Drawing reveals hallmarks of children with autism

Autism spectrum disorder (ASD) as a kind of mental disorder, has become an internationally recognized serious public health problem. Paintings of autistic children have not been compared systematically to those from Typically Developed (TD) children. In this work, we construct an ASD painting database which contains 478 paintings drawn by ASD individuals and 490 drawn by the TD group. Through subjective and objective analysis, some significant hallmarks, such as structuring logic, face, repetitive structure, composition location, edge completeness, etc. are found within the ASD paintings. We further train a classifier of ASD and TD painters using those extracted features, which shows encouraging accuracy as a potential screen tool for ASD. This work sheds light on understanding the uniqueness of autistic children through their paintings. The database will be released to the public.