Predictive effect of digital phenotype on the diagnosis of Autism Spectrum Disorder: A systematic review

ABSTRACT

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder characterized by high variability in symptomatic profiles and a high prevalence rate. However, there are no biomarkers, and the diagnosis is primarily clinical. The objective quantification of possible phenotypes may help clarify the disorder’s etiology. In this context, the objective of the present study was to conduct a systematic review to assess the capacity of the digital phenotype to predict early diagnosis in autism. Searches were conducted in the Scopus, PubMed, Web of Science, and PsycINFO databases without time restrictions. A total of 345 articles were identified, of which only 8 met the eligibility criteria. The data found indicated that characteristics such as head and eye movements and recordings of social interactions captured with portable devices may help differentiate typically developing children under the age of six from those with atypical development. However, current procedures are in early stages, potentially generating false positives or negatives. Therefore, although promising, the results should be analyzed with caution and discretion.

KEYWORDS
Autism Spectrum Disorder; Clinical diagnosis; Digital health.