Machine-learning techniques in the search for neuroimaging biomarkers of autism spectrum disorder

by Alessandra Retico (PI)

Tuesday 17 Dec 2013, 15:00 → 16:30 Europe/Rome

131 (INFN edificio C)

Machine-learning and pattern recognition techniques applied to neuroimaging data are emerging as useful tools in the identification and characterization of a wide range of neurological and psychiatric disorders. They aim to provide clinically useful biomarkers of pathologies. Whereas the discovery of valid biomarkers for autism spectrum disorder still remains a hope, information gathered from neuroimaging data can help in autism characterization. We employed support vector machines to detect structural brain alterations in children with autism spectrum disorder and to evaluate whether the brain abnormalities exhibit gender-related morphometric differences.