Authors: Wen, T. H., Cheng, A., Andreason, C., Zahiri, J., Xiao, Y., Xu, R., Bao, B., Courchesne, E., Barnes, C. C., Arias, S. J., & Pierce, K. Journal: Scientific Reports
Although autism spectrum disorder (ASD) commences during pregnancy and early signs are noticeable within the first months postpartum, diagnosis is based on clinical judgment and typically does not occur prior to age 4.
Validating biologically based markers of ASD would facilitate a rapid, accurate and objective early diagnosis, leading to timely treatment at an early age.
One of the well-known key features of ASD is reduced attention to social information compared to typically developing individuals. Teresa H. Wen and her colleagues from the Autism Center of Excellence, University of California, conducted a study with Karen Pierce. They set out to validate an eye tracking based biomarker in the largest eye tracking study of ASD to date. The study found that during a social attention preference test, both a simple approach using a single met- ric (percent fixation) and a computationally advanced approach based on multiple eye tracking metrics and machine learning algorithms had high accuracy in identifying ASD.
The study included 1863 children separated into diagnostic groups based on the most recent diagnosis. This allowed for comparison among children with ASD and others (ASD features, global developmental delay, language delay, typically developing, typical toddlers with an ASD sibling, other). The children were also diverse in terms of age, sex, race, and ethnicity.
The test consisted of two simultaneously presented images, a geometric and a social image. Eye tracking data was collected using Tobii T120 and the Tobii Studio software. The main metric used was percent fixation, that is, the amount of time a child spent looking at the social image. In addition, the number of saccades between the two images was considered.
The study found that the test used can rapidly and accurately detect autism before the 2nd year of age in a subset of children, and hence serve as a biomarker for a unique ASD subtype in clinical trials. The authors combined diagnostic evaluation by licensed psychologists and a highly repli- cable quantitative eye tracking evaluation of individual´s attention to social images. With a very low false positive rate, and high specificity the test may be an excellent 2nd tier screen or diagnostic tool.
This study showed the potential of eye tracking metrics in becoming a key early biomarker of severely impacted ASD toddlers, allowing them to be diagnosed and commence treatment as early as possible, raising the chances of better outcomes in their development, and life quality.