.

Falck-Ytter et al. (2012) and Spriggs et al. (2015) already saw odd gaze in babies and toddlers. This paper shows the same signal still works in preschoolers and can act as a quick screen.

Harrop et al. (2018) adds a twist: autistic girls look at faces more like typical girls. If your clinic uses gaze cut-offs, you may need separate ones for boys and girls.

Shirama et al. (2016) found shaky fixation in autistic adults, while Guobin finds clear classification in kids. Same tool, different ages—together they show eye markers span the lifespan.

You can run this 10-second clip on any desk-mounted eye tracker while the child waits. If gaze time on mouth plus body falls below the cut-off, flag for further evaluation. It is fast, child-friendly, and adds hard data to parent reports.