Confounding factors in machine-learning analysis of multicenter brain MRI data in Autism Spectrum Disorders

Autism Spectrum Disorder (ASD) is a heterogeneous  condition that affects individuals with various degrees of severity. Brain magnetic resonance imaging (MRI) represents a valuable non-invasive technique to study this condition. In literature, there is a great amount of studies based on supervised classification algorithms used to distinguish subjects with ASD from Typically Developing Controls (TDC) through an analysis of their structural brain images. However, the results obtained are controversial and usually only the studies based on small samples obtain good classification performances.