Finding early-emerging and persistent biomarkers that can address the vast heterogeneity in autism spectrum disorders (ASD) has been an ongoing challenge for decades. A promising avenue for exploration is imitation deficits in ASD. Typically-developing (TD) children start copying others as pre-verbal infants1–3. Imitation plays a crucial role in the development of social interactions, language and skill learning4–8. A highly active and growing line of research not only shows that children with ASD are poorer imitators than their TD peers, but also that imitation ability is strongly linked to social-communicative functioning. Therefore, imitation deficits can be a biomarker for both predicting autism diagnosis and developing intervention methods that improve social-communicative skills across the autism spectrum ranging from non-verbal individuals to those with high intellectual ability.

These converging lines of evidence support an imperative for pursuing motor imitation as a phenotypic biomarker of autism. Crucial obstacles, however, remain. Typically, imitation is quantified using methods that rely on human observation coding, which are limited in objectivity, automaticity, and scalability. Addressing these obstacles, we recently developed and validated a 3D motion-capture Computerized Assessment of Motor Imitation (CAMI) method that quantifies imitation with superior diagnostic discrimination ability as compared to traditional human coding methods. This CAMI method, which employs a brief (one-minute) and engaging video-game format, has potential to be highly scalable to clinic and home settings.